HH11SK
KIRTLANDIA
CLEVELAND, OHIO NUMBER 39
•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 be-
tween collections of short papers and single issue-length studies.
Kirtlandia is distributed by The Kent State University Press, Kent, Ohio 44242.
Copyright © 1983 by The Cleveland Museum of Natural History.
Kirtlandia is abstracted in Zoological Record and Biological Abstracts and in-
dexed in Bibliography and Index of Geology.
ISSN: 0075-6245
KIRTLANDIA
THE CLEVELAND MUSEUM OF NATURAL HISTORY
Cleveland, Ohio
December 1983
Number 39
RECENT EXCAVATIONS AT THE EDWIN HARNESS MOUND,
LIBERTY WORKS, ROSS COUNTY, OHIO
N’OMI GREBER
with James B. Griffin, Tristine Lee Smart, Richard I. Ford, Orrin C. Shane
III, Raymond S. Baby, Suzanne M. Langlois, Stephanie J. Belovich,
David R. Morse, Kent D. Vickery
Abstract
Analyses of data recovered during the 1976-1977 excavations at the
remnant of the Edwin Harness Mound, a major Hopewell site in Ross
County, Ohio, are presented, with some reanalysis of data from earlier
(1840-1905) excavations. A large structure has been identified at the base
of the mound. Botanical, faunal, and soil analyses indicate that the physical
environment in the central Scioto Valley near a.d. 300, when the building
was in use, was similar to that found in the area in the late eighteenth
century. Detailed studies of ceramics, lithics, and human skeletal remains
are presented. The varied activities which took place within and near the
major structure, and implications for interpretations of materials from
Seip (Pricer) Mound on Paint Creek, are discussed.
Kirtlandia No. 39
0075-6245/83/1983-0039 $9.95
Copyright © 1983 by The Cleveland Museum of Natural History
• •
. ■ ■
'
■
CONTENTS
Figures 5
Tables 5
Acknowledgments 7
1. Introduction 11
2. The Excavations 13
3. The Site 23
4. The Ceramic Complex — James B. Griffin 39
5. Plant Remains — Tristine Lee Smart and Richard /. Ford 54
6. Vertebrate Faunal Remains — Orrm C. Shane HI 59
7. Analysis of Human Skeletal Material
Ohio Historical Society Collections — Raymond S. Baby and Suzanne M. Langlois 62
Cleveland Museum of Natural History Collections — Stephanie J. Belovich 62
8. Mollusc Identification and Analysis — David R. Morse 66
9. The Flint Sources — Kent D. Vickery 73
10. Discussion and Conclusions 86
FIGURES
1. 1 General view of excavations, 12 August 1976 1 1
1.2 General view of excavations, 21 July 1977 12
1.3 Base of heavy gravel wall (Feature 1) 12
2.1 General location of site 13
2.2 Edwin Harness excavation grid, 1976 15
2.3 Edwin Harness excavation grid, 1977 16
2.4 Post holes as first found 1976-1977 17
2.5 Sketches of site profiles 18-19
2.6 Remains of major activity floor and other features found at base of mound 20
3. 1 Cuts made by Mills in 1905 through the original floor layers 26
3.2 Floor plan of Harness Big House 28
3.3 Location of post holes with surrounding stains 30
3.4 Examples of post holes with surrounding stains 31
3.5 Pipe fragment found by F. W. Putnam 32
3.6 Sketch profiles of pit features 35
4. 1 Hopewellian and Check-stamped sherds 40
4.2 McGraw Cordmarked, McGraw Plain, and unidentified sherds 41
4.3 Turner Simple Stamped and unidentified simple stamped sherds 46
4.4 McGraw Plain and McGraw Cordmarked sherds 47
9. 1 Areas of outcrop of flint varieties 75
10. 1 Estimated floor plan of Seip Big House 88
10.2 Radiocarbon dates plotted as mean calendrical years 90
TABLES
2. 1 Summary of Features 14
3.1 Non-structural Posts Associated with Edwin Harness Big House 29
3.2 Radiocarbon Dates 34
4.1 Summary Tabulation of Harness 1976-1977 Pottery Collection 39
4.2 Griffin-Morgan Pottery Count of Ohio State Museum Harness Collection 42
4.3 The Occurrence of Turner Simple Stamped B in Ohio, based on Prufer (1968) 43
4.4 The Occurrence of Turner Simple Stamped A in Ohio, based on Prufer (1968) 45
5.1 Charred Posts from the Edwin Harness Mound Floor 55
5.2 Charcoal from the Edwin Harness Mound Features 56
5.3 Carbonized Seeds and Faunal Remains from the Edwin Harness Mound Botanical Samples 58
6. 1 Frequency of Identified and Unidentified Bones by Class 59
6.2 Frequency of Identified Vertebrate Remains from Edwin Harness Mound 60
7.1 Edwin Harness Mound Human Skeletal Material Excavated 1896-1905 63
7.2 Edwin Harness Mound Human Skeletal Material, non-cremated, 1977 64
8.1 Summary of Mollusc Taxa Identified, Edwin Harness Mound 67
8.2 Comparisons of Dimensions Melongenidae — Conch Shells 68
9. 1 Provenience of Debitage from Edwin Harness Mound 77
9.2 Correlation of Flake Type and Technique of Flake Removal, Edwin Harness Mound 78
9.3 Correlation of Debitage Type and Flint/ Chert Variety, Edwin Harness Mound 79
9.4 Correlation of Technique of Flake Removal and Flint/Chert Variety, Edwin Harness Mound 80
9.5 Frequency Distribution of Flint/ Chert Varieties for Cultural Collection and Comparative Local Scioto
River Gravel Pebbles 82
10.1 Values of Ranksum F for Seip Mound 1 (Pricer) 89
. ' . ■ . . ; .
■*
.
ACKNOWLEDGMENTS
It was exciting and somewhat unbelievable in Novem-
ber 1975 to see clear evidence that sections of a major
Hopewell mound were still unexcavated after 1 30 years of
intermittent digging. This was my first look at a very level
Edwin Harness Mound. Many people shared the effort
and work, which have resulted in significant new informa-
tion from this classic site. The work was made possible by
Robert Harness, who not only gave us permission to ex-
cavate on his land, but also helped in many practical mat-
ters. These ranged from letting me use the top of his trac-
tor as a photography tower in November 1975 to
backfilling over many refilled and cemented post holes af-
ter the 1977 field season. He and his wife, Marilyn, were
most hospitable to their many digging guests.
The first season of field work was under the auspices of
the Ohio Archaeological Council, whose president, David
Brose, and vice-president, Orrin Shane, were instrumen-
tal in initiating the support of the council and in encourag-
ing me to work at the site. In the field there was a perma-
nent staff of three — Wesley Clarke, Michael Hambacher,
and myself — and a total of 1 20 volunteers who worked at
the site for various lengths of time from June 7 to August
14, 1976. Funding for this season was provided in part by
a matching grant to the council from the Ohio American
Revolution Bicentennial Advisory Commission. The
match for this grant was more than met with the dona-
tions received by the council from Ohio corporations and
private sources. At the end of this season, when it became
evident that there was more detail intact than could pos-
sibly be completed within the original time limit, Mr.
Harness agreed that 1 could seek funds for a second
season.
Major funding for this second season, which was
under the auspices of the Cleveland Museum of Natural
History, came from the National Science Foundation.
Field school students worked with a core crew of expe-
rienced first season veterans, making a combined total of
43 crew members for the second season. Wesley Clarke
was my chief assistant. Franco Ruffini joined the supervi-
sory staff with special responsibilities for overseeing stu-
dents. Michael Hambacker did what might seem impos-
sible by maintaining an excellent, much needed, and often
used photo record of our work while still demonstrating
his exemplary field techniques. The second season began
June 14 and ended August 26, 1977.
Both summers, Alva McGraw gave us the use of one of
his farmhouses as well as help with innumerable practical
problems and invaluable moral support. We, as all ar-
chaeologists in Ross County, give him special thanks.
The curation and analysis began in the field in 1976.
Many hours were spent “after hours” cataloguing, clean-
ing, water screening, etc. Jane Busch was the chief field
cataloguer in 1976; Shaune Skinner in 1977. The work
continued in the winter of 1976-1977 at the museum with
Dennis Griffin. In the fall and winter of 1977-1978 Eloise
Gadus supervised the laboratory curation and was a
general research assistant. Museum volunteers and stu-
dents from Cuyahoga Community College, Case Western
Reserve University, and Cleveland State University
worked on various laboratory connected projects. The re-
sults of specialized analyses presented herein were per-
formed at the respective laboratories of the various con-
sultants noted in the chapter headings. Even with excellent
support from public foundations and with private dona-
tions of funds and services, choices had to be made con-
cerning which samples would be analyzed first, which
would wait. Those samples awaiting analysis are stored
and ready for new resources, new analytic techniques, and
new ideas.
The integration of information from earlier excava-
tions was an essential part of our work. Martha Otto and
Bradley Baker of the Ohio Historical Society and Stephen
Williams and Sally Bond of the Peabody Museum, Har-
vard University, were, as always, gracious and helpful in
allowing access to their collections and documents.
The maps and drawings of the field data were prepared
for the present report with patience and skill by Mark
Schornak. The chapters without author designations
were written by myself.
Although I have personally gained knowledge from
each archaeological excavation in which I have partici-
pated and know that each site represents a unique and
irreplaceable piece of prehistory or history, Edwin Har-
ness is a special place for me. To each and every one who
shared the planning, excavation, curation, and analysis I
give personal and professional thanks. Without your help
we would still only be able to speculate on the existence of
the Harness Big House.
N’omi Greber
THE EDWIN HARNESS MOUND EXCAVATION
Field Personnel, 1976
Director: N’omi Greber
Field Supervisors: Wesley S. Clarke
Michael Hambacher
Participating Supervisors
David Brose
Patricia Essenpreis
Marie Freddolino
Michael Pratt
Jeffrey Reichwein
David Stothers
Kent Vickery
Kelly Adkins
Robert Asumc
Katie Banzhaf
David Barna
Sandra Barna
Naomi Bauer
Tom Berreis
Robert Blickensderfer
Nancy Bocash
Tom Bogus
Jonathan Bowen
Ken Bowden
Helen Bradburn
Ken Broberg
Jane Busch*
Nancy Cameron
Ricky Cibulskas
Eileen Closs
Elizabeth Cole
Cindy Cubbison
Jane Claire Deward
John Doershuk
Jim Donald
Marion Donald
Sally Donald
Jim Donald, Jr.
Bob Drozda
David C. Ebosh
Jean A. Ebosh
Fran Erwin
Dale Fable
Brad C. Featherstone
Bill Fenton
Charles Ford
Jon Paul Fry
Bob Genheimer
Leanna Geriak
Mike Glenn
Amy Glin
* Adjunct Field or Laboratory Assistant
General Crew
Richard S. Gray
Lisa Greber
Rebecca Greber*
Dennis Griffin*
Dan Grossman
Joseph T. Hannibal
Bonnie Hannon
Dan Harmon
Crickette Harrell
Janet Hart
Laura Havasi
Chuck Heath
Neil Henderson
Connie Holden
George L. Holler
Frank Huntley
Frank Johnoff
Winifred Kelley
Debby Klein
Phil Kleinhenz
Lois Lambert
Peter Lemmerman
Timothy Light
Janet Lipstreu
Lisa Littman
Karen Lord
Bruce Leland Markley
Mike Marmostein
Harry Martin
Mark McDonald
James Trent Metcalf
Barbara Mihuta
Laura Mihuta
Dave Miller
Dward A. Moore*
Jeannine Moore
Dave Morrison
Glen W. Nyhart
Bradley A. Oen
Special Consultants
Raymond Cotrill
James B. Griffin
Don Grimes
James Kerr
Alva McGraw
James Marshall
Maynard Munsing
Orrin C. Shane III
Jose R. Oliver
Christine Opfer
W. Kevin Pape*
A. Lowell Randall
James Reckley
Frank Reiger
Alison Roberts
Betty Rock
Patricia Rubright
Franco Ruffini*
Ronald A. Salupo
Hattie Sells
Ruth Sheard
Michael Shore
Jon Singer
Clement Skehan
William Snodgrass
Kimberly Jo Solsman
Dana G. Staley
Patrick Steiner
Merry Steward
Ruth Streicker
Ted Sunderhaus
Helen Swanner
Christine Tailer*
Sue Tituskin*
Dale Thomas
Bryan Tupper
Millie Tupper
William D. Ullery
John Walters
Sally Warrick
Carmen Anna Weber
Matthew J. Weitendorf
A1 Wilson
Susan B. Wiseley
David B. Woodmansee
Nora Wright
THE EDWIN HARNESS MOUND EXCAVATION
Field Personnel, 1977
Director: N’omi Greber
Field Supervisors: Wesley S. Clarke
Franco Ruffini
Participating Supervisors: Dward Moore
Arthur Saxe
Mark Bir*
Jonathan Bowen
Robert Buck*
Daniel W. Bull*
Lois Cahill*
Karen Ciatyk*
Hattie S. Clarke**
Ann C. Cramer
Scott Detroiler*
John Doershuk
Marjory Forbes-Howard*
Lois E. Fultz*
Eloise Gadus*
Kathy Gladwin*
John T. Goerlich*
Linda Grand*
Richard S. Gray**
Rebecca Greber**
Dennis Griffin
Susan Hammond*
Field Photographer: Michael Hambacher
Janet Hart
Elizabeth Ippolito*
Mary Klock*
Britt S. Krebs**
Janet Lipstreu**
Caroline McLeod*
Bradley A. Oen**
W. Kevin Pape
Laurie Patti*
Jim Retzler*
Alison Roberts**
Daniel Simon**
Shaune M. Skinner
Roy Walsh*
William Snodgrass**
Christine Tailer
Robert E. Thornsberry*
Sue Tituskin
Bryan Tupper**
Sally Warrick**
Russell Weisman*
* Field School Student
** Volunteer
■
.
■
'
■
1 INTRODUCTION
The following site report on the 1976-1977 salvage ex-
cavations of the remnant of the Edwin Harness Mound,
Ross County, Ohio contains types of data which could
not have been included in the reports of earlier expedi-
tions. Technology has grown rapidly since 1907 when the
last report was published. The application of these new
analytic techniques was one of the major purposes of our
excavations. (Figs. 1.1, 1.2). A second major purpose was
to salvage data on the manner in which space had been
used by the inhabitants of the site before the mound had
been constructed. This data had not been available be-
fore, partly because of the checkered series of excavations
and the missing field records from major sections of these
excavations. These earlier excavations at Harness and
other major sites in southern Ohio established that differ-
ent cultures were represented among the “mound build-
ers.” Edwin Harness was assigned to Hopewell on the ba-
sis of artifact similarities (Mills 1 907 : 1 9 1 ). These artifacts
do show the high artistic and technical talents of the indi-
viduals who made them, and we still restudy them because
they are part of the basic data which allow new ideas to be
developed and refined. Less exotic, perhaps, but equally
important to this data base is the context in which these
objects were found. Fortunately some of the gaps in the
knowledge of the contexts at Edwin Harness can now be
filled in, both with newly found old records (e.g., Murphy
1978) and with new field data (Fig. 1.3).
The emphasis of the most recent field work at the site of
the Edwin Harness Mound was context; the following
chapters report on the data we found and on the special-
ized studies of that data which have been completed to
date. The design and construction of a major Hopewell
civic-ceremonial building, parts of its contents, and the
mound that covered it are described. We also now have
specific data on environment, subsistence, and chronol-
ogy from this classic site.
References
Mills, William C.
1907 Explorations of the Edwin Harness mound. Ohio Ar-
chaeological and Historical Quarterly 16:1 13-193.
Murphy, James L.
1 978 William C. Mills’s notes on the Edwin Harness mound
excavation of 1903. Ohio Archaeologist 28(3):8— 1 1.
Fig. 1.1. General view of excavations, 12 August 1976. The remains of the heavy gravel wall
(Feature 1) ringed secondary mound fill and defined the major activity floor. View looking west.
Fig. 1.2. General view of excavations, 21 July 1977. View looking south.
Fig. 1.3. Base of heavy gravel wall (Feature I), west side. Unit N537.5 E485, 28 July 1976.
2 THE EXCAVATIONS
Previous Excavations
Putnam began his trench on the north end, inside the
heavy cobbles which ringed the mound, and gradually cut
down towards the floor. He widened the trench when he
reached the first charred areas (Greber 1 979: Fig. 6.4).
Moorehead continued south from the end of Putnam’s
trench. He did not dig from the surface to the floor but
used tunneling techniques to reach the south end of the
mound (ibid.). Mills began at both the south and southeast
edges of the mound in his first season. He spent 12 days in
1903 digging to and through the floor of the mound to the
approximate end of Moorehead’s tunnels. When he re-
turned in 1905, he began on the northeast side of the
mound east of Putnam's trench and finished on the west
side of this trench. The relatively intact main floor extend-
ing east-west near the N525 line and the similarly rela-
tively extensive remains on the west side of the structure
are in areas uncovered, if at all, at the end of each of
Mills’s two expeditions.
Squier and Davis describe the mound as egg shaped in
plan, with the larger end at the north. The height of the
mound varied from approximately 20 ft (6 m) in the north
Fig. 2. 1. General location of site. Topographic contour intervals, in feet, were taken from
U.S.G.S., 7.5 minute series quad, Chillicothe East.
The Edwin Harness Mound (lat. 39°15.'4N, long. 82°
52.'6W) was the largest of the 14 mounds associated with
the Liberty Earthworks, Ross County, Ohio (Fig. 2.1).
Excavations of this mound, as for many of the large clas-
sic Hopewell mounds in Ohio, began in the early part of
the nineteenth century and have continued intermittently
to the present. The results of the first two test shafts in the
mound were reported by Squier and Davis in 1848. Sub-
sequent digging was done by local schoolboys; by Freder-
ick W. Putnam ( 1 885) of the Peabody Museum, Harvard;
by Warren K. Moorehead (1897) and William C. Mills
(1907, 1903) of the Ohio Historical Society. Mills quotes
in some detail from the reports on earlier excavations and
from Putnam’s account of the materials found by the
schoolboys. One of the pits dug to the floor of the mound
by the boys was next to Squier and Davis’s Pit B. This
pit, which has been recorded by all the excavators except
Moorehead who worked only at the south end of the
mound, has provided one of the major reference points
for integrating data obtained over a 130 year time span
(see Table 2.1:Fea. 24).
14
N’OMI GREBER
No. 39
TABLE 2.1
Summary of Features
Major mound stratum: 1,41
Local mound stratum: 20, 21, 43
Floor stratum: 3, 3C, 33, 50, 65
Local floor stratum: 3B, 72, possibly 39, 96
Main activity floor
South Section
Shallow burned area: 82B, 92
Disturbed grave: 70, 82A; immediately west, 71
Middle Section
Shallow burned area: 47
Disturbed grave: 83, 84
Heavily burned area: 36
Pit: 19, 30, 79, 89
North Section
Shallow burned area: 34, 78; immediately west, 32
Disturbed grave: 27, probably 28, 29; all immediately west
Burned log: 26, 35
Probably heavily burned prepared clay basin: 18
Pit: 91
Prepared basin: 62
Depression: 54
Log mold: 48
Pit: 17
Post Hole: 11, 13, 14, 38, 40. 42, 64, 66, 67, 87
Small trench with stake holes: 59, 85, 88, also in 22
Shallow burned area or deposit of burned materials: 6-10, 12,
16, 31,44-47, 49, 51-53, 76
Historic pit in main floor: 24, 25, 77, 93, 94, 95, probably 58, 86, 90
In prior backfill
Bone concentration: 2, 4, 5, 23, 61, 74, 75
Log: 15
Below main activity floor
Shallow deposit of burned materials
On Feature 65: 54A, 97
On Feature 50: 57, 73
Associated with outer areas
Burial: 56, 60
Major stratum: 69, 69A
Deposit within 69: 68, 80, 81
Deposit within 69A: 55
Pit: 63
Tap Root: 37
to approximately 1 1 ft (3.3 m) at the southern end. Mills
reported the maximum height at the north as slightly less
(16.75 ft, or 5. 1 m). The recorded decrease is likely due to
Putnam’s cut through the crest of the mound and some
erosion. The site was backfilled after Mills’s more exhaus-
tive excavations to approximately 5 ft (1.5 m) in height.
Based on photographs and descriptions of Mills’s work
and on our recent excavations, it seems likely that the
backdirt was deposited behind the excavators as they
worked section by section. The outer boundary of the
mound described by all work prior to 1977 was the heavy
cobble and bedded stone mantle shown in Squier and Da-
vis’s estimated cross section (1848:Fig. 6f). Major por-
tions, if not essentially all, of the lower sections of this
wall were left intact until heavy power equipment was
used in 1975 to level the remnant of the mound and the
backfilled excavations to the general ground level. Data
from previous work, though varied in their documenta-
tion, have been integrated into the already completed re-
ports on special aspects of the research goals (Greber
1979; Gadus 1979; Greber, Davis, and DuFresne 1981;
Bender, Baerreis, and Steventon 1981) and into the pres-
ent site report.
Field Methods 1976-1977
During the 1976 season, the remains of the very bottom
of the outer stone wall (Feature 1) which had surrounded
the mound were defined. Hand excavated trench units
were concentrated in the least disturbed sections of the
site on the west and in the north in order to obtain undis-
turbed prehistoric ecofacts and as much stratigraphic in-
formation as possible. These units were taken down
through all cultural deposits into the underlying natural
soils. Two large areas in the central portion of the site
were cleared using power equipment (Fig. 2.2). During
the second season, additional hand excavated trench
units were placed on the east and in the south, again to give
stratigraphic information, particularly as a guide in using
power equipment. Backhoe trenches were dug to aid in
determining the horizontal extent of the floor strata (Fea-
tures 3, 3B, 3C, 33, 39, 50, and 65) and mound strata (Fea-
tures 4 1 , 69, and 69 A) found both by the hand excavations
and the first backhoe trenches. These trenches were num-
bered as they were dug. After the hand excavation trench
units were completed, the interior of the site (within Fea-
ture 1) was cleared by the backhoe so that the building
patterns and also the activities on the remaining floor out-
side the building but within the stone wall could be de-
fined by final hand clearing (Fig. 2.3). At the request of
Mr. Harness the stones of this wall were removed and
kept separate from the other backdirt, thus simplifying
future farming of the land. Mr. Harness himself very
kindly did the backfilling. A 2.5 m unit which contained
original mound strata just within this wall was left unex-
cavated, and it is hoped that sections of this will remain
below the depths reached by farming equipment.
At first, all excavated soils were dry screened through
!4 in. mesh. As it became apparent that the mound load-
ings at the remaining mound levels were culturally sterile
and that archaeological redeposited materials were of
secondary importance, general dry screening was discon-
tinued. When significant deposits of prehistoric materials
were found in the archaeological backfill, soils were dry
screened. Flotation samples were taken from undisturbed
areas of features and post holes as well as systematically
from mound loadings in each excavation unit. All the re-
maining excavated soils from features and post holes were
EDWIN HARNESS 33Ro22
Excavation Grid, 1976 #
Fig. 2.2. The Excavation grid was set along the major axis of the mound Grid north is 30° west of
magnetic north (1976). The units indicated by dotted lines were excavated to the base of the gravel
wall. Grid point N540 E485 is at 624 ft ( 190 m) elevation and at the Ohio Plane System Coordinate
(1,893,234; 457,750).
480 485 490 495 500 505 510 515 520 525 530
Fig. 2.3. Trench units excavated by the backhoe are numbered in sequence as dug. After completion
of the hand excavated trench units, the area within the gravel wall was excavated by a combination of
backhoe and hand clearing (see chap. 2 and Fig. 2.2). Magnetic north as of 1976 indicated.
EDWIN HARNESS 33Ro22
1976-1977 Excavations
POST HOLES
Fig. 2.4. Post holes as first found. Some variations in elevations due to priorexcava-
tion (see chaps. 2 and 3).
-DISTURBED-
$
a
a
Fig. 2.5. Sketches of site profiles. Vertical scale exaggerated for clarity. The directions indicated on
profile sections are grid W and grid N. Magnetic north ( 1976) is indicated on the location key. Section
EE' illustrates composite section through Big House; details discussed in chap. 3.
EDWIN HARNESS 33Ro22
1976-1977 Excavations
Fig. 2.6. Remains of major activity floor and other features found at base of mound. See chap. 3 for
explanation of elevation depths represented. Magnetic north as of 1976 is indicated.
1983
EDWIN HARNESS MOUND
21
water screened through fiberglass window screening. The
flotation equipment, which was constructed by museum
personnel, was basically that described by Patty Jo Wat-
son (1976). Screens of 4 mm, 2 mm, and 500 /urn were
used. Flotation was done at the Scioto River about 8 km
from the site. Water screening was done largely at the field
station.
One hundred two features were recovered. These fell
into several categories, which are tabulated in Table 2.1
and discussed in more detail in later sections.
Post holes were tabulated separately from features. Oc-
casionally a disturbed post hole was first recorded as a
feature and then given a post hole number when it was
later excavated and identified ( Fig. 2.4). Once the prehis-
toric loadings or historic backfills were removed, the
structural post holes were readily identified. The subsur-
face depths at which these holes were identified varied.
There was a downward slope built into the prehistoric
floor from the center towards the outside. However, the
major factor affecting the identification depth was the de-
gree of historic disturbance (Fig. 2.5). With very few ex-
ceptions (e.g.. Fig. 2.6: Feature 87) all major building
posts had been set deep enough to be identifiable below
extensive disturbances which were mainly due to Mills’s
excavations. Smaller, more shallowly set stakes, which
may have been placed in sections of the building floor
previously excavated by Mills, could not, of course, be
found.
Since the 1976-1977 excavations were salvaging infor-
mation from a severely disturbed site, samples for ecolog-
ical data had to be taken from a “what’s left” universe. All
samples for soils and pollen analyses were taken only
from within an archaeologically undisturbed context.
Within this restriction every attempt was made to take
samples which represented the horizontal and vertical ex-
tent of the site as well as the various types of features (i.e.,
mound strata, floor areas, pits, etc.). During the 1 976 sea-
son, two sets of pollen samples were taken. A set of 12
one-inch core samples were taken from within the undis-
turbed mound loadings and directly under Feature 1 (the
outer stone wall). Also, four columns which had square
cross-sections, 6 cm on each side, were cut through the
mound loadings and building floor strata down into the
natural subsoils. These columns were wrapped in plaster
soaked cloths for shipment to Tinda Shane at the Pollen
Laboratory, University of Minnesota. It was hoped that
uncontaminated pollen might be found by taking samples
from these columns under controlled laboratory condi-
tions. No pollen was found in any of the samples ana-
lyzed. Broken and fragmentary phytoliths were found in
samples from the building floor; however, it could not be
determined whether these were of ancient origin within
the clays used to construct the floor or whether they were
associated with flora blown or carried into the building
during the time the floor was in use.
During the second season. Dr. Shane came to the site to
collect samples which were to receive preliminary process-
ing immediately at facilities of Ohio University in Chilli-
cothe. She determined that the best chance for finding
intact pollen was on the under side of large in situ pieces of
charred wood. Unfortunately, once again, no pollen was
found.
Soil samples for simple comparative chemical analyses
were taken from representative mound and floor strata
and floor features. Soil samples for thin section analyses
were taken from archaeologically undisturbed soils within
Feature 1 and within and below Feature 3C in order to
study the soil structure. James Kerr of the U.S. Soils Ser-
vice did additional field studies of soil structure and pos-
sible origins on representative profile walls as well as in
the local vicinity of the site.
Charcoal samples for possible radiometric dating as-
says were collected from all charcoal deposits found in an
undisturbed context. In addition, Jeffrey Friedland of
Earth Sciences took several burned clay samples from
burned areas of the main floor (Feature 3) for archaeo-
magnetic dating analysis.
The major goals of the excavation were to collect dat-
ing and ecological samples and to determine, if possible,
the building pattern which was not reconstructible from
previous field work. These goals were met.
References
Bender, Margaret M., David A. Baerreis, and Raymond L.
Steventon
198! Further light on carbon isotopes and Hopewell agri-
culture. American Antiquity 46(2):346— 353.
Gadus, Eloise F.
1979 The Harness copper plate. Ohio Archaeologist 29(3):
27-29.
Greber, N’omi
1979 A comparative study of site morphology and burial
patterns at Edwin Harness mound and Seip mounds 1
and 2. In Hopewell archaeology: the Chillicothe con-
ference., edited by David S. Brose and N’omi Greber,
pp. 27-38. Kent State University Press, Kent, Ohio.
Greber, N’omi, Richard S. Davis, and Ann DuFresne
1981 The micro component of the Ohio Hopewell lithic
technology: bladelets. Annals of the New York Acad-
emy of Science 376:489-528.
Mills, William C.
1903 Diary. On file. Library Archives, Ohio Historical
Center, Columbus, Ohio.
1907 Explorations of the Edwin Harness mound. Ohio Ar-
chaeological and Historical Quarterly 16:1 13-193.
Moorehead, Warren K.
1897 Report of field work carried out in the Muskingum,
Scioto, and Ohio Valley during the season of 1896.
Ohio Archaeological and Historical Quarterly 5:
165-274.
Putnam, Frederick W.
1885 Explorations of the Harness mounds in the Scioto
Valley, Ohio. In Peabody Museum 18th and 19th An-
22
N’OMI GREBER
No. 39
nual Reports (1884-1885), bound in Peabody Mu-
seum Reports 3(5-6):449-466.
Squier, George Ephram, and E. H. Davis
1 848 Ancient monuments of the Mississippi Valley. Smith-
sonian Contributions to Knowledge 1. Washington,
D.C. Reprinted 1973 with introduction by James B.
Griffin as Antiquities of the new world: early explora-
tions in archaeology (Vol. 2). A.M.S. Press, New
York, for Peabody Museum, Harvard University.
Watson, Patty Jo
1976 In pursuit of prehistoric subsistence: a comparative
account of some contemporary flotation techniques.
Midcontinental Journal of Archaeology 1(1 ):77— 1 00.
3
THE SITE
Introduction to the Stratigraphy: Site Profiles
The site of the mound was a small knoll situated on a
second terrace approximately 3 km east of the Scioto
River (Fig. 2.1). The underlying geological strata are
sorted glacial outwash (Wisconsin) on which have gener-
ally developed Fox-Ockley soils with some associated
Warsaw series. These soils are described by the U.S. Soil
Survey.
Fox Series
In the Fox series are well-drained soils that developed on
deposits of calcareous gravel and sand of Wisconsin glacial
age. These soils are mainly on terraces (glacial outwash
plains and valley trains) but locally are also on kames, eskers,
and parts of moraines on uplands. Fox silt loams formed in
1 2 to 18 inches of silty material over gravel and sand, whereas
the coarser textured Fox soils formed in loamy material over
gravel and sand.
On terraces the Fox soils occur with the Thackery, Sleeth,
Westland, Wea, Warsaw, and Ockley soils. On uplands they
occur principally with the Kendallville soils, though in a few
places they are close to the Miami, Lorenzo, and Rodman
soils. Fox silt loams resemble the Ockley soils but are not so
deep to parent material. (Petro et al. 1967:133)
Warsaw series
The Warsaw series consists of dark-colored, well-drained
soils on terraces that developed on stratified, calcareous
gravel and sand outwash. These soils occur closely with Wea,
Fox, Ockley, and Westland soils.
Unlike the Fox soils, which developed under hardwood
forest, the Warsaw soils developed under grass and have a
darker colored A horizon containing more organic matter
than the Fox soils. The Warsaw soils have a less silty upper
solum, are shallower to calcareous gravel and sand, and are
less acid than the Wea soils. (Petro et al 1967:151)
In the following descriptions the Munsell color desig-
nations are given in parentheses. The color name used in
the text is the common visual color.
The major part of the soils and gravels used in con-
structing the various mound strata, the floor of the build-
ing at the base of the mound, and the various features on
that floor both within and without the building itself were
the local Fox-Ockley soils and underlying gravels. The
only exception was the outer cobble and bedded stone
mantle (Feature 1). These stones were brought to the site
from Dry Run, the banks of the Scioto to the west, or the
hills at the east end of the Scioto Valley, here 6 km east of
the site (Fig. 2.1). Within these stones was found a dark
(10YR 4/4 dark yellowish brown and 5YR 3/2 dark red-
dish brown) soft loam or fine silt which has characteristics
associated with the Warsaw series. Jerry M. Bigham of
the Department of Agronomy, Ohio State University,
analyzed thin sections made from samples that James
Kerr of the U.S. Soils Conservation Service had taken
from within these soils (see Appendix 3. 1). These soils do
not appear to have developed in place under grasses
which may have covered the stones since the mound was
built. It is possible, though probably unlikely, that they
were placed there as fill by the original builders. It is more
probable that they filtered down from upper mound lay-
ers before overgrowth stabilized the upper strata. We at-
tempted, with the assistance of Mr. Kerr, to find possible
local sources of such soils. Unfortunately historic and
possible prehistoric land use has disturbed the local area
too heavily to enable such areas to be found. Pocket prair-
ies did exist in this area of the Scioto Valley historically,
for example, at Prairie Station 8.5 km north of Liberty.
Also, in situ prairie soils were found under a section of the
High Banks Earthworks 8 km north of the Edwin Harness
site (Shane 1973, personal communication 1982). Thus,
although the characteristics of the soils within the mantle
are clear and different from the other soils we found in the
mound, the exact origin of these dark soils has not been
found.
Prehistorically the surface of the knoll had been cleared
into the B soils horizon with some minor filling required
before the desired building level was achieved. Evidence
of this filling was found in small pockets of soils with little
to no structure at the upper edge of the undisturbed B
horizon, as, for example. Sample II in Appendix 3. 1 . It is
likely that materials which had been burned in the land
clearing were incorporated into the layer which underlay
much of the site (Feature 3C). This was a generally thin (5
cm), dark (7.5YR 3/2 dark brown) clay stratum (Sample
I, Appendix 3. 1 ). It occurred, as did all of the floor strata,
within the area defined by the heavy cobble wall (Feature
1 ). At the outer edges of this area Feature 3C was the only
floor stratum present.
Sketch profiles of the site are given in Figure 2.5. Prac-
tical difficulties prevent the presentation of one-to-one
scale profiles. The field data are of course available for
study. The original land clearing, construction, and other
cultural activities resulted in a complicated stratigraphy
of which only disjointed and truncated remnants were left
in 1975. Fortunately there was still enough information to
allow reasonable reconstructions. The physical character-
istics of the strata follow directly; the horizontal extent
will be discussed further as the floor maps are discussed.
Three profiles along grid N-S lines and three along grid
E-W lines are shown. The profile along the E507.5 line
(Section DD') gives a general profile of the building floors
and remaining mound strata. Since the major axis of the
building is east of grid north, no grid line gives a symmet-
ric profile of the structure. Section EE' is a composite pro-
file based on the floor strata along the E495 line with post
24
N’OMI GREBER
No. 39
holes based on data farther to the east; thus this section
represents a line parallel to the major axis of the Big
House.
The number of superimposed floor layers varied from
one (Feature 3C) at the outer edges (east end of Section
AA', north and south ends of Sections DD' and EE') to
five (middle of Section DD').
All the tombs, pits, post holes, and almost all the
burned areas were found on the topmost layer. Feature 3.
This layer was a heavy gray clay (10YR 5/3 yellowish
brown), which, as noted by previous excavators, appears
to have been puddled, that is, mixed with water when put
down. The major layer beneath this clay (Feature 65) was
a very sandy orangish brown clay (7.5YR 5/8 strong
brown), at times mottled with gray. Between these two
heavy layers, a 2 mm layer of hard, reddish orange sand
(7.5YR 5/6 strong brown) was frequently found. Feature
33, which was distinctive in color and texture, was found
below Feature 65 on the E507.5 line but not on the E495
line. This stratum was composed of a red, sandy clay
(5YR 5/6, 5YR 5/8) mottled with dark gray, pink, yellow,
and light brown clays, and relatively evenly scattered
charcoal bits, 0.5 to 3 cm in size. There were limited sec-
tions within this layer, such as Feature 72 (a 50 cm by 75
cm area in N535 E509), which were of the same texture
and mottling as the major extent of the feature but within
a brown rather than a pink-red matrix. Feature 50, a soft,
moist, dark gray-brown, slightly sandy clay (7.5YR 4/2
strong brown) was found below Feature 33.
A construction break in Feature 3 was found in the
E495 profile near N543. It appears likely that such a break
also existed on the east side along E507.5, although the
extensive prior excavations, as illustrated in the deep cut
seen in the north end of Section DD', have destroyed the
evidence. It is dear that north of this point, although the
activity floor level and the construction materials appear
to be the same as those of Features 3 and 65, there are
differences. The floor is less consistent in total thickness,
and there is no consistent ordering of gray clay over orang-
ish brown sandy clay. Areas of both types of clay are in-
terspersed as can be seen in Section AA'.
The area shown at the south end of Section DD' shows
another variation in floor construction. Here Feature 39,
brown clay (7.5YR 5/6 strong brown) mixed with sands
and gravels, lies above Feature 3C. This stratum has a
different color and texture from that of Feature 65, which
was found directly beneath the main gray floor under the
structure.
In the N-S profiles, the large posts which were set to the
depths of the natural sandy gravel layer can be seen. In
Section DD' the remains of the posts defining the East
Structure can be seen at the base of the backfill. The loca-
tion of Putnam's and part of Mills’s excavations can be
seen in Section AA', an E-W profile along the N560 line.
This is near the beginning of the wider portion of Put-
nam’s trench (Greber 1 979: Fig. 6.4). Here the bottom of
his trench was sloping down towards the main activity
floor, which it had not yet reached. Mills did, as he wrote
in 1907, dig up to the edge of Putnam’s trench and
through the mound floor.
The remnants of the mound construction itself can be
seen in all the sketch profiles. What was an early, if not the
first, covering over the parts of the Big House can be seen
at the base of Putnam’s trench as well as farther east on
the N560 line and at the ends of the N-S profiles. This
covering is a brown sandy clay (7. 5YR 4/4 strong brown)
with a few pebbles (ca. 12 mm). Loadings are usually dis-
tinct and 25-50 cm in size. All of the soils we excavated
appeared to contain no cultural debris. In some areas,
such as at the north end of Section DD' where these soils
have been placed over a cleared soil horizon rather than a
prepared clay floor, the boundary between the fill and the
in situ soils was less distinct. Several small features are
probably a part of the early mound raising. Features 20
and 21, found in unit N560 E495 (Fig. 2.2), appear to be
very large loadings or very small mounds of gravels and
sands. Feature 43, found in unit N530 E492.5 (Fig. 2.3),
was composed of sheet loadings as contrasted with con-
tainer type loadings. Alternating sheets of sand, gray
sandy clay, and charcoal were found in a very restricted
area close to Feature 3. These were found at the edge of a
historically disturbed area and could not be traced east to
any significant extent.
Covering the base of the first fill in most areas is Fea-
ture 41, a 10 cm layer of pea gravels. These gravels are
identical to those found in the local C horizon. This same
type of material has been found as a major cover in many
Hopewell contexts, for example, within the Seip Earth-
works (Baby and Langlois 1979; Greber and D. Griffin
1982), at Mound City (Brown and Baby 1966), and in a
number of other sites where Squier and Davis ( 1 848) de-
scribed it as “sand.”
A second stratum of mound fill found over Feature 41
is composed of a matrix similar in color to the first fill, but
with a higher percentage of gravels so that loadings are
more difficult to detect. These two strata are in effect a
reversal of the natural B horizon in which the deeper soils
contain more gravels. This suggests that the second cover-
ing came from a greater depth, perhaps from the same
borrow pit. The only artifacts which we found in the soils
of this secondary fill were associated with Feature 56, a
cremated burial.
Feature 1, the bottom edge of a heavy cobble wall, was
found over this secondary fill. The base of this wall varied
from 1.5 to 2 m from inner to outer surfaces. It was com-
posed of waterworn pebbles 10 cm or less in size and
larger stones ranging up to 50 cm in length. Included were
flint cobbles, sandstone, and limestone, both waterworn
and bedded. All the stones in two 50-cm wide sections
through the wall at N517 and N528 on the east side were
saved as a representative sample. The average depth re-
maining of the wall was 10 cm. In general, the smaller
1983
EDWIN HARNESS MOUND
25
cobbles were found on the inner edge with the larger
stones on the outside. It is likely that the edge row of large
stones extending almost 6 m north from N532 on the west
side of the site is the end of the elliptical “ring” which Mills
( 1907, 1903) described on the west Hank of the mound.
Over the secondary fill, which, as previously noted, ex-
tended under and beyond Feature 1, were soils that ap-
parently eroded down from higher parts of the mound. At
times the boundary between these soils and the secondary
mound fill was difficult to distinguish, possibly because
these soils had a similar geological origin. The north end
of Section DD' shows an example of this type of stratig-
raphy. Auger samples taken on a line continuing north
from this profile show that the slough layer ended within
10 m.
To summarize, the remnant mound strata found from
the southwest corner of the mound, continuing clockwise
around the perimeter of the mound towards Backhoe
Trench 2, were a first fill covered by a thin layer of pea
gravel, followed by a secondary fill capped by a stone
wall. The lower fills extended beyond the base of this wall.
The secondary fill was covered over time with soil eroding
from higher up the sides of the mound. There is a different
stratigraphy on the southeast and south edges.
In Section FF' Feature 69A is seen directly over the pea
gravels of Feature 41. Here these gravels are more spo-
radic and appear to have been disturbed, probably by the
placement of this upper stratum, which is composed of a
distinctive, dark sandy loam ( 10YR 3/2 very dark grayish
brown) with a high density of small to medium gravels.
We found no artifacts in these soils, nor in the similar soils
of Feature 69 (Section BBO- Within both of these ex-
tended dark strata were deposits or discrete sections
which were darker (2.5YR 2.5/0 black) and more com-
pact. These areas (Features 55, 68, 80, 81) contained a
high density of rock, much of which was apparently fire
cracked, and large pieces of charred wood and charcoal.
Feature 55 extended 3 m by 2.6 m and was 55 cm deep.
The rocks within the feature were predominantly lime-
stone and ranged from pea gravels to 15 cm rocks; the
average maximum fracture edge length was 10 cm. Below
Feature 55 was a pit extending down into the C horizon.
A bundle burial (Feature 60) was found at the base of this
pit. No other pits were found associated with any of the
other rock concentrations recorded.
It is likely that the intrusion of Feature 69A under the
rock wall which is seen in Section FF' is the result of pre-
historic disturbance. The major archaeological distur-
bances on the east side of the site prevent the location of
the first fill in some profiles (see Section DD'). Here again
the outer dark, humic layer has been placed up to or under
the heavy rock wall. A small section of secondary fill ap-
pears under the dark layer in Section BB' but not in Sec-
tion FF'. The relationship of Features 69 and 69A to each
other is not clear from the available evidence.
To summarize the mound strata found in 1976-1977,
the lowest fill found was composed of soils likely taken
from the upper levels of the local B horizon. These soils
contained no artifacts and were deposited over the main
activity floor and outer fringes surrounding this main
area. Over this fill, up to some unknown depth, was a thin
layer of pea gravel identical to gravels found in the local C
horizon. This thin layer extended outward and usually
beyond the lower edges of the first fill. On top of this
gravel was a second fill composed of soils apparently
taken from lower levels of the local B horizon. These soils
contained no artifacts except those associated with a
cremated burial found on the east side of the site. On top
of this fill was placed a stone wall which was probably
intended to retain upper layers of mound fill. This wall
was above the irregular outer edges of the prepared knoll
surface; however, underlying mound strata extended
beyond this wall. On the southern and southeastern edges
of the site a dark, humic soil containing gravels but no
artifacts was placed up to or upon the mound. There are
indications that earlier mound strata were disturbed as
this soil was added to the complex. Discrete areas or de-
posits within this dark soil contained very dark, compact
soils mixed with burned rock, predominately limestone,
and burned wood and charcoal. No artifacts were found
in these very dark soils. The exact horizontal extent of the
one or several areas of these outer dark soil strata is not
known.
The Horizontal Extent of Floor Strata
Mills’s general description of the “clay floor” is in ac-
cordance with what we found; viz., “in some places it was
only three or four inches in thickness, in other places from
ten to twelve” (1907: 1 38). He also describes Feature 3C
and the original soils below it: “The original surface of the
site was covered with ashes and charcoal. . . . Nowhere
were there evidences of any prolonged fire on the original
surface, rather only the burning of small limbs as evi-
denced by the charcoal remains” (1907:138). The exten-
sive areas of Feature 3C which we found did not have
large pieces of charcoal, even from the “burning of small
limbs.” It may be the lack of large pieces rather than direct
evidence of limbs which Mills is describing. The layer, as
discussed above, was not in situ. We also did not find any
evidence for burning on the natural soil horizon. Includ-
ing Mills’s description. Feature 3C, the dark clay stratum,
underlay the Big House and the entire activity area; this is
generally the same area as defined by the inner edge of the
heavy cobble wall. Feature 3C was under or within a me-
ter or so of this wall edge around the entire site.
The overall extent of the main activity floor, Feature 3,
appears to have been from the outer edge of the portico on
the north and south, definitely beyond the portico on the
west. Feature 3B was a single layer above Feature 3C and
abutted the western edge of Feature 3. It was more grav-
26
N’OMI GREBER
No. 39
elly but generally similar to Feature 3. It was under the
west end of Feature 6 and extended several meters south
and appears to have been just a local variant in construc-
tion materials.
Feature 65 appears to have directly underlain Feature 3
in most areas of the site. At the outer edges there were
some minor variations as to which extended farther (e.g.,
see Fig. 2.5). Feature 39 may have been a part of Feature
65; however, the heavy historic disturbances prevented
the establishment of a direct connection. The slight differ-
ences in texture may have been accidental or perhaps as-
sociated with the end of the portico area.
Feature 33 and the underlying Feature 50 appear to
have been only in the northeastern area under parts of the
Big House and portico ( Fig. 3.1). The horizontal extent of
these features does not appear to coincide with any single
portion or portions of the Big House. As described above.
Feature 33 was the most distinctive floor stratum found.
If we assume that the portions found in the same strati-
graphic sequence around disturbances caused by Mills’s
major excavations are the same stratum, then the west-
ernmost piece of Feature 33 was near N546 E500, the
northwest corner near N553 E505, and the southeastern
edge near N535 E509. Although the exact shape cannot be
known, the north and south edges appear generally to be
oriented magnetic east-west. In the relatively extensive
areas of this feature, which were hand troweled, no post
pattern was found. We were rather hoping that an “Ade-
na” house would appear. It is possible that a small struc-
ture with shallowly set posts (less than 10 cm) could have
existed in the southwest heavily disturbed section of the
feature. I think it is more likely that the area had been
used without a major structure. Perhaps the mixed
burned clays and charcoal bits were culturally significant
remains from the land clearing and other activities asso-
ciated with the first use of the site. Such an origin, as care-
fully cleaned up and redeposited debris, would be in keep-
ing with the character of various types of features found
on the main activity floor.
The Building Post Pattern
At the base of the Edwin Harness Mound were the re-
mains of a large building, which in keeping with known
Native American languages we can call “Big House.” This
Fig. 3. 1. Cuts made by Mills in 1905 through the original floor layers. The layer (Feature 65), thus exposing the reddish clays of Feature 33. Unit
main activity floor (Feature 3) has been cut through, as has the secondary N550 E5 10, level 5, 29 June 1977.
1983
EDWIN HARNESS MOUND
27
is a translation, for example, of the Shawnee m ’sikamekwi
(Greber 1979:28) or the Creek tcoko-thlako (Hudson
1976:221). The same word is used in Shawnee for stomp
dance ground (Charles Callendar, Case Western Reserve
University, personal communication, 1983); thus, the
English translation has connotations not of a large resi-
dential house but of a special activity area. Also at the
base of the mound were the remains of cultural activities
which took place about the Big House within an area
which, though not symmetric, was apparently well defined.
There are several major parts of the Big House (see
Figs. 2.4, 2.6). The floor plan outline of each of the largest
two parts is a classic Ohio Hopewell nearly rectangular,
rounded corner design (Baby 1971; Baby and Langlois
1979; Brown 1979). These two parts, the North Section
and the Middle Section, are joined by a rectangular hall.
The South Section, which is circular in floor plan, is
joined directly to the Middle Section. On the east is a
small, again classic in floor plan, structure which is joined
to the other elements by a corridor area. A wall extends
around the north side of the East Section. The post pat-
tern of these basic parts has been abstracted from Figure
2.4 and presented in Figure 3.2. Because of the historic
disturbances prior to 1976, the building posts were found
at differing depths and with varying degrees of distur-
bance from none to total. The latter was rare. The size of
each post hole indicated in Figure 2.4 is the size first found
in 1976 or 1977. In Figure 3.2 adjustments have been
made to an estimated original floor.
There is a portico-like area about the entire complex
which is demarcated by large posts on all sides. On the
south, west, and north along the line of these posts is a
shallow, clay-filled trench containing small stake holes.
These were likely supports for a screen or narrow wall
which would have formed an enclosure about the Big
House. There was no evidence of such a shallow trench on
the east side of the house because of the extensive removal
of the main floor area in the 1903 and 1905 excavations
(Mills 1903, 1907). However, considering the east-west
asymmetry of the complex, there may not have been such
an enclosure on the east side.
The two major elements are similar in general architec-
tural design. There is a set of 48 inner posts (average post
size 24 cm) which forms the structural strength of the
building. These posts are arranged in seven rows N-S and
E-W. There is some bending of the rows to turn the
corners and some space left at the center of each building.
The E-W separation of the rows is the same for both,
averaging 1.6 m center to center. The N-S separation,
however, is greater in the North Section (1.9m compared
to 1.6 m). Thus there is a more spacious floor plan in the
North Section, but the same number of structural sup-
ports. The size and placement of these supports strongly
suggest that there had been an upper floor or platform
area. There was at least one post from the North Section
which was just over 10 ft (3 m) high. This charred post.
which was photographed by Mills in 1905, is probably the
same one which was sketched by Putnam in his field notes
(1884). Within the identified sample from the 1976-1977
excavations, the major construction timbers were young
hickory trees (Table 5. 1). The exact length available from
a tree which could have fitted into the recorded post holes
cannot be calculated because of the many complex envi-
ronmental variables involved. However, for similar en-
vironments, 58 ft (17 m) heights are recorded for 18 cm
diameter trees in Silvics of North America (Fowells
1965:126).
The next outward series of posts in both the North and
Middle Sections architecturally appear to have supported
a facade or, more likely, a roof for the area defined by the
48 inner posts. There is some variation possible in the
number of posts contained in this series because of the
common boundaries with the hall and the South Section.
This series contains posts of different diameters and dif-
ferent spacing in the two sections. The M iddle Section has
a paired post corner arrangement and at least 2 larger
posts in the western line. The majority of these posts aver-
age 18 cm in diameter; a few are 25 cm, while the 2 large
posts on the west are 40 cm. There are a total of 39 posts in
the series plus, possibly, Posts 378 and 387 (Fig. 2.4). For
the North Section there are probably 42 posts in this se-
ries, plus possibly Posts 42 and 72. The average size of the
northern posts is 23 cm. In general, these posts are more
widely spaced than those of the Middle Section. There is
relatively closer spacing at the corners. The larger post
size for this post series in the North Section may be related
to the longer spacing between posts. It is possible that in
the original building design the same number of posts had
been used in this series in both sections.
The last series of posts differs on the east and west sides
of both structures. Those on the east (13 cm average di-
ameter) are evenly spaced and may form a formal facade
or decorative line. Those on the west side are also mainly
small, but their placement is irregular and scattered. They
do not appear to be structural posts. Four short lines of
posts form a hallway or corridor between the North and
Middle Sections.
As is shown in Figure 3.2, the South Section is outlined
by posts of three size ranges: 7 cm average diameter, 15
cm, and 24 cm. The central posts of this element may have
served as structural supports; however, the contents of at
least one of these (Post Hole 216) suggest that they also
served other functions. The materials recovered from post
holes are described in another section of this chapter.
The East Section is outlined by small posts (10 cm).
This, as the South Section, is of small scale. It does not
appear likely that there had been an upper level in this
structure. The outer wall on the north side is composed of
posts 18 cm in average diameter.
The outer post holes which ring the complex average 27
cm in diameter on the west side and 1 8 cm on the east side.
Next to or about these posts on three sides of the structure
indicated are estimates of original sizes on the main floor of the building prior
1983
EDWIN HARNESS MOUND
29
was a shallow (8-13 cm), narrow (15-20 cm), clay-filled
trench (Features 22, 59, 85, 88) which contained small
stake holes ranging in size from 2 to 5 cm (average 3.6 cm).
Along the east side, these holes were at varying angles in
the clay such that stakes set in the holes would tilt, appar-
ently at random, towards the north or south but would
remain in the same plane. These stakes likely were part of
a screen or fence which was constructed after the main
floor had been completed. All the floor graves appear to
have been within the area enclosed by the trench and
outer posts. Other types of remains of cultural activities —
mainly burned areas and deposits, a few post holes, and
pits — were found outside this boundary.
Other Posts
The remaining post holes found fall into three size
classes: large (over 25 cm), medium ( 10-20 cm), and small
(less than 10 cm). The number of each size associated with
the several parts of the Big House are given in Table 3.1.
The range in the depth from the top of the main floor
(Feature 3) to the bottom of the medium-sized post holes
which were found basically intact was 23 to 40 cm. The
corresponding depth for the small ones was 7 to 19 cm.
Thus there may have been additional small posts in sec-
tions where Mills removed both Feature 3 and Feature 65
(total depth ca. 12 cm). The main sections where this had
been done were near the center of the site in the vicinity of
the hall and the East Section. It is less likely that evidence
of medium-sized post holes was completely destroyed in
the earlier excavations.
The small posts found associated with the South Sec-
tion are generally about the perimeter. The four medium
and large posts are more centrally located. In the Middle
Structure, the largest number (12) of the medium posts
found form the last series of posts on the west side of the
structure as described above. Approximately half (17) of
the small (here 3 cm) posts are located along the east side
from the center line of the structure to about 4 m south of
the center. The remaining small posts are scattered less
densely (Fig. 2.4).
TABLE 3.1
Non-structural Posts Associated with
Edwin Harness Big House
Location
Corridor
East
Diameter
North
Middle
East
Central
Section
South
(cm)
Section Section Section
Hall
to Hal!
Section
<10
39
31
0
0
2
14
10-20
40
25
1
7
5
4
>25
2
6
0
0
0
1
Total
81
62
1
7
7
19
In the North Section the medium posts are also in an
irregular line on the west side. Others were found within
the last western row of the central structural posts, out-
side the northern end, in the northeast corner, and as part
of the grouping which forms the central focus of the
building.
A few posts holes were found outside the portico: Post
Hole 2 (20 cm diameter) on Feature 3 near Feature 1 7 and
four post holes a short distance north (79,59) or east
(156,206) of the main complex (see Fig. 2.6).
Contents and Description
In general the main structural post holes were straight
walled, frequently with a clay lining below floor levels,
and also frequently bottoming in the C horizon gravels
where drainage is excellent. An exact inventory of the
contents of these posts cannot be given due to Mills’s ex-
tensive excavations (which were appropriate) and the loss
of his field notes (which is unfortunate). Describing the
contents, he says, “Very frequently these molds would
contain broken animal bones, mussel shells and occa-
sionally a piece of mica. We have never been able to find
in the great number of molds examined, any implements
or ornaments” (1907:138). Any of these objects could
have been accidentally swept into the hole by the prehis-
toric occupants of the building during the cleaning up of
activity debris from nearby floor areas. Such cleaning
would be in character with the apparent manner in which
shallow deposits of mixed materials had been made on the
floor (see following section). The majority of objects
which we found within post holes were similar to those
found by Mills and appear to have been accidental inclu-
sions. The small rocks, pieces of fabric, and perhaps
sherds, which we also found, were likely used as wedges.
However, there were incidences of apparently deliberate
filling of some post holes. These post holes were con-
structed in the same manner as were others of similar
diameter.
Although numbers of posts were burned before the first
stage of mound building, there were posts which were re-
moved. This procedure has also been found in our current
work within the Seip Earthworks. James Brown of
Northwestern University has discussed with me similar
findings at Mound City, and there are reports of similar
activities (for example, at Garden Creek, North Carolina)
from Middle Woodland sites outside Ohio (Chapman
and Keel 1979).
At least one post hole was filled with a collection of
materials similar to the deposits on the floor. Post Hole
216 contained a variety of burned woods (Table 5.1), ash,
shell, bone, apparently fire cracked rock, and mica pieces,
some of which were 8 by 3 cm. The soil surrounding these
objects was oily. Another non-structural post which had
been removed was Post Hole 178, immediately north of
the Big House. A small egg-shaped clay basin had been
Fig. 3.3. Location of post holes with surrounding stains. Only those post holes which were sur-
rounded by rings of iron stain are indicated by solid circles. Small posts have been relatively enlarged
for visibility. See also post descriptions in chap. 3, Fig. 2.4, Table 3. 1 , and Table 5. 1 .
1983
EDWIN HARNESS MOUND
31
constructed over this hole. The basin (Feature 62) had
contained a fire which reddened some of the surrounding
clays (5YR 4/6 yellowish red). The fire, as in Post Hole
216, had contained several varieties of wood, including
the second recorded occurrence of pine (Table 5.1), and
six fragments of mammal bones.
The post had also been removed from Post Hole 25.
This was a 26 cm E-W by 30 cm N-S hole which extended
56 cm below the floor surface and was located along the
narrow trench on the west side of the North Section. In
this hole a human skull and mandible was found. Parts of
the skull were colored red, apparently from red ocher.
Holes had been drilled in the ascending rami of the mandi-
ble in such a fashion that skull and mandible could have
been articulated. The skull has been identified by Ray-
mond Baby (personal communication 1977) as that of a
young adult male. No ocher, charcoal, or any other cul-
tural materials were found within this post hole.
One characteristic that divided the post holes into two
classes regardless of size was a “rusty” stained ring which
surrounded 85 of the post holes (Fig. 3.3). This ring,
which averaged 2 cm in thickness, was irregularly shaped
and ranged from less than 1 to 25 cm from the edges of the
post holes. It was surrounded in most instances by a fine
black line. The area between it and the post hole tended to
have a “crusty” texture. Putnam’s description of such
rings in his field notes coincides with our findings in the
central area of the North Section, which was within Put-
nam’s trench. This stain was present about the entire post
hole; thus, evidence for its existence was not destroyed by
previous excavations (see Fig. 3.4). These rings appear to
be caused by the movement through the soils of water
soluble forms of iron (this leaves an orange stain) which
are preceded by the movement of manganese (this forms a
thin black line). I have not yet been able to find suitable
chemical tests to prove that these rings were caused by
deterioration of coloring used on the posts; 1 believe that
they were. The patterning of these posts is shown in Fig.
3.3.
Of the nine structural posts with such stains in the
North Section, one was within the central focus area; six
were in the southeast corner; two were near the middle
hall. The north three posts on either side of the hall had
stains, as did the post in the Middle Section opposite the
center line of the hall and the stained post in the North
Section. The other stained structural posts in the Middle
Section were paired corner posts, two pairs on the west
side, one on the east. In the South Section, the stained
posts appeared to be symmetrically arranged with respect
to the two oak posts (Table 5. 1). This pattern may indi-
cate that the entrance to the structure was towards the
southwest, and Post Hole 216 would form a line with this
entry way. The posts forming the west side of the East Sec-
tion and one additional post on the north also had ac-
companying rust stains. The area between the East Sec-
tion and the middle hall had such posts along the north
and the west sides.
Fig. 3.4. Examples of post holes with surrounding stains. The rings of
iron stain continued the length of the posts as illustrated in the middle
sections of the site where Mills removed extensive portions of the floor
strata and post holes were found in 1 976 and 1977 under extensive back-
fill. Unit N537.5 E497.5, level 3, 75 cm below base of SW stake, PM 34,
37, and 38, 29 July 1976. These posts are on the west side of the Middle
Hall.
Of the non-structural posts found in the North Struc-
ture which had rust stains, six were within the central fo-
cus, single posts were at the southwest and northeast
corners, and four posts were just outside the north end. In
the Middle Section most of the small clustered posts on
the east side had stains, as did several within the last south
row of the main structural posts.
In summary, at least half of these specially marked
posts were located near entranceways or near the perime-
ters of parts of the Big House. The others appeared to be
marking various special areas within the interior of the
building.
Central Focus of the North Section
The center of the North Section had a more compli-
cated configuration than that of the Middle Section. This
was an area which had been excavated by all previous
expeditions except those of Moorehead. The most infor-
mation on the original deposits about this area is found in
32
N’OMI GREBER
No. 39
Putnam’s field notes. The small (Post Hole 432) and
medium-sized (Post Hole 444) posts set on either side of
the geometric center of the building were surrounded by
rusty stains, as were the other non-structural posts about
the main posts 23 and 462 in the next row towards the
west. The southern edge of the large (ca. 1.5 by 4.5 m)
deposit of burned matting and charcoal described by
Putnam was apparently located at the northern end of the
area marked by these stained posts. The long axis of the
deposit was E-W; thus, it extended relatively symmetri-
cally on both sides of the building center line. In the
Peabody Catalogue (Number 34982) the deposit is de-
scribed as being “1 to 4 inches thick and on burnt clay
about 1 foot from the bottom of the mound.” The field
notes call the floor on which the tombs were found “hard
pan”; this is the same as our Feature 3. The stratigraphy
containing the charcoal deposit is given in a label for a
sketch as “hard pan” at the base followed by “clay &
gravel 6 inch [s/c], sand /i inch, pure clay 4 inches, char-
coal layer 2-4 inches, clay, gravel & loam.”
Three feet (ca. 1 m) above this deposit was a 1 -ft (30-
cm) thick deposit of sands and gravels which extended 3
by 4 ft ( 1 by 1.3 m). A “basin-like cavity” apparently asso-
ciated with these sands and gravels contained mica, shell,
flint, carved bone, and an unusual human effigy pipe (see
Fig. 3.5). The horizontal location of the gravels and cavity
are not given; thus, it is not clear whether this cavity was
deliberately created or the result of a cave-in of the roof of
one of the tombs underneath the gravel deposits. One of
the few extended burials found in the mound was located
at (or possibly on) the eastern edge of the matting. This is
“Skeleton 3” in “Burial Chamber No. 6.” A copper plate
had been placed lengthwise on the chest of the individual
whose head was towards the southeast. A second ex-
tended burial “Skeleton 1,” also with head towards the
southeast, was located just north of the northwest corner
of the burned deposit.
An extensive ash deposit was found southwest of the
charcoal deposit in the vicinity of Feature 24 at the
southwestern edge of the central space. Another unusual
specimen, a carved stone sphere (Willoughby 1916: PI.
lOi), was found at the edge of this ash area, which is de-
scribed in the field notes as “distinct masses occurring
from a few inches to 3 feet above the clay.” In summary,
the archaeological evidence about the central space of the
North Section suggests much more extensive activity than
was found in the center of the Middle Section. The scale
of these deposits is also greater than that of any of the
other deposits which we found. The major deposit found
by the schoolboys was probably part of this central focus.
No other extensive deposits have been reported in any of
the available records.
Burned Areas on the Floor
All the burned floor areas found, except Features 18
and 36, were reddened and fire hardened to a depth of 2
Fig. 3.5. Pipe fragment (4.5 cm X 3.5 cm) found by F. W. Putnam over
deposit of burned matting immediately north of geometric center of the
North Section. Photograph courtesy of Peabody Museum of Ethnology
and Archaeology, Harvard University.
cm or less. Similarly, the individual charcoal deposits
were also thin. Features 18 and 36, which were within the
Big House, were different from each other. Feature 18,
which was in the southwest corner of the North Section,
appears to be the remnant of a heavily fired prepared
basin. There was only a strip 40 cm long by 1 cm wide
remaining in place, but large chunks (7X 15 cm) of heavily
burned clay of the same color and texture were found in
the nearby backfill.
Feature 36, which was at the center of the Middle
Section, contained the deepest evidence of fire harden-
ing found. Here in an area 120 cm N-S by 70 cm E-W,
hard, red sand (2.5YR 4/6 red, 5YR 5/8 yellowish red)
was found through the gray clay floor. The soils were red-
dened and hardened to the depths of the underlying C
gravels. Thin (1-2 cm) layers of cemented sand and pea
gravels and ash (0.5-4 cm) were found on top of the sand.
Unfortunately, this, as all fired areas found within the Big
House, had been archaeologically disturbed to some de-
gree. In a sample taken from the in situ layers of Feature
36, pine wood has been found. This is one of only two
examples of this type of wood in the identified floral sam-
ples (Table 5.1).
Mills in his report describes fires set on top of graves.
Such fires may have been the origin of the burned areas on
the portico immediately west of the North Section (Fea-
ture 78, about Features 28 and 29); and the areas within
the South Section (Features 82B and 92). No artifacts
were found associated with these or any of the other fired
areas within the Big House or the portico area.
1983
EDWIN HARNESS MOUND
33
More complete stratigraphy is known for the fired
areas found on the west and north sections of the site. In
some of these areas, ash and any burned matter had been
cleaned away and only the reddened and fire hardened
clay floor was found (Feature 22 and sections of Feature
16). Other discrete areas (Features 6, 7, 8, 9, 10, 49) were
fire reddened floor areas covered with a thin (2 cm or less)
lense of ash and charcoal, and then a final covering of
sand, gravels, or clay. None of these features contained
any artifacts.
The areas in which some materials were found (Fea-
tures 31, 44, 45, 46, 51, 53, and 53A) consisted of thin
deposits of mixed charred and sometimes uncharred
materials, placed on burned and un burned sections of the
floor. As is described in the discussion of stratigraphy, the
floor west of the Big House was somewhat different in
character from that on the north. North of the building
there were more patchy and more mottled sections of
floor. Frequently there was only one layer above Feature
3C. This peripheral floor was gray clay, as Feature 3, or
orange-brown sandy clay, as Feature 65, which was stra-
tigraphically directly beneath Feature 3 under the Big
House. Thus the charcoal deposits shown in Figure 2.6
which are not on a gray floor were found on an orange-
brown sandy clay floor.
The deposits of materials were not the results of in situ
burning. For example, in Feature 3 1 a thin ( 1-4 cm) layer
of sterile brown sand had been placed over a cleaned red-
dened floor area; then a layer (1-2 cm) of thoroughly
mixed charcoal, fired clay nodules, and an unidentified
mammal bone fragment were placed on top of the sand.
In Feature 44, a series of thin layers containing burned
beads, cut mica fragments, broken bladelets, and bone
fragments were deposited over burned and unburned
areas of the peripheral mound floor. The three whole plus
several fragments of clay beads and the single fresh water
pearl found were likely part of a necklace, which included
many canines (see chap. 6). These burnt and broken
drilled canines included fox and raccoon. Other faunal
material in the deposit included 51 turtle shell fragments;
2 bird bones; 1 bird talon; 3 catfish vertebrae, one of
which was modified; 13 mammal bones; and 42 unidenti-
fiable bone fragments. All bone was burnt. Twenty-seven
bladelets have been identified from 35 pieces.
Feature 45 was similar in construction to Feature 44
but contained less material. There were mica fragments;
clumps of burnt sand; 28 burnt mammal bone fragments,
3 unburnt; 3 bird bone fragments; 5 unidentifiable bone
fragments; several fossil fragments; and one bladelet mid-
section (see also Tables 5.2 and 5.3).
Feature 46 contained only 5 baked clay nodules and 2
burnt mammal bones mixed in with the charcoal deposit.
Feature 51 was smaller in extent than the other deposits
and contained a flint chip, 1 1 mammal bones, and 1 left
distal deer radius.
In Feature 53, a 1-2 cm layer of gray clay had been
placed on a fire reddened surface (Feature 53A); another
layer of dark clay ( 1 cm thick) had been placed over this
gray clay. On top of the dark clay was found 1-2 cm of fire
reddened clay covered with scattered charcoal, bladelet
fragments, and bone fragments; all these were covered by
a mottled gray clay. The 3 bladelet pieces represented 2
different bladelets. Seven mammal bone fragments, 1
burnt mammal bone fragment, 1 unidentifiable bone
fragment, and several flint flakes were in the deposit.
To summarize, each deposit outside the Big House was
carefully prepared; the number of artifacts and volume of
deposit were small. The unburned objects found may
have been used in the activities immediately associated
with the lighting of the fire which burned the remaining
materials. No pottery sherds were found in any of these
deposits.
Feature 54A is an unusual deposit associated with an 80
cm diameter, 19 cm deep depression found in the main
floor west of the Middle Section (Fig. 2.6). Feature 3
thinned into a negligible thickness at the center of this
depression. Feature 65, which was about 2 cm thick here,
followed the line of the depression. A thin layer of dark
clay with deposits of charcoal and charred twigs was
found between the two main floor layers. At 80-100 cm
from the depression center Features 3 and 65 were, as
more usual, 5-6 cm thick. The deposit on Feature 65 was
offset from the center of the depression (see Fig. 2.6).
Soils which were the same in color and in texture as the
local B horizon were found extending into the yellow C
gravels immediately below the depression. A thin layer ( 1
m diameter) of charcoal was found 19 cm below the center
of the depression; however, not the center, but the south
edge of this deposit was directly below the center of the
depression. A few apparently fire cracked rocks and flint
flakes were found within the charcoal on Feature 65; noth-
ing was found within the lower deposit. The origin of the
depression is not clear. Perhaps it resulted from the set-
tling of a relatively large section of fill needed during the
original land clearing; perhaps it was intentionally con-
structed. There was no open arch above this depression as
was usually found over the collapsed tomb roofs (e.g.,
Putnam 1884). The small trench about the portico area
was cut through the depression; the trench also cut along
the east edge of the heaviest section of the charcoal de-
posit on Feature 65.
There were three other deposits of charcoal found
below the main activity floor (Feature 3). Feature 57, a
deposit of hard ash and charcoal, 1-2 cm thick as usual,
extended 40 cm N-S by 60 cm E-W on Feature 50 (below
Feature 33) at N548 between E506 and E505. There were
scattered bits of charcoal found 10 to 75 cm about this
deposit. There was a small (5 X 15 cm) area of orange stain
found near one edge, but there was no definite evidence
for in situ burning. Similarly, Feature 97 was found on an
unburned surface. A thin (less than 2 cm) layer of
powdery charcoal lay between Feature 3 and Feature 65
at the edge of the disturbed main floor (N542.5 E495). It
extended N-S, and was apparently the western edge of the
34
N’OMI GREBER
No. 39
original deposit; no significant extent was found in trac-
ing this deposit west. DIC-662 is from this feature (Table
3.2). Feature 73 was a 90 cm N-S by 33 cm E-W layer of
charcoal deposited on gray clay east of the North Section
and below Feature 33. A dark red burned area was found
at the southern edge of the feature.
The charcoal was covered with a thin layer of mottled
yellow-gray clay which contained many small limestone
pebbles. There were also limestone pebbles mixed with
the charcoal found in the northern end of the feature. Ex-
cept perhaps for the deeper deposits below Feature 54, the
deposits of charcoal found under the main activity floor
appeared to be similar in character to those found upon
the floor; that is, they were carefully constructed, thin,
and with few artifacts.
Pits
Sketches of various pit shapes found are shown in Fig-
ure 3.6. Feature 17 was the only pit found which was on
the main floor outside the Big House. Features 19, 30, 79,
and 89 were within the Middle Structure; Feature 91 was
in the North Structure. This known distribution may be
close to the original distribution of relatively deep pits
since at least the bottom of such features would have been
noted in the heavily disturbed areas. If the known distri-
bution is the original distribution, then such pits occurred
only on the west side of the Big House and the portico.
Shallowly cut features, if any, were lost.
The apparent major difference between the contents of
the pits and the floor deposits discussed above was the
presence of pottery in three pits. Probably one cord-
marked vessel was represented in Feature 19 and two in
Feature 30, while Feature 89 contained cordmarked and
Hopewell series sherds, parts of possibly six to nine ves-
sels (see pottery analysis in chapter 4). Feature 19 also
contained 3 mammal canines; 75 unidentified bone frag-
ments (40 burnt, 35 unburnt); 7 shell fragments; and a
mica fragment. Feature 30 contained similar items. These
were 3 mammal canines; 1 fragment of deer ulna; 96 other
burnt bone fragments, 9 unburnt; 13 shell fragments (see
shell analysis in chapter 8); 2 mica fragments; 15 flint
flakes; 4 fire cracked rocks; and 4 fossils. Feature 89 con-
tained 7 burnt canines; 55 burnt unidentified bone frag-
ments, 45 unburnt; 6 shell fragments; 10 pieces of worked
flint; 2 bladelets; 1 clay bead; 12 fossils; and numerous
small mica fragments. Although Features 79 and 91 were
heavily disturbed, they contained some charcoal as did
the other pits. Also within Feature 91, 12 unidentified
bone fragments and 30 identified shell fragments were
found.
TABLE 3.2
Radiocarbon Dates
Radiocarbon
Lab. ft
Provenience
Material
Years (B.P.)
Comment
DIC-661
Fea. 17
Wood Charcoal
1490± 65
DIC-662
Under Fea. 3
Wood Charcoal
2150 ± 155
sample size, 4 g**
DIC-663
Fea. 19
Wood Charcoal
I620± 65
D1C-664
Fea. 30
Wood Charcoal
1500± 60
D1C-664 Rerun
Fea. 30
Unused Benzene from original burn
1600 ± 65
D1C-665
PM. 32
Wood Charcoal
1820 ± 70
D1C-801
PM. 36
Wood Charcoal
1900 + 460
sample too small; indicator date only
-500
DIC-802
Fea. 3 1
Wood Charcoal
1630+ 70
DIC-860
Fea. 53A
Wood Charcoal
1500+ 50
D1C-II89
Fea. 69
Charcoal [Gleditsia (Sp) 100%]*
Modern
burnt honey locust root over Fea. 68
DIC-1 187
Fea. 62
Charcoal
[Carya (Sp) 55%
Quercus (white) 40%
Quercus (Sp) 5%]*
1770+ 50
DIC-I 188
Fea. 8 1
Charcoal
[Carya (Sp) 55%
Fraxinus (Sp) 15%
Juglans (Sp) 20%
Quercus (white) 10%]*
1 140 + 60
D1C-I 190
Fea. 55
Charcoal
[Quercus (red group) 100%]*
1110+50
D1C-1635
Fea. 56
Burnt bone
1200+ 65
♦Percentages of 20 random pieces identified by University of Michigan Ethnobotanical Lab.
♦♦Using long-term average background count, the date is 1980 ± 1 55 b.p.
Clay
FEA. 79
Fig. 3.6. Sketch profiles of pit features: Fea. 17, 120
cm N-S X 72 cm E-W, 45 cm deep; Fea. 30, 48 cm N-S
X 53 cm E-W, 77 cm deep; Fea. 79, 40 cm X 37.5 cm, 9
cm deep, N-S diagonal 43 cm, E-W diagonal 44 cm;
Fea. 89, top 34 cm N-S X 32 cm E-W, base 55 cm N-S
X 53 cm E-W, depth 71 cm. (Grid north referenced.)
36
N’OMI GREBER
No. 39
Feature 17, which was in situ, showed careful cutting,
filling, and covering. Within it were found 4 fragments of
human bone, fragments of at least 2 canines, 1 bird bone,
108 mammal bone fragments, 129 unidentified bone
fragments, and 2 flint flakes. Large amounts of wood
charcoal were found as well as wild food remains (see
flora analysis in chap. 5). The rather idiosyncratic nature
of the pit profiles and the limited quantities of artifacts
found within the pits appeared to reflect single use. They
likely contained the remains of one cultural event, which
was probably associated with the performance of a ritual
or ceremony.
Other Features
The historic pits noted in Fig. 2.6 are obvious, but of
course are not the only spots which were literally dug up
prior to 1976. Feature 24 was, as previously described,
associated with the work of Squier and Davis ( 1 848) fol-
lowed by just about everyone else. The digging style
shown in Feature 77 I would identify with “The Boys,”
even though Putnam’s notes may indicate another loca-
tion for the schoolboys’ second pit. The remaining pits 1
assume were dug about post holes, aboriginal pits, or
other possibly deep features.
The bone concentrations were usually associated with
smears of charcoal in the backdirt. No artifacts can be tied
to these.
The disturbed graves (see Fig. 2.6) were very disturbed.
The location of several outside the Big House but within
the area defined by the portico posts does give useful in-
formation on the accepted use of these spaces. Within
Feature 84 the remains of an extended infant skeleton
were found. These remains are discussed in chapter 7 and
in Greber (1979). Mica fragments were found in the dirt
above the grave, but no artifacts were found associated
with the remains in 1977.
Two in situ burials were found in the outer areas of the
site. Feature 56 was a deposit of cremated human remains
centered at N534.4 E512.2 within a secondary mound fill
(see Fig. 2.5A Section CCT Associated with the remains
were a copper plate (22.6 X 12.7 cm) (see Gadus 1979), a
small copper adze (9X 5 X 1.06 cm thick) with some fabric
still intact, and a slab of sandstone. This slab was nearly
rectangular with maximum dimensions 84.4 X 32.5 cm,
weight 54.7 kilos. There was a small amount of pecking at
the corners, but the 5 mm depth of the weathering rind
suggests that this was a weathered rock before it was
placed in the second mound fill (Paul Clifford, Curator of
Geology, CMNH, personal communication, 1977). No
evidence of a prepared grave was found in the gravelly
clay matrix in which the bones were found (see Introduc-
tion to the Stratigraphy in this chapter).
The second burial found had been placed at the bottom
of a pit which was under Feature 69A (see Fig. 2.5B Sec-
tion FF'). This was a bundle burial which was accompa-
nied by a cut marine shell. David Morse, whose identifica-
tion and analysis of the molluscs recovered is given in
chapter 8, has separately described this shell.
The shell has been modified by the removal of the exterior
spikes and trimmed along the edge; probably for use as some
type of container.
This shell belongs to the species Busy con contrarium
(lightning welk) and is native to the western Atlantic coast
from North Carolina to Florida. The animal lives primarily
in shallow waters. Of all the species in the genus Busvcon ,
this species is one of the most southerly and restricted species
described in the zoological literature. Originally the shell was
a whitish buff in color with violet and brown vertical streaks.
The shell measures 17.5 cm in length which is average for
shells of this species today, but is one of the smallest speci-
mens known from the Edwin Harness Mound, (personal
communication, 1978)
A variety of burned hardwoods as well as a fragment of
Zea mays and other seeds were found in flotation samples
taken from the earth surrounding the redeposited bones
(see Table 5.3).
Feature 63 was a small pit centered 2 m south of the
edge of the pit containing Feature 60, in the east wall of
Backhoe Trench 4. The pit appeared to be generally oval
(21 X 35 cm) and 49 cm deep. Charcoal flecks were found
in the lower half of the pit within a dark reddish sandy
clay matrix (5YR 4/2 yellowish red). No artifacts were
found.
Feature 37 was a tap root centered at N522.75 E508.7
near the edge of Feature 1.
Radiocarbon Assays
The results of a series of radiocarbon assays on samples
collected during the 1976-1977 excavations are given in
Table 3.2. Two of these are not useful DIC-801, an indi-
cator date, and DIC-1 189, which is modern. Feature 69 is
an extensive deposit of dark earth and rock at the outer
edges of the mound with occasional discrete concentra-
tions of darker earth, fire cracked rock, and charcoal.
Apparently a modern root intruded into this feature di-
rectly over but not into one of these darker deposits (Fea-
ture 68). DIC-1 190 and 1188 are dates from deposits
which are similar to Feature 68 and in similar contexts.
The state of the art in radiocarbon work has changed
even since 1977, so that samples submitted today can be
smaller than the suggested 10-g size with no adjustments
necessary in procedures. At the time of the DIC-662 as-
say, 4 g of submitted charcoal resulted in a very small
prepared sample (0.6869 g). For such a prepared sample
size the background count used in calculating the radio-
carbon years may be based on the average background
count over a month rather than the average for only the
1983
EDWIN HARNESS MOUND
37
two days adjacent to the day on which the sample count is
taken. (Irene Stehli, Director, Dicarb Radioisotopes,
Inc., personal communication, 1977). Either of the calen-
drical dates (200 b.c. ± 155or30B.c.± 155) will be strati-
graphically in order with the remaining dates. The sample
was taken from a small deposit of burned material be-
tween Feature 3 (the central main activity floor) and Fea-
ture 65 (sandy clay layer directly below Feature 3) as dis-
cussed more fully in the section on the horizontal extent
of floor strata.
DIC-663 (a.d. 330 ± 65) and DIC-664 (first run a.d. 450
± 65, second run a.d. 350 ± 65) are from pits within the
Middle Section. These pits were cut into and through the
main floor (Feature 3). DIC-665 (a.d. 1 30 ± 70) is from a
post hole immediately northeast of the Northern Section.
DIC-661 (a.d. 460± 65) is from a pit cut through the main
floor west of the North Section, while DIC-802 (a.d. 320
± 70) is from a deposit of burned material on the same
floor, again, west of the Middle Section. North of the en-
tire building, DIC-860 (a.d. 450 ± 50) comes from a sim-
ilar deposit of burned material, and DIC- 1 187 (a.d. 1 80 ±
50) comes from burned hardwoods found in a small basin
(Feature 62) constructed over a post hole (see Fig. 2.6 for
locations).
The date for Feature 56 ( DIC- 1 635, a.d. 750 ± 65) is
again stratigraphically in order since this burial was in an
outer stratum of the mound construction. The data is very
close to the dates from the concentration of black earth,
burned rock, and charcoal in the outer strata (DIC-1 188,
a.d. 81 0± 60 and DIC-1 190, a.d. 840 ± 50). There can be
questions concerning the retention of humic materials in
bone use for radiocarbon assays; the standard procedures
used for removing these younger contaminates from
wood charcoal destroy bone collagen. This destruction
would give too young a date. It seems reasonable to as-
sume that bone, particularly burned bone, will not ac-
cumulate as much humic matter. The bone used for DIC-
1635 was somewhat protected from water by its general
location, which was within a gravelly, easily draining soil
and beneath an upper strata of heavy stone. In the contin-
uing series of dates from Ohio Hopewell sites, other sam-
ples of bone and charcoal from a single provenience are
being processed. For now, this single date is in proper
sequence for site stratigraphy, but it is considered ten-
tative.
The application of corrections for the Seuss Effect does
not significantly change the calendrical years at the time
period of this study. For example, using the Arizona cor-
rections (Damon, Ferguson, Long, and Wallick 1974)
DIC-1635 becomes a.d. 770 ± 83. Other examples using
both the Arizona and the Masca correction procedures
for dates from Ohio Hopewell sites have been given pre-
viously (Greber 1976: Fig. 24). The increase in the stan-
dard deviations, which is a result of the correction, may be
worth consideration. Comparative materials will be dis-
cussed in more detail in the report conclusions.
Appendix 3. 1
Thin Section Analysis of Soil Samples From
Edwin Harness Mound
Jerry M. Bigham
The Ohio State University
Sample I. (SW end of backhoe trench 7; N W wall. Feature 3C). 1
found no evidence of clay films in this sample and little indica-
tion of the structural aggregation which characterizes undis-
turbed soil materials. Vertical thin sections revealed several con-
tinuous lenses of charcoal (one very prominent), but charcoal
fragments were also dispersed throughout the sample. Iron
stains often, but not always, paralleled the charcoal lenses. The
thin, lenticular nature of these iron stains is quite uncharacteris-
tic of soils. Since soil will often redden when heated, perhaps
these stains represent materials which have been fired. In addi-
tion to these features, I also noted several filled worm casts, but
I cannot say if the activity was recent or relict.
Sample II. (As for I, but directly below Feature 3C). I saw no
evidence of clay films or charcoal in either vertical or horizontal
thin sections from this sample. Iron stains, however, were com-
mon. In soils, iron oxides often segregate to form mottles and
concretions. In contrast, the stains in this sample usually oc-
curred as coatings on large grains and/or rock fragments
(generally siltstone). In some instances, the stains proved to be
completely oxidized fragments of unknown origin. The contacts
between the iron stains and the surrounding matrix were gener-
ally quite abrupt, suggesting perhaps a mixture of materials.
Sample II also contained numerous worm casts and was quite
porous (but not as much so as Samples 1 and III).
Sample III. (N522.5 E487.5, South Wall, Feature 1). Most of
this sample was too loose to impregnate. However, we did man-
age to save one unit consisting of a siltstone fragment overlain
by unconsolidated material. In thin sections, the siltstone frag-
ment appeared to be quite fresh with no evidence of iron stains
or weathering rinds. The unconsolidated material was exceed-
ingly porous, and I saw no evidence of clay skins, iron stains, or
charcoal.
In all three samples the sand grains appeared to be fresh, rela-
tively angular and unsorted. All three samples also contained
more feldspar minerals (I, III, II) than I am accustomed to see-
ing in soils. I doubt, however, if there is any significance to this
observation. All samples were organic stained to some extent,
but I saw little evidence of primary (undecomposed) root
tissues. Subsurface layers in soils of this region often contain
“clay skins” along root channels and the surfaces of soil aggre-
gates due to the dispersion and downward movement of colloi-
dal particles in waters percolating through surface layers. Trans-
located clay is, therefore, indicative of soil formation. I saw no
evidence of translocated clay in any of the samples I examined.
Without further knowledge of the excavation site, 1 would con-
clude that these samples were taken from disturbed and/or re-
cently deposited materials that have not been subjected to soil
forming processes over a significant period of time. This state-
ment is based on the dark color, absence of clay films, unusual
porosity, charcoal content (where present), and lack of natural
aggregation (soil structure) in these materials.
38
N’OMI GREBER
No. 39
References
Baby, Raymond S.
1971 Prehistoric architecture: a study of house types in the
Ohio Valley. Ohio Journal of Science 71(4): 1 93— 198.
Baby, Raymond S., and Suzanne M. Langlois
1979 Seip Mound State Memorial: nonmortuary aspects of
Hopewell. In Hopewell archaeology: the Chillicothe
conference , edited by David S. Brose and N’omi
Greber, pp. 16-18. Kent State University Press, Kent,
Ohio.
Brown, James A.
1979 Charnel houses and mortuary crypts: disposal of the
dead in the Middle Woodland Period. In Hopewell
archaeology: the Chillicothe conference , edited by
David S. Brose and N’omi Greber, pp. 21 1-219. Kent
State University Press, Kent, Ohio.
Brown, James A., and Raymond S. Baby
1966 Mound City revisited. MS on file. Department of
Archaeology, Ohio Historical Society, Columbus,
Ohio.
Chapman, Jefferson, and Bennie C. Keel
1979 Candy Creek-Connestee components in eastern Ten-
nessee and western North Carolina and their relation-
ship with Adena-Hopewell. In Hopewell archaeology:
the Chillicothe conference , edited by David S. Brose
and N’omi Greber, pp. 157-161. Kent State University
Press, Kent, Ohio.
Damon, P. E., C. W. Ferguson, A. Long, and E. I. Wallick
1974 Dendrochronologic calibration of the radiocarbon
time scale. American Antiquity 39(2) part 1:350-366.
Fowells, H. A.
1965 Silvics of forest trees of the United States. Division of
Timber Management Research Forest Service. U S.
Department of Agriculture Handbook No. 271.
Gadus, Eloise F.
1979 The Harness copper plate. Ohio Archaeologist 29(3):
27-29.
Greber, N’omi
1976 Within Ohio Hopewell: analysis of burial patterns
from several classic sites. Ph.D. dissertation. Depart-
ment of Anthropology, Case Western Reserve Uni-
versity. University Microfilms, Ann Arbor, Michigan.
1979 A comparative study of site morphology and burial
patterns at Edwin Harness mound and Seip mounds 1
and 2. In Hopewell archaeology: the Chillicothe con-
ference, edited by David S. Brose and N’omi Greber,
pp. 27-38. Kent State University Press, Kent, Ohio.
Greber, N’omi, and Dennis P. Griffin
1982 Comparison of excavations and subsurface remote
sensing data from sections of the Seip Earthworks
Complex, Ross County, Ohio. Paper presented at the
1982 Annual Meeting of the Southeastern Archaeo-
logical Conference, Memphis, Tennessee, October
28-30.
Hudson, Charles
1976 The southeastern Indians. University of Tennessee
Press, Knoxville, Tennessee.
Mills, William C.
1903 Diary. On file. Library Archives, Ohio Historical
Center, Columbus, Ohio.
1907 Explorations of the Edwin Harness mound. Ohio Ar-
chaeological and Historical Quarterly 16:1 13-193.
Petro, James H., William H. Shumate, and Marion F. Tabb
1967 Soil survey, Ross County, Ohio. United States De-
partment of Agriculture Soil Conservation Service in
Cooperation with Ohio Department of Natural Re-
sources.
Putnam, Frederick W.
1884 Field notes on excavation of Liberty Group, Ross
County, Ohio. On file, Peabody Museum, Harvard
University.
1885 Explorations of the Harness mounds in the Scioto
Valley, Ohio. In Peabody Museum 18th and 19th An-
nual Reports (1884-1885), bound in Peabody Mu-
seum Reports 3(5-6):449-466.
Shane, Orrin C., Ill
1973 Report on excavations at the High Banks earthwork,
Ross County, Ohio. Paper presented at the Annual
Meeting of the Ohio Academy of Sciences, Cleveland,
Ohio.
Squier, George Ephram, and E. H. Davis
1 848 Ancient monuments of the M ississippi Valley. Smith-
sonian Contributions to Knowledge 1 . Washington,
D.C. Reprinted 1973 with introduction by James B.
Griffin as Antiquities of the new world: early explora-
tions in archaeology (Vol. 2). A.M.S. Press, New
York, for Peabody Museum, Harvard University.
Willoughby, Charles C.
1916 The art of the great earthwork builders of Ohio. An-
nual Report, Smithsonian Institution, Washington,
D.C.
4 THE CERAMIC COMPLEX
JAMES B. GRIFFIN
In this discussion of the pottery from N’omi Greber’s
1976-1977 excavations of the Edwin H. Harness Mound
floor, the specimens have been described and identified
according to their location (see Appendix 4. 1). This col-
lection has provided information on some of the pottery
which was in existence before the completion of the
mound and in some instances before different sections of
the mound floor had been formed. These pottery frag-
ments are not the pottery complex of the Harness people
but a very small segment of their pottery production. This
mound has had a long history of excavations since at least
the mid- 1840s, and the pottery preserved from those ex-
cavations in the Peabody Museum of Archaeology and
Ethnology, Harvard University, and the Ohio Historical
Society probably represent only a small part of the speci-
mens in the area excavated. 1 did not find any pottery
from Harness in the British Museum Squier and Davis
Collection. In addition, the ceramic material from the
several segments of the Harness mound area forms a
small portion of the ceramic production of the popula-
tions who lived at and near the Harness site and partici-
pated in the activities centered there. The same is true of
all of the other excavated major and minor Hopewell sites
in Ohio. Only the Turner site has a pottery sample which
can be regarded-as representative, yet that collection also
presents many difficulties of interpretation. If the major
Ohio Hopewell sites existed over a period of several
generations with a population associated with each of
some hundreds of people, their year-by-year pottery pro-
duction and breakage would reasonably come to a total
far in excess of that recovered from the excavations. The
McGraw site is one small unit of a village presumably as-
sociated with a small Hopewell earthwork, and it pro-
duced almost ten thousand sherds. The Turner and Mar-
riott sites had 3,806 sherds according to Prufer’s (1968)
analysis of the pottery. There are about 100 counted ex-
amples from the Edwin Harness floor in the present col-
lection, although the total is somewhat greater, for, when
possible, sherds belonging to a single vessel were counted
as one example. Prufer’s count, from the Harness mound
excavations at the Peabody Museum and the Ohio His-
torical Society, plus the Russell Brown collection, totals
some 1, 1 57 sherds.
In the Greber collection of 98 sherds, McGraw Cord-
marked is the dominant surface finish and McGraw Plain
a distant second (Table 4. 1 ). There are a few examples of
the Hopewell Rim and Chillicothe Plain Rocker-stamped.
There are a few examples of Turner Simple Stamped A,
but none of the micaceous or other sand tempered simple
ware, although there were some from the earlier collec-
tions. There are a few sherds of Turner Check Stamped,
almost certainly from one vessel. The most unusual pot-
tery is the vessel represented by limestone tempered
Turner Simple Stamped A, sherds which could well have
been an import. I have not provided a percentage figure
for the identified “types” because they would be even
more misleading than usual and can be easily obtained by
anyone wishing to do so. A description of each sherd is
given by provenience in Appendix 4.1. Photographs are
presented in Figures 4. 1-4.4.
TABLE 4.1
Summary Tabulation of Harness 1976-1977
Pottery Collection
Sherd Count*
McGraw Cordmarked
Body
56
McGraw Plain
Body
7
Rim
3
Chillicothe Plain Rocker Stamped
Body
4
Hopewell Rim
2
Turner Simple Stamped A
Rim
1
Body
3
Turner Check Stamped
Body
I
Unidentifiable
21
Total
98
* All pottery pieces which can be fitted together into one unit are
considered to be one sherd.
In 1968, Olaf H. Prufer published an analysis of Ohio
Hopewell pottery from a study he had made some years
earlier. His typology will be followed in this study in order
to enhance comparability. In 1943 Richard G. Morgan,
Curator of Archaeology of the Ohio State Museum (now
the Ohio Historical Society), and I made a similar study,
which I used briefly in a report on Adena pottery (Griffin
1945), but our complete study was not published. Our
sherd count is presented in Table 4.2 and is given in Pruf-
er’s terminology where possible. We did not use type
names in our original study except for the identification
of the highly distinctive Hopewell Rim style.
Fig. 4. 1 . Hopewellian and check-stamped sherds: a(24E), 6(40A),
and d{ 225B), Hopewell Rim; c(2E), e(24C), J[ 103E), g(103E),
/?(35B), y'(24H), A:(24G), and /(24B), Chillicothe Plain Rocker
Stamped; ;'(103E), Hopewell sherd with incised line; m( 290C),
n( 290C), o(290G), p(290C), and g(290C), thin check-stamped
sherds.
Fig. 4.2. McGraw Cordmarked, McGraw Plain, and unidenti- Cordmarked; e(290C),y(138A, 23 IB) unidentified, g(2B), h( 1 23),
fied sherds: a(308C), />(2D,I2H), c(37A),and ^(309A), McGraw /(219V), and /(229C). McGraw Plain.
42
JAMES B. GRIFFIN
No. 39
TABLE 4.2
Griffin-Morgan Pottery Count of
Ohio State Museum Harness Collection
Sherd Count*
Sub-total Total
McGraw Cordmarked
Body
78
Rim
2
80
McGraw Plain
Body
18
Rim
7
25
Turner Check Stamped (square)
Body (all from one Footed vessel)
23
Turner Check Stamped (diamond)
Body
8
Rim
3
34
Turner Simple Stamped A
Grit Tempered Body
6
Limestone Tempered Body
7
Limestone Tempered Plain Rims
(Probably Turner Simple Stamped A)
3
16
Hopewell Rims
4
4
Chillicothe Plain Rocker Stamped
Body
17
Dentate Rocker Stamped
Body
8
Banded Dentate Rocker Stamped
Body
1
Zoned Dentate Rocker Stamped
Body
2
28
Total
187
187
* All pottery pieces which can be fitted together into one unit are
considered to be one sherd.
Discussion
McGraw Cordmarked
The McGraw Cordmarked vessels from Greber’s Har-
ness floor collection display some variety. On some the
cord impressions are close together and on others are
widely spaced. Some of the sherds were smoothed while
others were not. There is also some variation in thickness,
from 4 to 8 mm. There is variation in the amount and
character of the grit temper. All of this variation suggests
somewhat less attention to the acquisition of raw mate-
rials and vessel manufacture than was evident in sherds of
the Hopewellian Series. The same can be said regarding
the few McGraw Plain sherds in this collection. The
McGraw Cordmarked sherds illustrated by Prufer ( 1968:
PI. 3) are representative of the sherds at hand.
There was a time when some archaeologists did not like
to think that cordmarked pottery was a part of the Ohio
Hopewell Complex. However, this pottery is a strong
component of almost all known Ohio Hopewell sites ex-
cept Tremper. While usually regarded as "utility’' ware,
which it certainly was, cordmarked pottery is also found
in burial association as whole or broken vessels at Turner,
Hopewell, and the Martin Mound in Coshocton County
in the upper Walhonding valley (Mortine and Randles
1978), where it was the surface finish on a small tetrapod
vessel.
The size of the reconstructed McGraw Cordmarked
vessel 20A, of the McGraw Plain vessel 27 1 0, and of the
Turner Simple Stamped 308A suggests that they were
probably whole vessels which had been used for some
function shortly before they were deposited. What that
function (or functions) was is not clear. The same may be
said for the larger Hopewell Rims and plain rocker-
stamped sherds which seem to be from the same vessel.
Hopewell Style Pottery
The Hopewell Rims and Chillicothe Plain Rocker-
Stamped sherds are the only representatives of Hopewell
style pottery in Greber’s collection from the Harness
floor. This is well-made pottery, and the slight camber of
the rim I regard as indicating the vessels were made fairly
early in the life-span of this complex. The reconstructed
body fragment (Prufer 1968: PI. 5a) is almost certainly
part of the same vessel as my Figure 4. Ic, f-h. The illus-
trated Hopewell Rims from the older Harness collections
are not as well made and appear to have a more pro-
nounced camber (Prufer 1 968: PI. 5b-c). The Hopewell
Zoned-dentate Rocker-stamped vertical compound jar
from Putnam’s Harness 4 in the Peabody Museum is a
unique specimen, and the large rim section from Russell
Brown Mound 1, one would think, might well fit into the
reconstructed area on the Peabody vessel (cf. Prufer
I968:P1. 2 and 46). This would raise the possibility of Put-
nam’s Mound 4 being the same mound as Frank Soday’s
Mound !. Close comparison, however, by Prufer and
others does not support the idea of a single vessel. There
are a few other sherds of the Hopewellian Series from
Harness and the Brown mounds, but my impression is
that Harness does not have the variety of this ceramic
complex that is found at Seip, Turner, or Mound City.
The McGraw site ceramic collection is also much more
varied, and I interpret its Hopewellian Series as later than
the available material at Harness, Mound City, and Seip.
The Harness examples should be close to the initial ap-
pearance of the Hopewellian Series in the Scioto Valley.
There is no evidence in southern Ohio for a develop-
mental sequence of ceramic decoration which could have
developed into the Hopewell Zoned decorated style.
While some of the late Adena tablets in Ohio and Ken-
tucky do present conventionalized bird designs, the only
northern area where zoned stamped decoration appears is
in the lower half of the Illinois Valley, where it is found
apparently in a time period which precedes the Ohio
Hopewell development. I have not seen a vessel or any
vessel fragments from Ohio which could be correctly as-
1983
EDWIN HARNESS MOUND
43
signed to an Illinois source. On the other hand, such ves-
sels from the southern half of the Illinois Valley were car-
ried into northern Illinois, Michigan, and Wisconsin. The
most logical area for the generation of the Hopewell
Zoned style is in the Illinois Valley even though it is by no
means a certainty.
Turner Simple Stamped B
The known distribution of Turner Simple Stamped
ware recorded by Prufer is given in Table 4.3. There are
no examples of Turner Simple Stamped B from Greber’s
excavation of the Edwin Harness Mound, nor were any
identified from the collections made by earlier excavators.
There were none from Russell Brown Mounds 1 and 2.
There were none reported by Prufer (1968) from Hope-
well, Rockhold, Ater, or Marriott 1, which is part of the
Turner site. No examples are reported from sites ex-
amined in the Scioto Valley survey (Prufer 1967) nor were
any mentioned as present from the excavations of the
McGraw site (Prufer et al. 1965).
At the Turner site, which, of course, has the strongest
representation of Turner Simple Stamped B (see Table
4.3 and Prufer 1 968: PI. 34a and c, PI. 40, PI. 44d and g),
TABLE 4.3
The Occurrence of Turner Simple Stamped B
in Ohio, based on Prufer ( 1968)
Site
Provenience
Body
Rim
Tetrapod
" 'feet "
Total
Turner
Mound 1 , fill
2
2
Mound 3, unit 2
unit 4
12
3
Mound 4, unit 1
40
4
3
Embankment trench
13
2
Cemetery, unit 6
2
Ginther
Mound
4
2
Tremper
2
1
Mound
Mound 13
1*
City
General
7
Seip
Mound 1,
unit 957/ 161
3
1
1
unit 957/ 237 & 238
1
unit 957/260
1
Mound 2, general
5
1
General
10
1
1
Fort
Ancient
1
4
Fort Hill
1
Russell
Brown
Md. 3
3
108
12
14
134
Willoughby had noticed that “some of the clay used in
making the smaller and more delicate vessels was tem-
pered with sand instead of crushed stone.” He mentioned
a tetrapod base and ten other “feet.” He also observed of
the complicated stamped sherds from Turner that “sherds
showing ornamental paddle marks were extremely rare.
Such vessels may have been brought from the southern
Appalachian region, or they may possibly have been
made by captured women from the South” (Willoughby
and Hooton 1922:93).
Prufer published a type description for Turner Simple
Stamped B (1968), which he included in his Southeastern
Series. He felt that type was imported into Ohio from the
southeast, primarily because of its sand temper, which is
rather rare in the Ohio Valley, and because of the narrow
stamp impressions which were similar to Deptford and
Mossy Oak Simple Stamped of the Georgia area. He also
recognized, as had others, similarities to Paintville Simple
Stamped of eastern Kentucky and to Bluff Creek Simple
Stamped of northwest Alabama. The connections to the
north Georgia area were thought to be particularly
strong, and this area was considered as the probable
source of imports into Ohio because of the presence in
some of the examples of small mica flakes in the paste.
Since the publications of Prufer, excavations in south-
western North Carolina and in eastern Tennessee have
produced ceramic data which serve to alter significantly
interpretations of the source for Turner Simple Stamped
types. In southwestern North Carolina a number of Mid-
dle Woodland sites have been excavated and identified as
members of a Connestee phase. One of the sites is Garden
Creek Mound No. 2. One of the ceramic types is Connes-
tee Simple Stamped with about 300 examples. Keel
(1976:1 10) says that “some of the sherds . . . could be
classified as Turner Simple Stamped. ’’One lower rim and
upper body (Keel 1 976: PI. 16d) has annular punctates in a
horizontal row at the base of a smoothed lower rim. In his
discussion of trade pottery at Garden Creek, Keel identi-
fies the same fine sand or finely ground limestone tem-
pered pottery as Turner Simple Stamped (1976:120) and
illustrates (PI. 181) two lower rim and upper body sherds
with annular punctates in the horizontal row at the base
of a smoothed rim which he calls Turner Simple Stamped
B. But in the description of Tennessee types he states that
the 92 limestone tempered sherds are all regarded as im-
ported from Tennessee, although other sources were pos-
sible. He particularly identifies the Ice House Bottom site
with its Connestee material as similar to the limestone
Connestee phase pottery at Garden Creek (1976:1 18). I
am puzzled why Chapman and Keel (1979:157) regard
sherds identified as Turner Simple Stamped sand tem-
pered as probably being from Ohio, when Prufer and at
least some other archaeologists have regarded this type as
trade pottery from the southeast into Ohio. If the sherds
illustrated by Keel ( 1 976: PI. 171) as Turner Simple
Stamped B are typical, then there are no known examples
at Edwin Harness Mound of this type. In comparing the
*whole vessel
44
JAMES B. GRIFFIN
No. 39
lower rim punctates at Garden Creek with the Mound
City whole vessel from Mound 13, one should note that
the former punctates are annular while the latter are verti-
cally placed hemiconical punctates. It may be doubted
that any simple-stamped sherds at Garden Creek were
traded from Ohio.
There are, however, Hopewell Cross-hatched rims and
plain and dentate rocker-stamped sherds that might well
be Ohio imports along with Ohio Flint Ridge blades. It is
of special interest to note that Keel identifies the figurine
specimen at Garden Creek as having been made from lo-
cal paste similar to that of the Connestee pottery.
At the Ice House Bottom site in Monroe County, Ten-
nessee on the Little Tennessee River there is a ceramic
complex which includes simple-stamped pottery that has
even more interesting similarities to the Ohio and North
Carolina pottery with the same surface finish (Chapman
1973). The Connestee Simple Stamped sherds from this
site were so named because of their striking similarity to
the Garden Creek and other North Carolina sites with
Connestee Series pottery. In contrast to Garden Creek,
the Ice House Bottom Collection has both annular as well
as angular punctates, and it is the later technique which is
found at Mound City, Ohio in Mound 13. The rectangu-
lar punctates also appear on what could be Connestee
Plain rims; both annular and angular punctates are also
found on Connestee Brushed.
The Connestee Series pottery at Ice House Bottom
constitutes about 20% of the ceramic assemblage, while
the Candy Creek Series with limestone temper is, at 71%,
the dominant pottery. Chapman uses the term “Bluff
Creek” for the simple-stamped pottery with limestone
temper following the practice initiated by Haag ( 1 942a) in
northwestern Alabama and subsequently followed by
others for materials from that area. Its use in eastern Ten-
nessee, however, is perhaps a misnomer even though there
are obvious resemblances in the material from the two
areas. The Ice House Bottom Bluff Creek Simple Stamped
pottery does not have punctates similar to those of Con-
nestee, and the vessel rims are primarily vertical instead of
flaring, which is the dominant form on Connestee Simple
Stamped. Over half of the lips of Bluff Creek are notched
transversely, which does not appear on Connestee wares
at this site or in Ohio. Ice House Bottom perhaps does
have a simple-stamped ware which is close to the lime-
stone tempered vessel from Harness. The rim and lip
treatment at Harness is not, however, either illustrated or
described by Chapman or by Gleeson (1970) from earlier
excavations.
The excavations at the C and O Mounds and village site
in Jonathan County, eastern Kentucky, on Levissa Fork
of the Big Sandy River recovered a small number of
simple-stamped sherds. However, this pottery, described
as Paintsville Simple Stamped (Haag 1942b), presents
some problems in interpretation for several reasons. In
his description of this type Haag says that perhaps all of
the sherds could be from one vessel. The large rim (Haag
1 942b:Fig. 17:3) has a high flaring upper rim and annular
punctates in a horizontal row at the base of the lower rim.
Also an exact provenience within the two excavated
mounds and village is not given. Since the flint projectile
points from the site range in age from Early Archaic to
perhaps Fort Ancient, and the pottery range is from an
Adena complex to perhaps Fort Ancient, the attribution
of Paintsville Simple Stamped is a bit difficult. There are
Montgomery Incised examples and the Adena pottery
complex as a whole would seem to be late. The Paintsville
Simple Stamped is probably a trade vessel or vessels from
eastern Tennessee. This would fit well with the presence of
mica at the C and O Mounds and its probable derivation
from western North Carolina. There are no specifically
Hopewellian artifacts at the C and O Mounds, and if the
Adena occupation there were in existence during the life-
span of Ohio Hopewell, one would expect to find some
indication of such contemporaneity.
A look at one final possible southern connection indi-
cates that the three sand tempered simple-stamped sherds
illustrated in the Tunacunnhee site report from northwest
Georgia (Jefferies 1976) cannot be considered close to
Turner Simple Stamped B, and the one limestone tem-
pered simple-stamped sherd is not close to Turner Simple
Stamped A.
Turner Simple Stamped A
Turner Simple Stamped A sherds at Edwin Harness
probably all belong to one vessel (cat. nos. 308A and 237C,
plus other limestone tempered examples), except for two
grit tempered specimens with a thinner body and less con-
spicuous lands and grooves. The sherds of the large vessel
are very close in appearance to the sherds illustrated by
Prufer ( 1968:P1. 4b and c) from older Harness collections
and from Russell Brown Mound 2, which was a part of
the Harness Earthwork Complex. For a listing of the oc-
currence of Turner Simple Stamped A as stated by Prufer
see Table 4.4.
As already mentioned, the Ice House Bottom Bluff
Creek Simple Stamped illustrated examples do have an
appearance similar to Turner Simple Stamped A. This is
also true of other Middle Woodland sites in eastern Ten-
nessee. At the Pittman-Alder site in Marion County, a
short distance southwest of the bridge on which U.S.
Highways 41, 64, and 72 cross the Tennessee River, a mi-
nority ware (58 examples) of the Middle Woodland com-
plex is Bluff Creek (Faulkner and Graham 1965:P1.
XXIV). One of these rims has a notched lip (Faulkner and
Graham 1965:59). Only two sherds from the site were
identified as Benson Simple Stamped, which is the sand
tempered variant name for northeastern Alabama (Heim-
lich 1952). On the south side of the Tennessee River oppo-
site Pittman-Alder is the Lay site, where the continuing
excavations in the Nickajack Reservoir by the University
1983
EDWIN HARNESS MOUND
45
TABLE 4.4
The Occurrence of Turner Simple Stamped A in Ohio,
based on Prufer ( 1968)
Site
Provenience
Body
Rim
“feet" bn*
Total
Rockhold
1
1
Hopewell
Mound 2
1
I
Mound 17
1
1
General
40
1 1
40
Tremper
Mound
3
3
Seip
Mound 1 , unit 957/216
1
1
Mound 2
1
1
General
20
2
22
Fort Ancient
12
Turner
Mound 3, unit 4
1
1 1
1
Mound 4, unit 2
3
1
2
Mound 6
3
3
Mound 9, unit 2
1 1
Embankment
3
1
2
Harness
Edwin Harness Mound
13
8
5
Russell Brown Mound 1
34
2
36
Mound 2
153
4
157
Mound 3
4
4
293
8
4
305
*limestone temper
**grit temper
of Tennessee uncovered more and better examples, some
919 sherds, of Bluff Creek Simple Stamped (Faulkner and
Graham 1 966: PI. XIII, 37-38). They identified this type
with late Early Woodland in eastern Tennessee at the time
of writing the report, but a later placement in Middle
Woodland is probably a better assessment.
At the Doughty site in Loudon County, Tennessee, just
east of the 1-75 bridge over the Tennessee River, a few
sherds of a Bluff Creek Simple Stamped vessel were found
(McCollough and Faulkner 1 973: PL 21G). At the nearby
Higgs site there were also a few examples (McCollough
and Faulkner 1973:89-91). In the Tims Ford Reservoir
along the Elk River in Franklin County, Tennessee, at the
Mason site, a very minor type is Bluff Creek Simple
Stamped (Faulkner 1968:78 and PI. IXG). Other sites in
this area with a Middle Woodland limestone tempered
complex should also have this type as a minor component.
When working with the Norris Basin pottery in the mid
1930s, I identified simple-stamped specimens as “combed”
or striated (Griffin 1938). A few such examples were
found at Rock Shelter-cave Sites 3 and 12. Examples are
illustrated in W. S. Webb’s section of the report (Webb
1938:P1. 13) from Saltpeter Cave (Site 3), and from Wal-
lace Cave (Site 12) in my section of the report on Plate
1 52. While limestone tempered, none of the examples are
similar to the Harness simple-stamped example. The sites
were located in the Clinch-Powell river drainage a short
distance north of Norris Dam in Campbell County, Ten-
nessee. It is unfortunate that these shelters could not have
been excavated in a manner which might have aided a
recognition of successive Woodland occupations, for the
pottery suggests a time span from Early Woodland cer-
tainly well into the Middle Woodland time period. Two of
the open sites, the Harris Farm, Site 9, and the Cox
Mound, Site 19, also had a few specimens of simple-
stamped or brushed grit tempered surfaces (Griffin
1938:305).
In the lower Ohio Valley there are two sites which have
simple-stamped pottery pertinent to the presence of this
ceramic technique in Ohio. The Mann site in Posey
County, Indiana, southeast of Mount Vernon, is a large
site of some 200 acres with a strong Hopewellian compo-
nent(Adams 1949; Kellar 1979, 1973). In the collection of
pottery available to Adams, simple-stamped sherds were
a minority type, but there are indications from later Indi-
ana University work of areas of the site where it occurs in
much larger numbers. Adams emphasizes that this type at
Mann was sand tempered with small fragments of mica in
the paste. He referred to similar inclusions in southeast-
ern simple-stamped pottery, and he also noted that there
were levels at the Angel Site with micaceous sands identi-
fied by soil tests (Adams 1949:59). Kellar’s resume men-
tions the Mann site and illustrates a simple-stamped rim
with closely spaced lip notches (Kellar 1973:45-46). There
Fig. 4.3. Turner Simple Stamped and unidentified simple stamped sherds: a(308A), />(237C), c and
4309A), e( 1 38C), /(308A), gand h( 23 1 A), i(23 1 F), /(237C), and Ar(237E), Turner Simple Stamped A;
/and m(2!9A), unusually thin hard simple stamped.
Fig. 4.4. McGraw Plain and McGraw Cordmarked sherds: a(2710), McGraw Plain;
6(20A), McGraw Cordmarked.
48
JAMES B. GRIFFIN
No. 39
are also complicated stamped sherds from the Mann site.
One of these is illustrated by Kellar, and six presumably
from there are illustrated by Adams ( 1 949: PI. V).
In the Rutherford Mound in Hardin County, Illinois,
overlooking the Saline and Ohio river bottoms, M. F.
Fowler (1957) excavated a small simple-stamped vessel
with tetrapods in association with Burial 6, which was
placed with four other burials in the primary mound
about 2.5 feet above the mound floor. These burials may
or may not be significantly later than the effigy platform
pipes, panpipes, copper earspools, and axes placed with
burials on the mound floor. While there is a Carbon-14
date of about a. d. 432 ± 1 00, 1 believe it is too late for the
Hopewell items and for the tetrapod vessel, for this mate-
rial should date close to a.d. 100.
While the stamped paddle pottery seems to disappear
after the Hopewell occupation in southern Ohio, such is
not the case in the lower Wabash Valley. The survey re-
port for the Illinois side of the river by H. D. Winters
( 1 963) shows simple-stamped pottery which he has named
Embarrass (locally pronounced “Ambraw”) Simple
Stamped. Some of the lips are notched. This type is asso-
ciated with the FaMotte-Allison complex, which I believe
straddles the rather arbitrary dividing line between late
Middle Woodland and early Fate Woodland. Check-
stamped pottery is also associated with this complex.
Check-stamped
There are five body sherds with check-stamped impres-
sions from Harness. Prufer has called the Ohio Hopewell
examples Turner Check Stamped and placed them in his
Southeastern Series. Some of the Ohio examples may
really be trade vessels from the east Tennessee area. The
Harness examples of Greber’s collection, however, are
blackened on both surfaces, and their paste characteris-
tics are like the paste of the Hopewell Rims and Chilli-
cothe Plain Rocker-stamped Harness sherds and are
probably the product of Harness potters. I do not believe
I have seen very many similar check-stamped sherds from
other Ohio Hopewell sites and Prufer (1968) does not il-
lustrate any. I have not seen such sherds from southeast-
ern collections or in illustrations from that area. The con-
cept of check-stamping did, however, almost certainly
reach southern Ohio from the southeast, probably the
east Tennessee area. I think that the absence of check-
stamping on most of the Adena pottery in Kentucky,
Ohio, and adjoining areas probably has temporal signifi-
cance. One limestone tempered check-stamped sherd was
at the late Adena Wright mounds (Haag 1940:8 1) and one
grit tempered sherd from Jo9 (Haag 1942b:348). Haag il-
lustrates two check-stamped sherds from Wright Mound
6 ( 1 940: PI. 52n and r).
Prufer has identified a diamond check-stamped in his
Harness site study, a Turner Check Stamped at Russell
Brown Mound 1, and two of the same type at Russell
Brown 3. None of those are illustrated from the Brown
mounds but the Harness example is illustrated in Plate 4.
Such sherds also occur in very small frequencies at Rock-
hold; at Seip General and Seip Mound 2 General; at Fort
Hill; and at Turner Mounds 1, 3, 4, 7, and 9, the Great
Embankment, and the cemetery area. The lower rim and
body sherd from Turner Mound 4 is illustrated by Prufer
( 1 968: PI. 39a) and has a horizontal incised line separating
the plain lower rim from the stamped body. The sherd is
placed sideways on the plate instead of vertically. It has
stamp size and black outer surface similar to the five spec-
imens from the Edwin Harness floor (cat. no. 290C).
Prufer does not report check-stamped sherds from Hope-
well, Ater, Ginther, Mound City, Tremper, Fort Ancient,
or the Marriott Mound at the Turner site. There are cer-
tainly more sherds from the Turner excavations than
from all the other Ohio Hopewell sites where such sherds
are known to occur. This is probably in part because of
the much larger amount of pottery excavated and pre-
served from the Turner site. At Turner, check-stamped
sherds are found on both grit and limestone tempered
paste, the latter being most common. One sherd from
Seip General is limestone tempered, and the diamond
check-stamped sherd from Seip 2 General is sand tem-
pered. The few other sherds are grit tempered.
The check-stamped presence in Hopewell is even more
difficult to pin down in terms of its derivation than was
the simple-stamped technique. None of the Ohio exam-
ples are very large and very few have rims which might be
helpful. In eastern Tennessee, check-stamped is found
along with simple-stamped on many of the sites with
Middle Woodland components. This is probably the area
from which the check-stamped vessels or the concept was
moved north into southern Ohio. One large rim and body
sherd from the Turner cemetery is identified by Prufer
( 1 968 : PL 45a) as an Untyped Complicated Stamped be-
cause it has large squares with a raised central circular
area, which is unusual. Another location where such a
pattern appears in the southeast is near Savannah, Geor-
gia. Quite a few years ago J. C. Caldwell sent me illustra-
tions of sherds that he identified as the Oemler complex,
which had a limited distribution in the coastal area. Such
material is not mentioned or illustrated in the Waring
papers, although the Oemler site is identified on maps
(Williams 1 968: Figs. 35 and 37). The Oemler style of
stamping is regarded by some archaeologists to date
about 600 b.c. (DePratter 1979). A connection between
the Turner and Oemler examples is not implied, for they
are quite different in appearance, and the vessel shape at
Turner with its fairly high rim which is angled outward is
quite different in shape from any of the Oemler complex.
Three sherds with this design are known from the Mann
site in southwest Indiana (Kellar 1979: 103). Thediamond
variant of the check-stamp is uncommon both in Ohio
Hopewell and in the southeast but does occur at the
Yearwood site in southwestern Tennessee (Butler 1979:
1983
EDWIN HARNESS MOUND
49
Fig. 20.9). The ones at Harness in the older collections
and other Ohio examples are probably local products.
Complicated-stamped
While there are no complicated-stamped sherds at
Harness, they do occur at Seip and Turner particularly.
The illustrated examples appear to have their closest con-
nections to Early to Middle Swift Creek types. Sites in
eastern Tennessee of the Middle Woodland period which
have check-, complicated-, and simple-stamping, with ev-
idence of trade sherds and other items probably made in
Ohio, are the most likely candidates to be on the right
temporal level to have furnished those pottery specimens
or concepts to Ohio Hopewell.
As mentioned, complicated-stamped pottery occurs at
the Mann site (Kellar 1979:103). One rim illustrated by
Adams (1949: PI. V) has closely spaced lip notches while a
second has an undulating appearance. Adams expressed
the opinion that these specimens — because of their grit,
sand, and clay temper designs and notched lips — more
closely resembled Swift Creek pottery in Georgia than the
geographically closer Pickwick Complicated Stamped of
northern Alabama. This was also my reaction when I saw
them in the early 1940s (Griffin 1946:71) and still is.
Rocker-stamped
The presence of rocker-stamped pottery in the Greber
collection or in the earlier collections from the Edwin
Harness site is not easily explained. The Hopewell Zoned
Plain Rocker-stamped vertical compound vessel, recon-
structed from sherds obtained by Putman from his exca-
vations, might be regarded as having been deposited as a
whole vessel. The plain rocker-stamped sherd illustrated
by Mills ( 1907:Fig. 36) is on the right hand side of the four
glued sherds illustrated by Prufer ( 1 968: PI. 5a), and these
appear to be part of those in the Greber collection. The
latter specimens were relatively close together in the
northwest part of the southern sub-rectangular structure.
It is also possible that Mills’s ( 1 907 : Pig. 37) and Prufer’s
( 1 968: PI. 4b) are from the same vessel as Greber’s check-
stamped sherds (cat. no. 290C) from Feature 66, which is
also in the southern sub-rectangular structure. Notes
from the Griffin-Morgan study state that the check-
stamped sherds from the Mills Harness collection seem to
be from one footed or tetrapod vessel.
Conclusions
This brief discussion and survey of the distribution of
stamped wares, particularly simple-stamped, indicates
the general area of possible southeastern connection for
Ohio Hopewell. At present, eastern Tennessee and south-
western North Carolina are the more probable loci for the
derivation of simple-stamped and check-stamped vessels
into the central Ohio Valley or for the manufacturing
techniques which produced them. At some of the sites
mentioned there are pottery types and other artifacts
from Ohio. Some part of this interarea diffusion probably
represents the activity of individuals from southern Ohio
who participated in the acquisition of mica from North
Carolina and in the acquisition of marine shell and other
items from the Florida Gulf Coast. A more intensive ef-
fort should be made to identify other sites in northeastern
Tennessee, along the headwaters of the tributaries of the
Tennessee, and in eastern Kentucky and West Virginia,
which have occupations of the Middle Woodland peri-
ods. Whether this proposed interarea traffic took place
along streams or by trails is not definitely known. Proba-
bly both were used.
The temporal span of both Ohio Hopewell and Middle
Woodland sites in the eastern Tennessee area is not too
well known, and it is difficult or impossible to be precise
in terms of calendar or Carbon- 14 years about when these
contacts took place. Given the probable time span of
each, of some 400 years or more, the known amount of
reasonably identifiable trade goods is not very great, nor
can we yet be very specific about which Ohio sites were
prime movers of these goods. Such may come with a re-
finement of identification techniques of the clays and
tempering material.
Appendix 4. 1
Description and Sherd Count
Harness Mound Pottery — Season 1976
Field Cat. No.
1 Surface
Small McGraw Cordmarked specimen with cords
closely spaced and smoothed. Thickness 4.5 mm. I
2B, 2D, 2E, 2J, N535 E492.5 Disturbed area
2B Small vertical rim sherd of McGraw Plain with nar-
rowed flattened lip 3 mm wide. Slight protrusion on
outer upper rim folded or smoothed down 3 mm on
that surface. Consequently that area has a thickness
of 5 mm as does the lower rim (Fig. 4.2g). 1
Small McGraw Cordmarked body sherds one of
which is 6 mm and the other 4 mm thick. 2
2D McGraw Cordmarked body sherds composed of 3
fragments which fit together and vary in thickness
from 5 to 12 mm (Fig. 4.2b). I
2E McGraw Plain, probably a rim section, 5 mm thick. 1
An upper shoulder area section of Chillicothe Plain
Rocker-stamp which is placed below a shallow hor-
izontal line 3 mm wide which delimits the smoothed
horizontal rim band from the decorated area on the
body. The Rocker-stamp impressions are convex to
the right. They are arranged in 2 visible horizontal
rows each 1 .9 cm high. Both the inner and outer sur-
50
JAMES B. GRIFFIN
No. 39
face have been smoothed. On the outer surface this
took place after the design application. Both sur-
faces are black, possibly from having been fired in a
reducing atmosphere. Lower rim 7 mm and body
from 7 to 5.5 mm thick (Fig. 4.1c). 1
2J McGraw Cordmarked body sherd with some large
tempering particles 3.5 mm thick extending from the
inner to outer surfaces. Thickness 4 to 5 mm. 1
1 1C, N535 E495, Feature 4
Small sherdlets of McGraw Cordmarked. Perhaps
from the same vessel. Thickness 4.5 mm. 4
20A, N537.5 E492.5, Feature 13, (in situ) I.S.
21 fragments of the side wall of a McGraw Cord-
marked vessel which was in 28 fragments when re-
ceived. Thickness 3 to 5 mm (Fig. 4.4b). 1
McGraw Cordmarked sherd with same field number
but perhaps a different vessel or near base of vessel
listed above. Thickness 7 mm. 1
24B, 24C, 24D, 24E, 24G, 24H, N535 E492.5, Feature 89, I S.
24B Outer wall fragment McGraw Cordmarked. 1
Chillicothe Plain Rocker Stamped carelessly exe-
cuted on body fragment with tan outer surface.
Probably not from same vessel as 2E. Thickness 5.5
mm (Fig. 4. 11). I
24C McGraw Cordmarked. Thickness 3.5 mm. 1
Chillicothe Plain Rocker Stamped with same execu-
tion and size of rocker stamping as 2E but only 4.5
mm thick. Same dark brown to black surfaces but
probably not from same vessel. Fine grit temper
(Fig. 4.1e). I
24D McGraw Cordmarked 5 mm thick. 1
24E McGraw Cordmarked with coarse grit temper of
whitish crushed temper which is not limestone. Very
friable and 5.5 mm thick. I
Well-made Hopewell rim. The fine incised cross-
hatched upper band is 1.6 cm high. The horizontal
row of hemiconical punctates below the cross-
hatching are 5 to 9 mm long and 4 mm high. The
lower rim area is well smoothed. Inner and outer sur-
faces are black. There is a slight camber to the rim
caused by a shallow channel on the upper inner rim.
The lip is rounded and 4 to 5 mm wide while both the
upper and lower rim are 7 mm thick. This rim could
be from the same vessel as 2E because color, paste,
and thickness are similar (Fig. 4. la). 1
24G Chillicothe Plain Rocker Stamped with light tan
outer surface. Swing of rocker is 2.3 cm high. Thick-
ness 5 mm (Fig. 4. 1 k). 1
24H Chillicothe Plain Rocker Stamped with darker tan
outer surface. Swing of rocker is 2.3 cm high. Thick-
ness 4 mm. Does not appear to be from same vessel
as 24G and neither are from same vessel as 2E (Fig.
4- lj). 1
3 IB, N535 E495 Disturbed
McGraw Cordmarked sherdlets 3 mm thick. 1
35B, N535 E492.5, Disturbed
Chillicothe Plain Rocker Stamped. Too small to
measure the rocker swing. The lines are markedly
narrower than those of 2E and it is 5 mm thick (Fig.
4. lh). I
37A, N535 E492.5, PM 16, I.S.
McGraw Cordmarked upper body and lower part of
rim which may have been smoothed. It is 6 mm thick
and has relatively fine grit temper (Fig. 4.2c). 1
40 A, N535 E492.5, PM 14, I.S.
40A Hopewell rim section with same features as 24G and
is almost certainly from same vessel. The hemiconi-
cal punctates are more closely spaced but otherwise
appearance of all visible features is almost identical
(Fig. 4. lb). I
44C, N535 E492.5, Feature 19, (pit) I.S.
McGraw Cordmarked 4 mm thick. 1
Three sherdlets probably McGraw Cordmarked,
one of which is 5 mm thick. These probably from the
same vessel. 1
McGraw Cordmarked 5 mm thick. 1
66A, N548.5 E507.5, Feature 23
Two McGraw Cordmarked sherdlets 4.5 mm thick. I
McGraw Cordmarked sherds 4.5 mm thick. I
103B, 103C, 103D, N532.5 E495 Redeposited-disturbed
103B Three McGraw Cordmarked sherds 4.5 mm thick
probably from same vessel. I
McGraw Cordmarked lower rim and upper body 5
mm thick. 1
103C McGraw Cordmarked sherd. 1
103D McGraw Plain sherd 7 mm thick. 1
McGraw Plain sherd 4 mm thick. 1
103E Two Chillicothe Plain Rocker Stamped sherds with
1.8 cm high vertical swing, of very similar color and
finish as 24E but are 4.5 to 5 mm thick. They could
be from same vessel as 24G but from a different sec-
tion of the body (Fig. 4. lf,g). 1
One fine paste Hopewell sherd with horizontal (?)
incised line 2 mm wide. Sherd is 4 mm thick (Fig.
4. li). 1
104 Bulldozed area N520 E500
Small McGraw Cordmarked sherds 5 mm thick.
Could be from same vessel but impossible to say for
certain. 4
107 Bulldozed area
McGraw Cordmarked with well-smoothed exterior
5.5 mm thick. I
123 Surface
McGraw Cordmarked 7 mm thick.
1
1983
EDWIN HARNESS MOUND
51
McGraw Plain rim with flattened lip 6.5 mm wide
and slight outer slope. Whitish grit temper. Rim is 7
mm thick (Fig. 4.2h). I
138A, N522.5 E495 Redeposited-disturbed
Small lower rim and upper body sherd 7 mm thick.
Is from same vessel as 23 1 B and is glued to that rim
(Fig. 4.20. 1
138C, N522.5 E495 Redeposited-disturbed
Probably Turner Simple Stamped A, 7 mm thick
(Fig. 4.3e). 1
155, N557.5 E495 in mound floor
McGraw Cordmarked 5 mm thick. 1
183A, 183B, 183C, 183F,and 183G, N525 E495, Feature 30(pit)
Some 24 sherdlets of 183B, six of 183A, seven of
183C, six of 183E, seventeen of 183F, and three of
183G are McGraw Cordmarked. While it is not cer-
tain that all of these are from one vessel, the 183A, a
183B, and a 183C fit together, as do a 183C and 183E
sherd. If there are at least two vessels, the larger
sherds belong to a vessel with 8 mm side walls and
another vessel is represented by side walls 4 mm
thick. 2
One baked clay fragment of 1 83C does not appear to
be from a vessel, or figurine, and is unidentified.
191 A, N532.5 E495, PM 84, I.S.
Inner section of McGraw Cordmarked or Plain,
probably the former. I
217, Provenience lost
This small sherd is either McGraw Plain or Cord-
marked but it cannot be identified with any cer-
tainty. 1
Harness Mound Pottery — Season 1977
219A, Surface
Turner Simple Stamped in Prufer’s( 1968) terminol-
ogy. This thin, hard, fine grit tempered sherd is cer-
tainly not the same vessel as 308A. Neither have sand
or micaceous sand temper. The inclusion of either in
a “Southeastern Series” would be a mistake. Thick-
ness 4.5 mm (Fig. 4.31 and m). 1
219E, Surface Backhoe Trench 1 backdirt pile
Fairly large smoothed over McGraw Cordmarked 6
to 7 mm thick. 1
219U, Surface
Two small McGraw Cordmarked sherds glued to-
gether. They are 4 mm thick. 1
219V, Surface
A McGraw Cordmarked sherd 5.5 mm thick. 1
A McGraw Plain small rim with a narrowed and
rounded lip 4 mm wide while the rim is 7.5 mm thick
(Fig. 4.2i). 1
22 1G, N530 E507.5, Bulldozed area
One sherd 8.5 mm thick called here McGraw Plain. 1
225B, N540 E507.5, Redeposited
Small Hopewell rim with reddish tan interior and
exterior color. The lip, narrowed and rounded,
slopes inward and is 3 mm wide. The incised cross
hatching is widely spaced and is 7.5 mm high. The
hemiconical punctates are 3 mm long and 3 mm
wide. They are spaced 3.5 to 4 mm apart. The upper
and lower rim thickness is 4.5 to 5 mm (Fig. 4. Id). I
229C, N550 E507.5, Redeposited
Lower rim and upper body sherd of McGraw Plain.
It is 8 mm thick and grit tempered (Fig. 4.2j). 1
230N, N522.5 E497.5
Two small McGraw Cordmarked sherds glued to-
gether. They are 4 mm thick. 1
231 A, 23 IB, 23 IF, 237C, 237E, N522.5 E490, Disturbed area
231 A Two small Turner Simple Stamped A sherds with
limestone temper. They are probably part of 308A.
The stamp depressions are 2 to 3 mm wide and the
lands are 1 to 2 mm wide (Fig. 4.3g, h). 1
23 1 B A rim sherd of unidentified type with a plain upper
body and a horizontally brushed, wiped, or simple
stamped outer rim. The lip is narrowed and rounded
and is 4.5 mm wide. The rim and body are 8.5 to 9.5
mm thick. It is grit tempered. 1
23 IF Turner Simple Stamped A body sherd with lime-
stone temper. Probably same vessel as 308A. Thick-
ness 7.5 mm (Fig. 4.3i). 1
237C Rim sherd and upper body of Turner Simple
Stamped A. Is very probably same vessel as 308A.
The rim has a slight flare or outer slope and is 7 mm
thick. The lip slopes to the interior and has shallow
depressions caused by thumb (?) impressions while
the clay was still soft. This vessel is a real stranger in
the Harness ceramic assemblage because of the
temper (Fig. 4.3b, j). 1
Turner Simple Stamped A body sherd from same
vessel. 1
237E Turner Simple Stamped A body sherd from same
vessel (Fig. 4.3k). 1
264E, N530 E485
McGraw Cordmarked 6.5 mm thick. 1
2710, N552 E500, PM 162, I.S.
A large McGraw Plain body of 4 glued fragments 5
mm thick (Fig. 4.4a). I
290C, 290G, N525 E502.5, Feature 66
290C Has 8 fragments of a thin checked-stamped vessel.
These sherds are thin, 3 mm, grit tempered and are
not trade material from the Southeast as far as I can
see. Both inner and outer surfaces are black and
smoothed (Fig. 4. lm-g). 1
Has 1 lower rim section of McGraw Plain. It is well
smoothed on outer and inner surfaces, which are
black. The sherd is 9 mm thick (Fig. 4.2e).
1
52
JAMES B. GRIFFIN
No. 39
290G Has 4 sherds of McGraw Cordmarked. They do not
seem to be from the same vessel. 4
291 A, N530 E502.5 on top of PM 160
A small sherd of McGraw Plain 4 mm thick. 1
295G, N530 E502.5 on top of PM 160
Two sherds of probably same McGraw Cordmarked
vessel. 1
Two other unidentified fragments. 2
295L, N530 E502.5, PM 160, I.S.
Two sherds of which one has two glued pieces. The
glued sherds are from a lower rim section of perhaps
McGraw Plain and are 8.5 mm thick. The second is a
small fragment of probably McGraw Cordmarked
and is 5 mm thick. 2
297G, N535 E509, partially disturbed
One McGraw Plain 5 mm thick body sherd. 1
One McGraw Cordmarked 4.5 mm thick. 1
300A, 300B, N525 E509, Redeposited
Both of the two sherds are very small. They may be
McGraw Plain or Cordmarked and are 3.5 mm
thick. 2
308A, N524 E485, PM 270, I.S.
Large body sherd of Turner Simple Stamped A
glued together from 9 sherds, and one other sherd.
This vessel has limestone temper of medium size
crushed fragments. There are fragments of this ves-
sel from other localities, and the vessel seems to be
the only limestone tempered one in this collection.
The paddle depressions are about 3 mm wide and the
bands are 2 to 1 mm wide. Thickness is 6 to 7 mm
(Fig. 4.3a, 0. 1
308C, N524 E485 PM 268
4 body sherds McGraw Cordmarked glued together,
reddish to tan in color, may belong to one vessel, but
uncertain. Thickness 6 to 9 mm. 1
309A, N524 E492.5 burned area on mound floor, partly dis-
turbed.
A lower rim (McGraw Plain) and upper body
(McGraw Cordmarked) 7 mm thick. Well fired tan
in color (Fig. 4. 2d). I
Two body sherds of Turner Simple Stamped A from
same limestone tempered vessel as 308A, 8 mm thick
(Fig. 4.3c, d). 1
504B, N527.5 E495, PM 410, I.S.
A McGraw Cordmarked sherd 4.5 mm thick. 1
700B, N522.5 E495, Redeposited
McGraw Cordmarked sherd 5 mm thick. 1
References
Adams, William R.
1949 Archaeological notes on Posey County, Indiana. In-
diana Historical Bureau, Indianapolis.
Butler, Brian M.
1979 Hopewellian contacts in southern Middle Tennessee.
In Hopewell archaeology: the Chillicothe conference,
edited by D. S. Brose and N. Greber, pp. 150-156.
Kent State University Press, Kent, Ohio.
Chapman, Jefferson
1973 The Ice House Bottom site 40Mr23. Report of Inves-
tigations, No. 13, Department of Anthropology, Uni-
versity of Tennessee, Knoxville.
Chapman, J., and B. C. Keel
1979 Candy Creek-Connestee components in eastern Ten-
nessee and western North Carolina and their relation-
ship with Adena-Hopewell. In Hopewell archaeology :
the Chillicothe conference, edited by D. S. Brose and
N. Greber, pp. 157-161. Kent State University Press,
Kent, Ohio.
DePratter, C.
1979 Ceramics. In The anthropology of St. Catherines Is-
land: 2. The Refuge-Deptford mortuary complex,
edited by D. H. Thomas and C. S. Larsen. Anthropo-
logical Papers of the American Museum of Natural
History 56(1): 109- 132, New York.
Faulkner, Charles H., and J. B. Graham
1965 Excavations in the Nickajack Reservoir: Season 1.
Tennessee Archaeological Society, Miscellaneous
Paper, No. 7.
1966 Highway salvage in the Nickajack Reservoir. Tennes-
see Department of Highways Contract No. 0402. De-
partment of Anthropology, University of Tennessee,
Knoxville.
Faulkner, Charles H. (editor)
1968 Archaeological investigations in the Tims Ford Res-
ervoir, Tennessee, 1966. National Park Service Con-
tract 14-10-0131-1631. Department of Anthropology,
University of Tennessee, Knoxville.
Fowler, Melvin L.
1957 Rutherford Mound, Hardin County, Illinois. Illinois
State Museum Scientific Papers 7( 1 ): 2— 43, Spring-
field.
Gleeson, Paul F. (editor)
1970 Archaeological investigations in the Tellico Reser-
voir, Interim Report, 1969. Report of Investigation,
No. 9. Department of Anthropology, University of
Tennessee, Knoxville.
Griffin, James B.
1938 The ceramic remains from the Norris Basin, Tennes-
see. In An archaeological survey of the Norris Basin in
eastern Tennessee, edited by W. S. Webb. Bureau of
American Ethnology, Bulletin 1 18:253-359, Wash-
ington.
1 945 The ceramic affiliations of the Ohio Valley Adena cul-
ture. In The Adena people, edited by W. S. Webb and
C. E. Snow. University of Kentucky, Reports in An-
thropology and Archaeology 6:220-246.
1946 Cultural change and continuity in eastern United
States archaeology. In Man in northeastern North
America, edited by F. Johnson. Papers of the Robert
S. Peabody Foundation for Archaeology 3:37-95,
Andover.
Haag, William G.
1940 A description of the Wright site pottery. In The
1983
EDWIN HARNESS MOUND
53
Wright mounds, sites 6 and 7, Montgomery County,
Kentucky, edited by W. S. Webb. University of Ken-
tucky Reports in Anthropology and Archaeology
5( 1 ):75— 82, Lexington.
1942a A description and analysis of the Pickwick pottery. In
An archaeological survey of Pickwick Basin in the ad-
jacent portions of the states of Alabama, Mississippi,
and Tennessee, edited by W. S. Webb and D. L. De-
Jarnette. Bureau of American Ethnology, Bulletin
129:509-526.
1942b The pottery from the C and O mounds at Paintsville.
In The C and O mounds at Paintsville, sites Jo 2 and
Jo 9, Johnson County, Kentucky, edited by W. S.
Webb. University of Kentucky, Reports in Anthro-
pology and Archaeology 5(4):34 1—349, Lexington.
Heimlich, Marion D.
1952 Guntersville Basin pottery. Geological Survey of Ala-
bama, Museum Paper No. 32, Tuscaloosa.
Jefferies, Richard W.
1976 The Tunacunnhee site: evidence of Hopewell interac-
tion in northwest Georgia. Anthropological Papers of
the University of Georgia , No. 1, Athens.
Keel, Bennie C.
1976 Cherokee archaeology: a study of the Appalachian
summit. University of Tennessee Press, Knoxville.
Kellar, James H.
1973 An introduction to the prehistory of Indiana. Indiana
Historical Society, Indianapolis.
1979 The Mann Site and “Hopewell” in the lower Wabash-
Ohio valley. In Hopewell archaeology: the Chillicothe
conference , edited by D. S. Brose and N. Greber, pp.
100-107. Kent State University Press, Kent, Ohio.
McCollough, Major C. R., and Charles H. Faulkner (editors)
1973 Excavation of the Higgs and Doughty sites: 1-75 sal-
vage archaeology. Tennessee Archaeological Society,
Miscellaneous Paper. No. 12.
Mills, William C.
1907 Explorations of the Edwin Harness mound. Ohio Ar-
chaeological and Historical Quarterly 16:1 13-193.
Mortine, Wayne A., and Doug Randles
1978 The Martin mound: an extension of the Hopewell in-
teraction sphere into the Walhonding Valley of east-
ern Ohio. Occasional Papers in Muskingum Valley
Archaeology 10.
Prufer, Olaf H.
1967 The Scioto Valley archaeological survey. In Studies in
Ohio archaeology , edited by O. H. Prufer and D. H.
McKenzie, pp. 267-328. The Press of Western Re-
serve University, Cleveland; rev. ed., Kent State Uni-
versity Press, 1975.
1 968 Ohio Hopewell ceramics: an analysis of the extant col-
lections. Anthropological Papers No. 33, Museum of
Anthropology, University of Michigan, Ann Arbor.
Prufer, Olaf H., D. H. McKenzie, O. Pi-Sunyer, H. C. Cutler,
R. A. Yarnell, P. W. Parmalee, and D. H. Stansbery
1965 The McGraw site: a study in Hopewellian dynamics.
Cleveland Museum of Natural History, Scientific
Publications (n.s.), 4(1).
Webb, W. S.
1938 An archaeological survey of the Norris Basin in east-
ern Tennessee. Bureau of American Ethnology, Bul-
letin 1 18, Washington.
Williams, Stephen (editor)
1968 The Waring papers: the collected works of Antonio J.
Waring, Jr. Harvard University, Peabody Museum of
American Archaeology and Ethnology, Papers 58,
Cambridge.
Willoughby, Charles C., and Earnest A. Hooton
1922 The Turner group of earthworks, Hamilton County,
Ohio. Harvard University, Peabody Museum of
American Archaeology and Ethnology, Papers 8(3),
Cambridge.
Winters, Howard D.
1963 An archaeological survey of the Wabash Valley in Il-
linois. Illinois State Museum, Report of Investiga-
tions 10, Springfield.
5 PLANT REMAINS
TRISTINE LEE SMART and RICHARD 1. FORD
The Edwin Harness Mound, located in the Scioto
River valley near Chillicothe, Ohio, has long been recog-
nized as a major Hopewell burial mound. While several
excavations were conducted at the site in the past, many
questions regarding the mound stratigraphy and sub-
mound features, structures, and activity areas remained
unanswered. Therefore, when unexcavated portions of
this mound were recently threatened by cultivation, sal-
vage excavations were conducted under the direction of
Dr. N’omi Greber of the Cleveland Museum of Natural
History. During the excavations, plant remains were col-
lected and submitted to the Ethnobotanical Laboratory
of the University of Michigan Museum of Anthropology
for identification and interpretation. These samples were
returned to the Cleveland Museum of Natural History af-
ter analysis for curation and storage.
Nature of the Samples
Plant remains recovered from the Edwin Harness
Mound included primarily charcoal samples and flota-
tion samples from post molds and features. In addition, 2
pieces of fabric, 1 seed, and a modern corn cob ( Zea mays)
were collected during excavation.
Flotation was conducted in the field and at the Cleve-
land Museum of Natural History using a SMAP-type flo-
tation system (Watson 1976). The light fraction was col-
lected in 4 mm, 2 mm, and .5 mm screens.
Analytical Techniques
The samples of plant material were examined micro-
scopically under magnifications ranging from 10 to 30X.
Carbonized seeds were separated from the samples and
identifications were attempted with the aid of seed manu-
als (Martin and Barkley 1961; Delorit 1970) and the com-
parative collections of the Ethnobotanical Laboratory.
Charcoal was identified by examining a transverse section
microscopically with the assistance of wood manuals
(Brown 1928; Panshinand deZeeuw 1970) and compara-
tive wood samples. A subsample of 20 pieces of charcoal
was selected for identification from samples containing
large amounts of charcoal.
Because the size and shape of a piece of charcoal affects
its identifiability, these factors influence the selection of
charcoal pieces for identification. Therefore, the charcoal
selected cannot be considered an unbiased subsample.
However, care was taken to select charcoal pieces of many
shapes and sizes, so the charcoal subsamples should be
fairly representative of the charcoal from each sample as a
whole.
Charcoal from Posts
The submound deposits from the Edwin Harness
Mound included several adjacent structures outlined by
post molds. Charcoal was recovered from a number of
these post molds (Table 5. 1 ).
Hickory ( Carya) was the most common timber used for
construction. Out of a total of 33 post samples examined,
25 (76%) contained hickory. All of the identified hickory
pieces had annual rings which showed an even, concentric
growth pattern. This suggests that these trees were grow-
ing in a location that was not stressful at the time the rings
were formed. All were saplings or young trees rather than
limbs or split sections from larger, more mature trees.
In 4 of the post samples, a few pieces of non-hickory
charcoal types were found in addition to hickory. These
included oak ( Quercus', Post 1 42), the more specific white
oak group (Posts 151 and 155), and maple (Acer, Post
146). These may have been from wooden wedges used to
tighten the post in place, or alternatively contamination
from the general fill or the charred remains of trees which
once grew on the site. A woven bast fiber textile was
found at the base of Post 1 3 1 and may have served a sim-
ilar purpose.
Eight post samples did not contain hickory charcoal.
The white oak group was present in 6 of these samples
( Posts 85, 122, 132, 1 98, 2 1 3), 1 post was elm ( Ulmus ; Post
138), and another was an unidentifiable diffuse porous
wood type (Post 128).
The posts from the circular structure at the south end of
the mound are interesting. The 2 samples from Post 216
actually contained charcoal of 5 tree types: hickory,
chestnut ( Castanea ), honey locust ( Gleditsia ), walnut
( Juglans ), and pine ( Pinas). This suggests that this feature
was not a post. Two of the remaining 3 posts that were
identified were white oak group, and it might be signifi-
cant that they were located opposite each other on the east
and west sides of the structure.
Charcoal from Features
Plant remains were recovered from a number of differ-
ent types of features at the Edwin Harness Mound. These
included mound loading (Feature 43); a large ring of cob-
bles which encircled the other submound deposits (Fea-
ture 1 ); a burned area at the center of the middle structure
TABLE 5.1
Charred Posts from the Edwin Harness Mound Floor
(Percentages based on a total of 20 identified pieces)
C.
■3
-C>
3
3
u
Possible fence around perimeter
Circular structure, south end
Middle rectangular structure
Northern rectangular structure
Eastern structure
116
F6
X
1 19
F4
X
125
F9
X
140
F40
X
151
F76
X
85
192A
142
230M
X
198
CCD61
216
CCD64
50
F99
65
127
F31
X
128
F28
131**
F27
X
F32
X
157
F85
X
230
¥91
X
122
F8
126
F19
X
132
F48
134
F39
X
136
F26
X
138
F42
141
F65
X
146
F66
X X
152
F75
X
154
F74
X
155
F77
X
162
F84
X
163
F82
X
213
F93
F95
121
F5
X
215
F98
X
O
"O
o
20 30
10 10
c
5,
3
C
'£
x
x
x
x
10
X
X
X
X
X
.§
3
£
a
§
X*
X
x = present
*charcoal minute and fragile
**woven blast fiber fabric present
TABLE 5.2
Charcoal from the Edwin Harness Mound Features
(Percentages based on a total of 20 identified pieces)
*3
§•
§■
©
s*
3
o
O
bo
*©
3
bo
3
3
"3
3
3
“3
-3
£
»)
3
3
-Si
§■
£
©
•Si
«3
-3
c.
c«
C
2
-3
-3
3
-3
3
3
C
'*
3
o
-3
.2
.3
*3
£
C
J2
.3
s.
3
O
3
■4*
3
O
3
o
s
*3
©
C.
3
£
O
ao
C3
a>
Ob
C
3
3
>3
$
<
u
u
Urn
U-
o
—>
a.
o
a
o
1
F10A
X
F23
X
17
F70AA
X
X
X
X
33
F25
67
33
F36
100
F41
X
F43
100
F44
100
F45
X
F57
40
60
F70
50
10
40
CCD19A
5
5
15
75
36
FI A
25
75
43
F29
100
X
45
F53
100
F60
10
30
60
X
F64A
30
10
60
F64B
80
20
X
F64C
30
5
5
60
F64D
30
30
40
F69
5
50
15
30
X
F69B
67
33
F69C
60
40
X
46
F47
20
40
40
F62
45
5
15
35
49
F46*
90
10
F49*
30
70
50
F58
100
52
F56*
100
53
F61
100
F63
70
30
54
F79
X
60
F7 1 A
75
5
5
5
5
5
62
F83
30
10
60
65
F35
100
**
F20
-
X
x = present
♦directly over feature in loading
**N560 E507.5, mound loading
1983
EDWIN HARNESS MOUND
57
on the mound floor (Feature 36); a depression in the main
mound floor (Feature 54); strata under the main mound
floor (Features 33, 50, and 65); a bundle burial (Feature
60); a large pit located outside of the submound structures
(Feature 1 7); a small oval-shaped basin found just north
of the submound structures (Feature 62); and various
strata located outside of the submound structures. The
latter included burned material on clay floors (Features
45, 46, and 53); burned soil (Feature 49); and mica pieces
in a dark stained area (Feature 52).
Charcoal identifications from botanical samples col-
lected from features are presented in Table 5.2. Ten tree
types were represented, including such types as beech
( Fagus ) and maple, which are found today in riverbottom
and lower slope communities, as well as such trees as
chestnut and pine, which are primarily found on the
upper slopes (Gordon 1 969). Additional tree types identi-
fied included hickory, ash ( Fraxinus ), honey locust, wal-
nut, red oak group, and white oak group.
As with the post samples, hickory and white oak group
were the most common charcoal types in the feature sam-
ples. Of the 35 flotation samples and 1 charcoal sample
from features, 26 (72%) contained hickory and 20 (56%)
contained white oak group charcoal. This suggests that
some of the charcoal incorporated in the feature deposits
could have come from the remnants of burned hickory
and white oak group construction timbers. However, the
charcoal in features could also be the remnants of wood
used as fuel. Hickory and white oak group wood both
burn with a hot, clean flame and may have been selected
for use as fuel because of their burning properties.
Interestingly, the burned area in the middle structure
on the main mound floor (Feature 36) contained 75% pine
charcoal. Pine has very different heating properties than
oak and hickory, producing a hot fire more quickly than
the hardwoods. The bundle burial deposit (Feature 60)
contained the greatest diversity of charcoal types includ-
ing hickory, chestnut, beech, walnut, white oak group,
and red oak group.
Seeds from the Edwin Harness Mound
Carbonized seeds were recovered from 10 of the feature
flotation samples, from 2 of the post samples (1 seed was
recovered during excavation), and from a charcoal sam-
ple used for C-14 dating (Table 5.3).
Corn kernels were the only type of domesticated plant
material recovered from the site. Unambiguous kernel
fragments were recovered from Feature 45 and Feature
60, the bundle burial. The identification of grains from
Feature 65 is more questionable because of their fragmen-
tary state. One uncarbonized, 12-row cob fragment of a
modern variety was found in Backhoe Trench 4 during
excavation. The presence of corn with the bundle burial
suggests the use of corn as a mortuary offering and hints
at the ritual significance of corn in Hopewell culture.
Carbonized seeds from wild plants recovered at the site
included goosefoot (Chenop odium), knotweed ( Polygo-
num), spurge ( Euphorbia ), and grass (Gramineae). The
seeds of the goosefoot and knotweed are difficult to inter-
pret. These ruderal plants grow prolifically on disturbed
or abandoned sites, and their seeds could be accidental
inclusions resulting from unintentional dispersal of seeds
from plants growing nearby. Alternatively, these seeds
could have been carbonized through the use of dried
weeds as kindling. However, both goosefoot and knot-
weed seeds are edible. Large quantities of these seeds were
recovered from Middle Woodland deposits at the Scovill
site in Illinois, indicating that they were used as food in
that area (Munson et al. 1971). In most of the samples
from the Edwin Harness Mound the low counts of goose-
foot and knotweed seeds suggest that they may well have
been accidental inclusions in the deposits. However, the
moderate number of goosefoot and knotweed seeds along
with corn in the bundle burial deposit hint at the eco-
nomic use of these seeds. Spurge seeds were also recov-
ered from the Feature 60 burial as well as from Feature
17, although these seeds are not known to be edible. The
presence of carbonized grass seeds and culms could also
have been due to accidental inclusion in the archaeologi-
cal record. Alternatively, these plant remains could have
come from thatching on the submound structures.
Hundreds of an unknown, egg-shaped seed type were
recovered from Features 17 and 60, and much smaller
numbers of these seeds were found in Features 33 and 65.
A sample of 25 of these seeds from Feature 60 had an
average size of 1 .7 mm X 1 .2 mm. In addition, other uni-
dentified seeds were found in many of the samples.
No carbonized nutshell fragments were recovered from
the site, even though hickory, oak, and walnut charcoal
was present in the post and feature samples. This is quite
unlike the reported plant remains from several other Ohio
Hopewell sites (Ford 1979:235) and probably reflects the
special purpose of the Edwin Harness Mound. The only
identified edible tree products from this site were 2 plum
pits ( Prunus americana).
Uncarbonized, modern weed seeds were also recovered
in the archaeological deposits from the mound. These in-
cluded not only modern goosefoot, knotweed, and grass
seeds, but also the grass genus Digitaria , purslane ( Por-
tulaca), campion ( Silene ), and blueberry ( Vaccinium ),
which did not occur archaeologically.
Other items recovered from the botanical samples were
bone and snail shells (Table 5.3). A piece of bone from
Feature 60 was identified as possibly human. The identifi-
able snail shells from the samples were Hawaiia minus-
cula (Amy Shraden Van Devender, Museum of Zoology,
Univ. of Michigan, personal communication, 1978). This
species prefers a moist, floodplain habitat, but it could be
found elsewhere as well (LaRocque 1970:639).
58
TRISTINE LEE SMART and RICHARD I. FORD
No. 39
TABLE 5.3
Carbonized Seeds and Faunal Remains from the Edwin Harness Mound Botanical Samples
(counts include half to complete seeds)
Sample
Provenience Number
Chenopodium
goosefoot
Euphorbia
spurge
Gramineae
grass
Carbonized Seeds
Prunus
Polygonum americana
knotweed plum
Zea
mays
( kernel )
corn
Unknown
(egg
shaped)
Fauna
Unident. Bone Snail
Post 141
F65
1
Post 142
230M
1 frag
Fea. 17
F70AA
4+
3
1
8
155+
34+ x
Fea. 33
CCDI9A
5+
F25
1
F36
l+(cf)
1
F70
22+
7(cf)
3
56 xx
Fea. 45
F64C
I
3
1 frag
F69C
3 culms
X
Fea. 46
F62
1
X
Fea. 52
F56
1
Fea. 60
F71A
4+
8
16+
1 frag+*
131
44+ x x
Fea. 65
F35
1 frag
+(cf)
10
8+
+ = additional seed fragment(s) not included in count
x = present
* one Zea mays kernel fragment plus many small cf Zea mays kernel fragments
Conclusion
The plant remains from the Edwin Harness Mound
were very specialized as one might expect in a mortuary
site. Most of the construction wood was hickory. The
charcoal from the feature deposits was primarily hickory
and white oak group, although other species were present
in low frequencies. One sample from a burned area at the
center of the middle structure on the mound floor con-
tained mostly pine charcoal. The charcoal types recov-
ered suggest that the aboreal environment was similar to
the contact forest; however, no nuts from this forest were
found at the site. Fragments of corn kernels and possibly
goosefoot and knotweed seeds found in association with a
bundle burial may represent mortuary offerings. Corn
kernel fragments were also present in one or possibly two
other deposits. Carbonized seeds from wild plants recov-
ered from other features were possibly refuse from a meal
or were simply the result of accidental dispersal into a
hearth or the fire that consumed the structures.
References
Brown, H. P.
1928 Atlas of the commercial woods of the United States.
The Bulletin of the New York State College of For-
estry at Syracuse University 1(4).
Delorit, R. J.
1970 Illustrated taxonomy manual of weed seeds. Agron-
omy Publications, River Falls, Wis.
Ford, Richard I.
1979 Gathering and gardening: trends and consequences of
Hopewell subsistence strategies. In Hopewell archae-
ology: the Chillicothe conference , edited by D. S.
Brose and N. Greber, pp. 234-238. Kent State Univer-
sity Press, Kent, Ohio.
Gordon, Robert B.
1969 The natural vegetation of Ohio in pioneer days. Bul-
letin of the Ohio Biological Survey 3(2).
LaRocque, Aurble
1970 Pleistocene mollusca of Ohio. State of Ohio Depart-
ment of Natural Resources, Division of the Geologi-
cal Survey Bulletin 62, Part 4.
Martin, Alexander C., and William D. Barkley
1961 Seed identification manual. University of California
Press, Berkeley, Calif.
Munson, Patrick J., Paul W. Parmalee, and Richard A. Yamell
1971 Subsistence ecology of Scovill, a terminal Middle
Woodland village. American Antiquity 36(4):4 1 0—43 1 .
Panshin, A. J., and Carl de Zeeuw
1970 Textbook of wood technology (V ol. 1). McGraw-Hill
Book Company, New York.
Watson, Patty Jo
1976 In pursuit of prehistoric subsistence: a comparative
account of some contemporary flotation techniques.
Midcontinental Journal of Archaeology 1(1 ):77— 1 00.
6 VERTEBRATE FAUNAL REMAINS
ORRIN C. SHANE III
During the 1976 and 1977 field seasons nearly 10,000
pieces of bone were recovered from the Edwin Harness
burial mound (33Ro22), Ross County, Ohio. Excava-
tions were directed by Dr. N’omi Greber of the Cleveland
Museum of Natural History as part of a project to salvage
information from the site prior to its final destruction by
agricultural activity.
The Harness Mound is the major earthen mortuary
structure associated with the Harness (Liberty) Earth-
work Complex, and represents a locus of Hopewellian
mortuary-ceremonial activity dating from the first few
centuries of the Christian era. The mound is located on
the east side of the Scioto River Valley, in an area charac-
terized in prehistoric times by a mixture of deciduous
forest and open grassland vegetational communities. The
site lies with the Carolinian Biotic Province (Dice 1943),
at the western margin of the unglaciated Allegheny
Plateau.
Methods
A variety of recovery techniques were employed in the
mound excavation, including dry screening through 'A
inch hardware cloth, water separation, and flotation.
Consequently, the rate of recovery of bone and other cul-
tural material was high, yielding large numbers of small,
often unidentifiable bone fragments. Therefore, each
bone piece was examined and initially placed into one of
three categories, either as bone identifiable below the
class level, bone identifiable to class only, or bone uniden-
tifiable to class. The criterion for identifiability below the
class level was the presence of the major portion of an
articulatory surface; bone was identified to class on the
basis of gross structure.
For that portion of the assemblage with bones bearing
an articulatory surface, identifications were made after
comparison with skeletal reference collections housed in
the James Ford Bell Museum of Natural History, Univer-
sity of Minnesota, and the Science Museum of Minne-
sota. Scientific and common names of mammals follow
Hall and Kelson ( 1959), while the names of birds are from
the Checklist of North American Birds of the American
Ornithologists’ Union ( 1957). The scientific and common
names of fishes follow Trautman ( 1957).
The minimum number of individuals (MNI) repre-
sented by the identifiable bones was determined by simple
osteological count of right and left elements. Because of
the small size of the sample and the absence of bones in-
dicative of the age of individuals, this method of deter-
mining MNI is probably satisfactory.
Several excavation units and features yielded mammal
canine teeth drilled for suspension. While many of these
specimens were burned and fragmentary, some teeth were
sufficiently complete for identification. A minimum num-
ber of drilled canines was obtained from a count of distal,
medial, and proximal fragments. Perforation for suspen-
sion was near the base of the tooth root, and fracture gen-
erally occurred at the point of drilling. Therefore, distal
fragments were defined as including the tooth portion
from the tip of the crown to the point of fracture at the
perforation. Medial fragments lacked the tip of the tooth,
while proximal fragments were defined as the tooth por-
tion from the base of the root to the point of fracture at
the perforation. The minimum number of drilled canine
teeth was determined from counts of distal, medial and
proximal fragments plus whole canines.
Results
A total of 9,762 pieces of bone were recovered. Of these,
3,036 pieces, or 31. 10% of the total, were very small frag-
ments unidentifiable to class. Of the remaining 6,726
bones, 1 76 pieces were identifiable to the family, genus, or
species levels. Table 6. 1 shows the frequency of identified
and unidentified bones by class; the frequency of identi-
fied vertebrate taxa is presented in Table 6.2.
Sixteen complete perforated mammal canine teeth and
225 proximal, medial, and distal fragments represent at
least 171 specimens of drilled canines. All but 5 of these
objects were burned, and some were completely calcined.
Seventeen specimens could be identified to the genus or
species level as follows:
Raccoon 10 specimens
Canis sp. 4 specimens
Grey Fox 2 specimens
Bobcat I specimen
Feature 44, described in the field as a deposit of burned
materials and ash, yielded 115 fragments of drilled ca-
TABLE 6.1
Frequency of Identified and Unidentified
Bones by Class
Class
Identified
%
Unidentified
%
Total
%
Mammal
133
1.98
6,211
92.34
6,344
94.32
Bird
24
0.36
260
3.87
284
4.21
Reptile
2
0.03
51
0.76
53
0.79
Fish
17
0.25
28
0.42
45
0.67
Totals
176
2.62
6,550
97.38
6,726
100.00
60
ORRIN C. SHANE III
No. 39
TABLE 6.2
Frequency of Identified Vertebrate Remains from Edwin Harness Mound
Scientific Name
Common Name
No. of Bones
%
MNI
%
Odocoileus virginianus
Deer
35
19.89
2
6.25
Procyon lotor
Raccoon
11
6.25
6
18.75
Canis sp.
Canid
5
2.84
2
6.25
Urocyon cinereoargenteus
Grey Fox
1
0.57
1
3.13
Lynx rufus
Bobcat
2
1.14
1
3.13
Sylvilagus floridanus
Cottontail
1
0.57
1
3.13
* Tamias s that us
Chipmunk
28
15.91
1
3.13
Peromyscus maniculatus
Deer Mouse
2
1.14
1
3.13
Rodentia
Small Rodent
1
0.57
1
3.13
Homo
Human
47
26.70
3
9.37
Total Mammal
133
75.57
19
59.37
Meleagris gallopavo
Turkey
15
8.52
2
6.25
* Gallus gallus
Chicken
5
2.84
2
6.25
Colinus virginianus
Bob-white Quail
1
0.57
1
3.13
Buteo sp.
Hawk
1
0.57
1
3.13
Anas sp.
Duck
1
0.57
1
3.13
Passeriformes
Songbird
1
0.57
1
3.13
Total Bird
24
13.64
8
25.00
Colubridae
Snake
2
1.14
1
3.13
Total Reptile
2
1.14
1
3.13
Ictalurus sp.
Catfish
7
3.98
1
3.13
Micropterus sp.
Bass
6
3.41
1
3.13
Catostomidae
Suckers
4
2.27
2
6.25
Total Fish
17
9.66
4
12.50
Grand Totals
176
100.00
32
100.00
♦Intrusive
nines representing at least 99 specimens. Also associated
with these canines was a fragment of a cut and polished
bobcat (Lynx rufus) mandible. This specimen is the ante-
rior portion of the right mandible including the alveoli of
the canine and the adjacent two premolars. The mandible
is cut parallel to the plane of the dentition at the base of
the tooth roots, and grinding is evident along the cut sur-
faces. A hole was drilled through the mandible, below the
first premolar and posterior to the canine. This bobcat
mandible fragment and the associated perforated canines
may represent a necklace or other similar ornament.
Discussion
Chipmunk and domestic chicken are both clearly intru-
sive to the site. The chipmunk is represented by 28 bones
found articulated in situ; apparently the animal died after
burrowing into the mound. The occurrence of domestic
chicken is interesting, for it is likely that these bones rep-
resent meals of such earlier excavation crews as those led
by Putnam, Moorehead, and Mills. All of the chicken
bones were recovered from previously disturbed portions
of the mound, in and under backdirt from prior excava-
tions and well below the modern surface.
Apart from the intrusive modern domestic chicken, the
faunal assemblage is in no way unusual for a Woodland
context in the Carolinian Biotic Province. No exotic spe-
cies are present, and those animals which are represented
are to be expected. White-tailed deer, raccoon, turkey,
and fish account for approximately 45% of identifiable
bone and almost 50% of the individuals present. These
species were also among those most numerous in the envi-
ronmental zones around the site.
What may be most significant about this assemblage is
its very small size. If the sample is truly representative of
animal utilization on the Harness Mound floor, then it
would appear that animal foods d id not play a particular-
ly important role in Hopewellian mortuary-ceremonialism
1983
EDWIN HARNESS MOUND
61
as practiced at Harness. While animal parts may have been
used as ornaments, certainly there is no evidence for mass
offerings of animals, or for large-scale processing of foods
for use on the mound floor.
The small size of the assemblage would seem to pre-
clude any large habitation at the mound as mortuary ac-
tivities were carried out. Furthermore, the paucity of fau-
nal remains argues against the mortuary floor being the
site of accumulation and exchange of food resources. If
anything, animals appear to have been used on the mound
floor in much the same manner as at habitation sites, but
in far smaller quantity.
References
American Ornithologists’ Union
1957 Checklist of North American birds. 5th ed.
Dice, L. R.
1943 The biotic provinces of North America. University of
Michigan Press, Ann Arbor.
Hall, E. R., and K. R. Kelson
1959 The mammals of North America. Ronald Press, New
York.
Trautman, Milton B.
1957 The fishes of Ohio , with illustrated keys. Ohio State
University Press, Columbus.
7 ANALYSIS OF HUMAN SKELETAL MATERIAL
Ohio Historical Society Collections
RAYMOND S. BABY and SUZANNE M. LANGLOIS
The fragmentary skeletal materials from the early ex-
cavations in the Edwin Harness Mound by Warren K.
Moorehead (1897) and William C. Mills (1907), presently
in the Ohio Historical Society collections, represent only
a small part of the total number of burials recorded in the
field. Thus general statements concerning the total popu-
lation from which the individuals came cannot be made.
There are cremated bone fragments in the collections
from the Harness “Rectangular Grave Exhibit.” At least
three individuals are represented: an adult male, an adult
female, and one immature individual. These bones prob-
ably represent a conglomeration of several burials exca-
vated and combined for exhibit. The duplication of parts
includes condylar fosses, mandible with tooth sockets,
right mandibular fragments, and vertebrae. The burning
of the remains follows the standard pattern for Hopewell
(Baby 1954). The analysis of these bones is included in
Table 7.1.
The cremated burial excavated in 1977 (Feature 56)
contained two thin skull fragments, one anterior inferior
parietal, one mastoid process, one mandible, and several
long bones. The coronal suture exhibits the beginning of
closure. These bones are from an adult female, around 30
years old. Again, the burning of the remains follows the
standard pattern for Hopewell.
References
Baby, Raymond S.
1954 Hopewell cremation practices. Papers in Archaeology
No. 1, Ohio Historical Society.
Mills, William C.
1907 Explorations of the Edwin Harness mound. Ohio Ar-
chaeological and Historical Quarterly 16:1 13-193.
Moorehead, Warren K.
1897 Report of field work carried out in the Muskingum,
Scioto, and Ohio Valley during the season of 1896.
Ohio Archaeological and Historical Quarterly 5:
165-274.
Cleveland Museum of Natural History Collections
STEPHANIE J. BELOVICH
All of the skeletal material included in this analysis was
excavated from the Edwin Harness Mound (33Ro22) in
1977 by field crews under the direction of N'omi Greber.
The skeletal material analyzed was sparse and fragmen-
tary, few bones being complete. As a result, analytic
procedures were limited to 1) inventory, 2) age and sex
determinations, where possible, and 3) gross macroscopic
examination for osteopathology.
Anthropometric measurements were not possible due
to the incompleteness of the bones. Skeletal age determi-
nations were based upon the following criteria: 1) dental
eruption (Brothwell 1972) and 2) epiphyseal fusion
(Schour and Massler, cited in Brothwell 1972; Bass 1971;
and Krogman 1962). Determinations of sex were based
upon pelvic examination for major sexing criteria (Bass
1971). The material was also examined for the presence of
four categories of pathological skeletal lesions: 1 ) develop-
mental, 2) degenerative, 3) infectious, and 4) traumatic.
The results of these analyses are summarized in Table 7.2.
The three burials recovered during the 1 977 field season
can hardly be considered a population, and only insuffi-
cient data can be obtained for the other Harness burials;
thus, it is impossible to address questions of population
dynamics. Nonetheless, some statements can be made.
Feature 60 was a charred and partially burned bundle
burial. An age determination of 19-22 years was based
upon the following observations:
1) absence of epiphyseal fusion of the iliac crest
2) partial epiphyseal fusion of the femur
3) dental development
4) dense, uniform trabecular bone
The skeleton was determined to be that of a female be-
cause of the presence of a wide sciatic notch and a deep pre-
auricular sulcus. No pathological lesions were observed.
Feature 75 was a primary inhumation recovered from a
disturbed context, which resulted from previous excava-
tions. An age determination, other than to classify this
individual as an adult, was not possible. The porous na-
ture of the trabecular bone and the presence of osteophy-
tosis, however, suggest that this adult was past middle
age. The major sexing characters of the innominate were
missing. A small portion of what appeared to be the be-
ginning of a deep pre-auricular sulcus was present. This
suggests that the individual was female. Osteophytosis
TABLE 7.1
Edwin Harness Mound Human Skeletal Material Excavated 1896-1905
O.H.S.
No.
Bone
Sex
Age
Comments
7/56
Cranium with facial mask
absent
Male
Approx. 45 years
Marked bifrontal flatten-
ing, probably mesocranic
7/B4,
14150
Facial mask only
Initial development of
supraorbital ridges sug-
gests male
Immature, 4-5 years
No evidence of bifrontal flat-
tening
7/B2,
13849, &
Complete right parietal,
part of frontal
Immature (child)
Not part of7/B4, 14150
13850
7/BI, Intact left ilium Greater sciatic notch sug- Under 13years Elements of innominate un-
13814 gests female united, iliac crest ununited
7/B6, Left tibia Immature Epiphyses ununited
14171
7/B4,
14152
Portion of right frontal
Adult probably male, from
superior orbital margins
—
Not related to 7/B4, 14150
13910/
1391 1
Nearly complete right
parietal, portions of right
occiput and frontal
Male
40 years by suture closure
7/53
Intact mandible
Male
Approx. 35-40 years
Not related to 7/ 56 but could
be to 7/ B4, 1 4 1 52 or 1 3910/
1391 1, slight erosion of
right condyle resulting in
more wear on right teeth
20070
Intact mandible
Male
Not related to 7/ 56, could be
to 13910/ 11 or 14152, large
( 1 2 x 10 mm) aperture in
right ascending ramus 4.5
mm below notch, due to
bone tumor; healing begun
on exterior surface,
draining-type abscess;
some trauma to right con-
dyle, marked reduction of
right ramu compared to left
7/
Parts of both skull and post-
cranial skeleton (ca. 19%)
represented
Acetabulum suggests female
Lapsed union on posterior
sagittal suture places age
above 45 years
Bone ranging from com-
pletely normal (unburned)
to completely incinerated
7/
(exhibit)
Male
ca. 50 years
Vault extremely thick,
arthritic lipping of con-
dylar fossa and upper
lumbar & lower thoracic
vertebrae quite extensive
7/
(exhibit)
Female
45-47 years
Slight arthritic lipping on
cervical, none on condylar
fossa, sagittal and part of
lambdoid suture com-
pletely closed
7/
(exhibit)
1 fragment of unburned
cervical vertebra
Immature
64
STEPHANIE J. BELOVICH
No. 39
was the only pathological lesion observed. Degeneration
involved lipping and destruction of the vertebral bodies.
Finally, copper staining was observed on several of the
bones. Table 7.2 details the extent of this observation.
Feature 84 was located in a grave cut through the
mound floor and into the natural gravels which underlay
the mound. Dental eruption and the length of the humer-
us were used to establish an age of 6-8 months. Sex, of
course, remains undetermined. No pathological lesions
were observed.
No developmental, infectious, or traumatic lesions
were observed for any of the individuals examined in this
study.
In summary, the three individuals recovered from the
Edwin Harness Mound during the summer of 1977 were
examined for sex and age determinations and osteopa-
thology. The small sample size and its fragmentary nature
limited the scope of the present study to inventory and
description.
TABLE 7.2
Edwin Harness Mound Human Skeletal Material,
non-cremated, 1977
Feature # Specimen Comments
60 partially burned bundle
burial female, !9-22yrs
frontal; missing small
portion of squamous
and L & R orbits
L temporal; missing
squamous
R temporal; missing
squamous and mastoid
L zygoma
L & R parietal
occipital; missing base, R
condyle and portion of
squamous
L & R nasal
L & R maxilla; missing
frontal process and
portion of body
L mandible
R mandible; missing
coranoid process, con-
dyle and portion of
ascending ramus
148 skull fragments
3 incisors
1 canine
3 premolars
12 molars
2 roots
L scapula; missing supe-
rior and medial borders,
glenoid cavity, coracoid
and portions of body,
axillary border and spine
5 thoracic spinous
processes
Feature # Specimen Comments
L articular facet of atlas
vertebra
17 vertebra fragments
36 rib fragments
L & R radius; shaft frag-
ments
L ulna; shaft fragments
R femur
L femur; mid-shaft frag-
ment, greater trochanter
and neck
L tibia; proximal and
distal fragments
R tibia; distal fragment
(L/R)? tibia; mid-shaft
fragment
L & R fibula; distal
fragments
84 long bone fragments
L & R patella
R innominate; articular
surface and iliac portion
of acetabulum
L innominate; small por-
tion of ilium and iliac
portion of acetabulum
5 innominate fragments
L & R calcanous
L talus
L cuboid fragment
L 2nd cuneiform
R 2nd cuneiform
navicular fragment
L 3rd metatarsal
fragment
L 4th metatarsal
fragment
R 5th metatarsal
fragment
foot phalanges: 5 prox-
imal, 2 medial, 1 distal
hand phalanges: 1 prox-
imal
1st metatarsal fragment
4 metatarsal/ metacarpal
(?) fragments
3 phalange fragments
1000 plus unidentifiable
bone fragments less
than 1 cm in size
75 primary inhumation,
adult female, osteo-
phytosis, copper stain-
ing on many of the bones
5 cranial fragments
L scapula; acromion and copper staining
body fragments
R scapula; glenoid and copper staining
acromion, 3 body
fragments
1983
EDWIN HARNESS MOUND
65
Feature # Specimen
Comments
Acknowledgments
I wish to thank R. P. Mensforth, C. O. Lovejoy, and
W. H. Kimbel for their review of the analysis.
References
Bass, William M.
1971 Human osteology: a laboratory and field manual of
the human skeleton. Missouri Archaeological So-
ciety, Columbia, Missouri.
Brothwell, D. R.
1972 Digging up bones. British Museum (Natural History),
London.
Krogman, Wilton Marion
1962 The human skeleton in forensic medicine. Charles C.
Thomas Publ., Springfield, 111.
L & R frontal orbits
L & R lateral portions of
occipital
L & R petrous portions
22 cranial fragments;
parietal, temporal,
frontal, occipital and
sphenoid represented
R mandible; condyle and
portion of ascending
ramus
L mandible; body and
condyle
1 incus
2 malleus
6 incisors
2 canines
6 deciduous molars
4 permanent molars
R neural arch of 1st
vertebra
24 neural arches
7 centrums
L scapula; portion of
axillary border and spine
25 rib fragments
L humerus fragment
1 sacral fragment
100 plus unidentifiable
bone fragments less than
1 cm in size
35 vertebral fragments
L femur; proximal head,
portion of shaft
R femur; mid-shaft
fragment
61 long bone fragments;
upper/ lower limbs
L patella
R patella fragment
R innominate; acetabu-
lum, portion of ilium
L innominate; portion of
acetabulum
14 innominate fragments
L talus fragment
calcaneus fragment
navicular fragment
L 2nd cuneiform fragment
L 3rd cuneiform fragment
19 metatarsal/ metacar-
pal (?) fragments
300 plus unidentifiable
bone fragments less than
1 cm in size
4 unidentifiable burned
bone fragments
5 with copper staining
copper staining
5 with copper staining
5 with copper staining
84
Infant, 6-8 mos.
8 MOLLUSC IDENTIFICATION AND ANALYSIS
DAVID R. MORSE
The archaeological salvage excavations of the Edwin
Harness Mound (1976-1977) by the Cleveland Museum
of Natural History and the Ohio Archaeological Council
recovered 332 molluscan remains (19 species) in various
cultural contexts associated with the mound. The shell
remains analyzed included local Ohio naiads (51%) and
terrestrial gastropods (31%) as well as marine gastropods
( 1 7%) from the Atlantic Coast. As this research project is
only the latest phase in the investigation of the Edwin
Harness Mound, it is likely that the mollusc material dis-
cussed in this report represents only a small sample of the
items that were originally associated with the mound
structure (over 3,000 shells are part of the Edwin Harness
Mound collection at the Ohio Historical Society). Al-
though the sample recovered in these recent excavations
is small, it is hoped that some interpretational context can
be given to the importance and use of mollusc species by
Hopewell groups (see Appendix 8.1).
As part of the overall project, two major areas of the
site were excavated. First, in order to define any remnant
sections of the mound structure, the entire floor was ex-
posed. Second, once the mound structure had been de-
fined, exploratory radial trenches were opened with the
aid of power equipment to expose any additional midden
areas or structures. Prehistorically used shell material was
recovered from both areas.
In order to characterize the range of the mollusc species
found on this site, an attempt was made to identify the
genus and species of all items. At this site, however, most
of the material consisted of only small sections of shell,
many times lacking the phenotypic characteristics used to
distinguish between species. As a result, the specific iden-
tification in this report is tentatively offered to facilitate a
discussion of dietary, environmental, and other variables.
The relative frequency of species (Table 8.1) has been
based on a tally of the number of locations from which a
species was recovered (maximum number of individuals,
MaxNl; Grayson 1973). Although this method has many
problems (Grayson 1973), in this case, as the material was
widely scattered over the site, it would seem unlikely that
fragments are from the same individuals.
To add to these problems, a third of the shell material
recovered from the 1976-1977 Harness Mound excava-
tions were found in a disturbed stratigraphic context.
Many of the early investigators of the Ohio Hopewell
complex included the Harness Mound in their field inves-
tigations. This resulted in much of the interior sections of
the mound being devoid of stratigraphically significant
material. Most of the mollusc remains discussed in this
report were recovered from the exterior edges of the
house structures and mound gravel ring (Feature 1).
In addition, the mollusc material from the Harness
Mound curated at the Ohio Historical Society (OHS) ap-
pears to be very selective. Most of this collection consists
of exotic marine shell fragments and shell beads. M ollusc
species indigenous to Ohio, all naiads, constituted less
than 1% of the OHS collection (93% of the material from
the 1976-1977 excavations consisted of indigenous spe-
cies). It is interesting to note that the relative frequencies
of local species that are represented in both collections are
similar, suggesting that both samples are part of the same
overall statistical population (see Appendix 8.2).
Naiads
Compared to other sites in southern Ohio, the range of
naiad species at the Harness Mound site is somewhat lim-
ited (see Table 8.1). A total of three species were identi-
fied from the Harness Mound site: Lampsilis ovata
(30%), Elliptio dilatatus (45%), Amblema costata (25%).
These species are some of the most common naiad
(freshwater pelecypods) species found in archaeological
sites and in recent collections from central and southern
Ohio (Stansbery 1965). Elliptio dilatatus (common filter
clam or spike mussel), a very adaptable species, is found
in a wide range of riverine habitats, which probably ac-
counts for its high frequency at Harness. On the other
hand, Amblema costata (common river mussel) and
Lampsilis ovata (ovate river mussel) are found in small,
slow moving, shallow streams or tributaries of large
streams, preferring sandy and gravel bottoms. It is likely
that these species somewhat reflect local conditions of the
Scioto River near the mound.
It is interesting to note that a wider range of naiads were
found at the McGraw site (Prufer 1965) and the Morrison
Village site (Prufer and Andors 1967), both of which are
located near Harness on the Scioto River in Ross County.
For example, in the McGraw site report, Stansbery ( 1 965)
lists 25 naiad species. It is likely that a wider range of
naiad species had existed in the Scioto River during the
occupation of Harness than were found in the recent ex-
cavations. The differences in the relative numbers of spe-
cies between these sites probably reflect some dissimilar-
ity in activities (e.g., Brose 1972). If the McGraw site is
primarily a habitation site, then one would expect that
there would be more evidence of food procurement and
tool manufacture activities, which seems to be the case.
The low number of naiad shells at Harness might be re-
lated to occasional manufacture of a limited range of arti-
facts, possibly including shell beads. Even though many
have assumed that most shell beads were made from ex-
TABLE 8.1
Summary of Mollusc Taxa Identified, Edwin Harness Mound
Max NI* Max NI*
Taxa 1976-1977 Exc % OHS %
Naiads
Fragments, species indeterminate
21
43.8
Lampsilis ovata
1
2.1
5
20.8
Lampsilis sp?
5
10.4
3
12.5
Total Lampsilis
6
12.5
8
33.3
Elliptio dilatatus
6
12.5
3
12.5
Elliptio sp?
6
12.5
10
41.7
Total Elliptio
12
25.0
13
54.2
Amblema costata
3
6.3
2
8.3
Amblema sp?
6
12.5
1
4.2
Total Amblema
9
18.8
3
12.6
Total
48
100.1
24
100.1
Terrestrial Gastropods
Fragments, species indeterminate
11
28.2
Stenotrema leaii
1
2.6
Mesodon sp?
1
2.6
Anguispira alternata
1
2.6
Anguispira sp?
4
10.3
Total Anguispira
5
12.8
Oxychilus sp?
1
2.6
Zonitoides arboreus
4
10.3
Discus cronkhnitei
2
5.1
Helicodiscus para.
8
20.5
Retinella wheatyi
1
2.6
Vertigo morsei
2
5.1
Pupilla sp?
1
2.6
Cionella lubrica
1
2.6
Hawaiia miniscula
1
2.6
Total
39
100.2
Marine Gastropods
Marginella sp?
2
28.6
500**
Jaspidella jaspidae
2
28.6
3
Olivella sp?
1
14.3
1400
Oliva sp?
4
Vitrinella sp?
6
Busycon contrarium
1
14.3
2
Busycon spiratum
1
Fasciolariidae
2
unidentifiable shell bead
1
14.3
2000**
Total
7
100.1
♦maximum number of individuals
♦♦estimate
68
DAVID R. MORSE
No. 39
otic marine gastropods, it is possible some naiads were
used in the manufacture of shell beads. The three naiad
species found at Harness are among those few species
thick enough for shell bead manufacture.
Terrestrial Gastropods
At the Harness Mound site, at least ten terrestrial gas-
tropod species were recovered from a total sample of 39
individuals (MaxNI). All items were collected as exca-
vated; however, most of the smaller sized (below 3 mm)
species were recovered from flotation samples of post
holes and features (mostly in the genera Zonitoides , Dis-
cus, Helicodiscus, Retinella , Vertigo , Pupuilla). All the
terrestrial gastropods in the sample are indigenous to
Ohio except for Oxychilus (cellar snail). This species was
introduced from Europe in the eighteenth century (La-
Rocque 1 970) and most likely is related to the excavations
early in this century. The single specimen identified was
found at the top of Feature 89.
The interpretation of gastropods in an archaeological
context is often related to paleo-environmental variables,
especially in a general description of the vegetation cover
of the site. Although it is difficult to relate many of the
species discussed here to specific environmental condi-
tions per se, some comments seem to be warranted. Most
of the species found at Harness, especially in the genera
Anguispira (cf. Alternata ) (striped forest snail), Zoni-
toides (zonite shell). Discus (common disk shell), and Heli-
codiscus (parallel disk shell), are most often found within
forest detritus: logs, stumps, and other decaying material.
So it seems reasonable to suggest that a certain amount of
forest cover existed in the area of the mound. Helicodis-
cus is also associated with second growth areas which
might have been present near the mound. The absence of
a significant number of species which favor either cleared
areas, especially in the genus Mesodon , or heavily
wooded areas, such as Anguispira kochi, is notable. This
is in contrast to other archaeological sites in south-central
Ohio (sites along Caesar Creek, Broseetal. 1979; Killen A
and B sites, Brose et al. 1979; McGraw site, Prufer 1965;
and Morrison Village site, Prufer and Andors 1967). Al-
though this sample is small, it suggests that the vegetation
cover near the mound could best be described as an open
forested area or parkland.
This range of gastropods at the Harness Mound varies
from the published list of molluscan fauna from the Mor-
rison Village (Prufer and Andors 1967) and McGraw
(Stansbery 1965) sites near the Harness Mound along the
Scioto River in Ross County. At both sites Anguispira
kochi and Allogona profunda (profound forest snail),
primarily a forest adapted species, are important at these
sites. So, it would seem that these sites were more heavily
wooded than was the area near the Harness Mound.
Marine Gastropods
Seven marine gastropods were recovered in the 1 976—
1977 excavations at the Harness Mound site. Taxonomi-
cally, these molluscs can be grouped into three major fam-
ilies: Olividae (olive shells), Marginellidae (marginella
shells), Melongenidae (conch shells). It was obvious that
these items were part of a much larger sample, as over
3,000 marine gastropods from the Harness Mound are
curated at the OHS. In general, all items are native to the
southeast Atlantic Coast or Gulf Coast of the United
States. Although the items recovered in the present exca-
vations seem to be finished artifacts, the earlier collec-
tions at the OHS also include a whole range of unmodi-
fied, partially finished and worked fragments.
The largest item recovered in the excavations was a
conch shell or lightning whelk. Busy con contrarium (Fea-
ture 60, bundle burial), which had been modified by re-
moving the columellae and trimming the edge surface of
the aperture. (The term “whelk” as applied to B. contra-
rium should not be confused with waved whelks in the
family Buccinidae.) In comparison with other shell of B.
contrarium , this item is somewhat small but still can be
considered a fully developed individual (see Table 8.2). In
the family Melongenidae , four species besides B. contra-
rium are found along the Atlantic coast in the United
States: B. carica( knobbed whelk), B. canliculatum (chan-
neled whelk), B. spiratum (fig whelk), B. perversum (per-
verse whelk). Of these, B. contrarium has the most south-
ern range, which extends from North Carolina to Florida.
In the Harness Mound collections at the OHS, two addi-
tional B. contrarium , one B. spiratum , and one Fascio-
lariidae (tulip shell or horse conch) have been identi-
fied tentatively. These species are common to many
Hopewell sites in the Ohio Valley. Seip and Mound City
have a few molluscs in the genus Crassis (helmet shell)
TABLE 8.2
Comparison of Dimensions,
Melongenidae — Conch Shells
Cat. No.
Busycon contrarium
Length
Width
OHS-Collections
1375G 7/34
20.5 cm
12.2 cm
OHS-Collections
7/32
27.5
16.8
1976-1977 Excavations
28 IE
17.5
8.5
Mean of OHS
Material
24.0
14.5
Overall Mean
21.8
Busycon spiratum
12.5
OHS-Collections
13757 7/34
21.6 cm
14. 1 cm
1983
EDWIN HARNESS MOUND
69
which are absent from Harness. Crassis is native only to
southern Florida, which might have been beyond the
range of southern contacts for the inhabitants of Harness.
The remaining items are within the families Marginel-
lidae ( Marginella sp?, marginella shells) and Olividae
( Olivella sp?, dwarf olive shells; Jaspidella jaspidae ,
jasper dwarf olive shells). The native range for these spe-
cies is very similar to Busycon (North Carolina to Flor-
ida), which seems to confirm that this region is the overall
acquisition area for the Harness Mound shell material.
These species, recovered from the 1976-1977 excavations,
characterized 97% of the marine shell at the OHS from
Harness and seem to be a good sample of the overall
collection.
Spatial Distributions
As mentioned, the molluscan faunal collection from
the Edwin Harness Mound is the result of excavations of
the remaining mound floor in its entirety and a sampling
of selective areas outside of the mound structure. This
procedure included the excavation and collection of flota-
tion samples from all features (burials, hearth, and activ-
ity areas) and all post molds. Although large sections of
the mound had been excavated previously, it was noted
that some of the mound structure and sections of the
mound floor were intact. In addition, the investigators
found post hole outlines of a major house structure on the
mound floor, part of an early construction phase. As mol-
luscs were found within a number of features, an analysis
of their distribution was considered important to the
overall understanding of the site. Assemblage variability
and frequencies of individuals found within a particular
feature of significant stratigraphic level were compared to
the site as a whole although sample size was not large.
A comparison of molluscs within and beyond the limits
of the mound found approximately 90% of the same local
aquatic and terrestrial species in both areas. That eight
times as many individuals were found within the mound
area is attributed to the amount of area excavated. No
significant differences in relative frequency or species dif-
ferences were noted between differing intra mound struc-
tures or between stratigraphic features and/or levels.
These data suggest that proximal vegetation patterns
within the general site area were consistently parkland
forest throughout the sequence of structure and mound
construction. Fifty percent of the smaller species ( Zoni -
toides , Discus , Helicodiscus , Vertigo) were directly asso-
ciated with post holes, and an additional 20% of these
molluscs were found near the highest density of post
holes. As these species commonly feed upon decaying
plant material, it seems that some of these posts rotted in
situ. Because such posts were evenly distributed in differ-
ent structures across the mound floor, it may be that these
structures are more or less coeval. At any rate, these data
strongly support the position that every structural por-
tion of the submound structure underwent some more-or-
less extended period of exposure prior to either burning
or entombment.
Since these samples were small, these tentative hypoth-
eses should be evaluated in future research.
Discussion
The identification and interpretation of mollusc mate-
rial from the Edwin Harness Mound excavations and
Ohio Historical Center collections has provided some
context for the discussion of environmental and cultural
variables, in spite of the small sample and extensively dis-
turbed areas. Although all of the recommendations were
provisional and inductive, this material adds to the gen-
eral data base for Hopewell sites. It is hoped that mollusc
samples will be collected in the future.
Overall, the Harness Mound material seems to exhibit
a different range of molluscs than were found on other
sites along the Scioto River. Principally, it is suggested
that the number of naiads were low because shell was used
mainly as raw material for a small number of specific arti-
facts (e.g., shell beads) and that indigenous gastropods
are related to an open parkland forest environment.
Analysis of the spacial distribution seems to indicate that
these environmental conditions were present on the site
during all construction phases of the mound.
Appendix 8. 1
Mollusc Identification Analysis
Edwin Harness Mound
1976-1977 Excavations
Cat. No.
Provenience
Taxa
2D
N535-E492.5, disturbed
1 frag., gastropod
2F
N535-E492.5, disturbed
1 frag., gastropod
3H
N547.5-E492.5, dis-
turbed but close to
original floor
1 frag., gastropod
3J
N547.5-E492.5, dis-
1 ventral margin frag..
turbed but close to
poss. Lampsilis ,
original floor
burned
13B
N547.5-E492.5, dis-
turbed but close to
original floor
1 frag., gastropod
2 1J
N547.5-E502.5, dis-
1 frag., gastropod, poss.
turbed near “Boys’
1880”
Anguispira
23A
N547.5-E502, disturbed
1 shell bead
24A
N535-E495, Fea. 89
1 Oxychilus
24C
N535-E495, Fea. 89
1 frag., naiad
1 Vertigo morsei
1 Helicodiscus
70
DAVID R. MORSE
24C
N535-E495, Fea. 89
1 dorsal margin frag.
(bag #2)
w/o beak. Elliptic)
dilatalus
24E
N535-E495, Fea. 89
1 Cionella lubrica
30
Surface, collected June
1 R dorsal margin frag.
1976
w/o beak, Elliptio
dilatatus
32
N540-E492.5, south
1 R dorsal margin frag.
balk surface
w/o beak, Elliptio
dilatatus , burned
37A
N535-E492.5, post hole
#16
2 frag., naiad
39B
N535-E492 5, post hole
#15
20 frag., naiad, burned
39B
N535-E492.5, post hole
1 frag., naiad, burned
(Bag #2)
#15
64A
N532.5-E492.5, dis-
turbed
1 frag., naiad
64C
N532.5-E492.5, dis-
1 dorsal margin frag.,
turbed
Marginella
103B
N522.5-E492.5, dis-
turbed but near other
archaeological de-
posited material
1 Jaspidella jaspidea
103C
N522.5-E492.5, dis-
1 frag., naiad, poss.
turbed but near other
archaeological de-
posited material
Elliptio
109
Surface, collected July
1 frag., naiad, poss.
1976
Amblema
1 17B
N520-E505, disturbed,
Fea. 1
1 frag., naiad, burned
1 18D
N535-E495, front
1 R dorsal margin frag.
loader cut #2, dis-
w/ beak, Amblema
turbed
costata
128 A
N535-E495, poss. post
hole #8
10 frag., 1 naiad
130
N535-E500, poss. post
hole #10
1 frag., naiad
138C
N522.5-E495, floor
surface
1 frag., naiad
138C
N522.5-E495, floor
4 frag., naiad, poss.
surface
Elliptio
139B
N520-E445, floor
surface
1 frag., naiad
168
N522.5-E495, post hole
1 R dorsal margin frag.
#80
w/o beak, Elliptio
dilatatus
168
N522.5-E495, post hole
1 ventral margin frag.
(bag #2)
#80
Elliptio
183E
N525-E495, Fea. 30
6 frag., naiad, poss.
Amblema , burned
183F
N525-E495, Fea. 30
1 dorsal section of
Marginella
1 frag., gastropod
188A
N532.5-E495, post hole
1 dorsal margin frag.
#40
w/o beak, poss.
Elliptio
217
No provenience
2 frag., naiad, burned
4 frag., gastropod,
burned
No. 39
1 frag., Amblema
costata
219A
Surface, collected June
10 frag., naiad, poss.
1967
Amblema, burned
223D
N520-E507.5, Fea. 37
1 Anguispira alternata
271 A
N552-E500, disturbed
8 frag., naiad
1 ventral margin frag.,
poss. Lampsilis
27 IQ
N552-E500, post hole
#179
1 Zonitoides
272C
N560-E512.5, backdirt
10 frag., naiad
1 dorsal margin w/o
beak, Lampsilis
28 IB
N517.5-E502.5, Fea. 60
15 frag., naiad, poss.
Elliptio
28 IE
N517.5-E502.5, Fea. 60
1 Busycon contrarium
281C
N517.5-E502.5, Fea. 60
4 frag., naiad
1 frag., poss. Steno-
trema leaii
15 frag., naiad, poss.
Anguispira
28 ID
N517.5-E502.5, Fea. 60
10 frag., gastropod,
poss. Anguispira
290C
N535-E502.5, post hole
#171, poss. disturbed
1 R dorsal margin frag,
w/ beak, Elliptio
dilatatus
295B
N530-E502.5, post hole
#160
1 Olivella
299A
N500-E502.5, Fea. 55
6 frag., 2 gastropods,
poss. Anguispira
305B
Front loader Area #3,
post hole #102
1 Helicodiscus
308K
N524-E485, post hole
#270
3 frag., naiad
3 10C
N537.5-E497.5, post
hole #252
1 F dorsal margin frag,
w/o beak, Elliptio
dilatatus
31 1C
N522-E495, mound
loading directly over
Fea. 3
1 L dorsal margin frag,
w/ beak, Lampsilis
312C
N540-E497.5, post hole
#66
30 frag., naiad
Flotation Samples
F2B
N520-E507.5, Fea. 37
7 Zonitoides arboreus
4 Helicodiscus par-
allelus
1 Discus cronkhitei
224H
N522.5-E485, east of
Fea. 1
8 frag., 1 gastropod
228B
N547.5-E510, poss.
post hole #20
3 frag., 1 gastropod
229 E
N550-E507.5, mound
2 frag., naiad, poss.
floor
Elliptio, burned
229Q
N550-E507.5, post hole
1 Helicodiscus par-
#141
allelus
230L
N522.5-E497.5, post
2 Helicodiscus par-
hole #143
allelus
2 Zoniloides
1983
EDWIN HARNESS MOUND
71
2330
240A
240 B
240C
262A
260 D
260P
269C
269C
(Bag #2)
269 D
270A
F71A
F100
F269C
CCD68
Cat. No.
13735-7/99
13736
13737
N530-E492.5, Fea. 43
1 frag., naiad
13738-7/99
1 frag., naiad, poss. Elliptio
1 frag., gastropod
13739-7/99
1 frag., naiad, poss. Elliptio
N547.5-E507.5, dis-
15frag., poss. Amblema
13740
1 frag., naiad, poss. Elliptio
turbed
13741
1 frag., naiad, poss. Elliptio
N547.5-E507.5, dis-
9 frag., naiad
13743
1 frag., naiad, poss. Elliptio
turbed
1 L dorsal margin frag.
13745
1 whole shell, Elliptio dilatatus
w / beak, Amblema
13746-7/99
1 ventral margin frag., poss. Elliptio
1 R dorsal margin frag.
13747-7/99
1 ventral margin frag., poss. Elliptio
w/o beak, Lampsilis
14582-7/99
1 whole shell, Lampsilis ovata
N547.5-E507.5, dis-
1 Helicodiscus
14583-7/99
1 frag., Lampsilis ovata
turbed
14584-7/99
1 whole shell, Amblema costata
N550-E509, above
1 L dorsal margin frag.
14585-7/99
1 whole shell, Lampsilis ovata
Fea. 33
w/ beak, Elliptio
14586-7/99
1 whole shell, Amblema costata
dilatatus
14587-7/99
1 whole shell, Lampsilis ovata
N557.5-E500, Fea. 44
1 dorsal margin frag.
14588-7/99
1 whole shell, Lampsilis ovata
w/ beak, Amblema
1 ventral margin frag., poss. Amblema
costata
14589-7/99
1 ventral margin frag., poss. Lampsilis
N557.5-E500, Fea. 44
6 frag., naiad
14590-7/99
1 frag., Elliptio dilatatus
backdirt
1 dorsal margin frag, w/
#7
1 frag., naiad, Amblema
beak, poss. Amblema
1 frag., naiad, poss. Elliptio
Backhoe trench #4,
1 frag., gastropod
1 frag., naiad, poss. Lampsilis
post hole #147 outside
of Fea. 1
not labeled
1 whole shell, Elliptio dilatatus
Backhoe trench #4,
12 frag., naiad
Conch
post hole #147
1 ventral margin frag.,
Lampsilis
1 frag., poss. Elliptio
6 frag., naiad
1 375G *-7 / 34
1 Busycon contrarium
N517.5-E502.5,
-7/32
13757-7/34
1 Busycon contrarium
1 Busycon spiratum
backhoe trench #4
13755-7/33
1 Fasciolariidae
Fea. 56
1 frag., gastropod
14590-7/99
1 frag., poss. juvenile of Pleuroploca
N517.5-E507.5, back-
7 Helicodiscus par-
gigantea
hoe trench #4, Fea. 60
allels
Post hole #147
1 frag., poss. Mesodon
Small Marine Gastropods
Post hole #147
3 Zonitoides
I frag., poss. Pupilla
40 Helicodiscus par-
allels
36 Relinella wheatleyi
66 Zonitoides arboreus
I I Vertigo morsei
7/36
4 Oliva sp?
1400** Olivella sp?
500** Marginella sp?
6 Vitrinella sp?
3 Jaspidella jaspidea
Modified Shell
Carbon Samples
Backhoe trench #6 3 Retinella sp?
18 Hawaiia minuscula
Appendix 8.2
Preliminary
Mollusc Identification Analysis
Edwin Harness Mound
Ohio Historical Society
Harness General 20 dorsal sections, Marginella , burnished
1 string — single row, ventral sections,
Marginella
1 string — double row, dorsal sections,
Marginella
1 string — double row, ventral sections,
Marginella , burnished
4 strings — double row, dorsal sections,
Marginella
1 string — single row, dorsal sections,
Olivella
Naiads
Taxa
1 frag., naiad, poss. Elliptio
1 frag., naiad, poss. Elliptio
1 frag., naiad, poss. Elliptio
1 frag., naiad, poss. Lampsilis
♦This is the catalogue number written on the specimen. Based
on the original catalogue of materials collected by Moorehead,
the number likely should be 13756.
♦♦estimate
72
DAVID R. MORSE
No. 39
References
Abbott, R. Tucker
1 974 American seashells (second ed.). Van Nostrand Rein-
hold, New York.
Brose, David S.
1 972 The molluscan fauna. In The Schultz site, compiled by
J. E. Fitting. University of Michigan Museum of An-
thropology Memoirs No. 4, Ann Arbor.
Brose, David S., Donald R. Bier, Jr., Judith Astramerki, Fred-
rick Chapman, Richard Ford, Robert Mensforth, David Morse,
and Paul Storch
1979 Prehistoric occupation of the Killen Electric Generat-
ing Station near Wrightsville, Adams County, Ohio.
Submitted to the United States Department of the In-
terior Interagency Archaeological Services, Atlanta;
Contract 5880-7-0070 by the Cleveland Museum of
Natural History, Cleveland.
Brose, David S., and Nancy M. White with contributions by
Stephanie Belovich, Fredrick Chapman, W. B. Clapham,
Richard I. Ford, John Lallo, David Morse, and Paul Storch
1979 Archaeological investigations of prehistoric occupa-
tion in Caesar Creek Lake: Clinton, Greene, and
Warren Counties, Ohio. Submitted to the United
States Department of the Interior Interagency Ar-
chaeological Services, Atlanta; Contract C-5555
(formerly CX5880-7-01 34) by the Cleveland Museum
of Natural History, Cleveland.
Cinadr, Thomas J., and David S. Brose with contributions by
Stephanie Belovich, Donna Benson, David Morse, Paul Storch,
and Nancy M. White
1978 Archaeological excavations in Caesar’s Creek Lake,
Ohio: Section 1 1 . The Carr Mill Race Site (33Wa75).
Submitted to the United States Department of the In-
terior Interagency Archaeological Services, Atlanta;
Contract CX-5880-7-0134 by the Cleveland Museum
of Natural History, Cleveland.
Grayson, Donald
1973 On the methodology of faunal analysis. American An-
tiquity 39:432-438.
LaRocque, Aurele
1970 Pleistocene mollusca of Ohio. State of Ohio, Depart-
ment of Natural Resources Division of Geological
Survey Bulletin 62(4).
Prufer, Olaf H., D. H. McKenzie, O. Pi-Sunyer, H. C. Cutler,
R. A. Yarnell, P. W. Parmalee, D. H. Stansbery
1965 The McGraw site: a study in Hopewellian dynamics.
Cleveland Museum of Natural History, Scientific
Publications (n.s.), 4(1).
Prufer, Olaf H., and Ellen Andors
1 967 The Morrison Village site (33Ro-3): a terminal prehis-
toric site in Ross County, Ohio. In Studies in Ohio
Archaeology , edited by Olaf H. Prufer and Douglas
H. McKenzie, pp. 187-229. Kent State University
Press, Kent, Ohio.
Stansbery, David H.
1965 The molluscan fauna from the McGraw site. In The
McGraw site: a study in Hopewellian dynamics ,
edited by Olaf H. Prufer, pp. 1 19-124. Scientific Pub-
lications of the Cleveland Museum of Natural His-
tory, (n.s.), 4(1).
9 THE FLINT SOURCES
KENT D. VICKERY
Introduction
Four collections of chipped stone items were analyzed:
1. artifacts and debitage recovered during the 1976
field season of excavation at the Edwin Harness
Mound, 33Ro22, consisting of 127 items;
2. 1,180 bladelets and bladelet cores surface collected
by Robert Harness in the vicinity of the Harness
Mound;
3. 627 surface collected artifacts and pieces of debitage
from Harness’s Site 18; and
4. a similar collection of surface material from Har-
ness’s Site 25, consisting of 720 chipped stone items.
Chipped stone material from the Harness Mound ex-
cavation consisted of 91 pieces of debitage, 23 bladelets, 1
bladelet core, and both unifacial and bifacial tools and
weapons. A bipolar core and 1 tool edge rejuvenation
flake were treated as artifacts (rather than debitage) for
purposes of this analysis. The Site 1 8 collection included 7
projectile points, 4 bifacial artifacts, and 4 bladelets in
addition to abundant debitage. Debitage also dominated
the Site 25 collection, which included 4 projectile points
and 14 bladelets. The remaining surface collection con-
sisted of 1,027 bladelets and 153 bladelet cores. For site
locations see Greber, Davis, and DuFresne (1981). The
present analysis focuses on flint/ chert raw material iden-
tifications for the 2,654 items from these four collections
and on the chipping techniques which produced the 91
pieces of debitage from the Edwin Harness Mound. This
latter analysis focuses on 1) incidence of heat alteration,
2) presence or absence of post-detachment utilization
and/ or retouch, 3) debitage type, and 4) technique of de-
tachment. Unmodified and broken chert pebbles which
were found in post holes and features of the 1976 excava-
tions of the Harness Mound were also examined for raw
material identifications. These pebbles lacked pronounced
bulbar scars and compression ring segments, platform
preparation, or other evidence of intentional fracture for
artifact production.
Methods
Flint raw material identifications were made by ex-
amining specimens under low magnification using a bi-
nocular microscope and comparing them with samples
obtained from verified outcrop locations. In addition,
cortex characteristics were recorded for the Harness
Mound debitage in an effort to determine the general con-
texts from which chert was obtained. “Pebble” chert is
material dislodged from its naturally occurring matrix,
transported by stream or glacial action, and subsequently
deposited some distance away from its source area. In the
process, such material acquires a hard, smooth surface
patination (cortex). Flakes with cortical retention on por-
tions of the dorsal face and/or platform, as well as arti-
facts, are referred to in this report as “Local Pebble” chert
if they cannot be matched to any of the samples from veri-
fied outcrop locations. Artifacts and debitage lacking cor-
tical remnants and unidentifiable as to flint source are re-
ferred to as “Unknown.”
Heat alteration was recorded in an effort to detect
whether or not thermal treatment was part of the se-
quence whereby artifacts were produced from parent raw
material. For those specimens exhibiting thermal modifi-
cation, macroscopic and microscopic inspection was the
basis for distinguishing between heat alteration and heat
damage. The former is commonly recognized by adjacent
lustrous and non-lustrous scars (Collins and Fenwick
1974; Greber, Davis, and DuFresne 1981:513), a smoky
or cloudy appearance, and/or slight discoloration. It is
only among these specimens that candidates for proper
heat treatment might be present. By contrast, heat dam-
aged specimens often showed extensive crazing, potlid
depressions, pronounced color changes, scalloped edges
or more serious heat-induced breakage, and / or a chalky
texture caused by water having been driven out under
thermal stress.
Utilization and retouch were determined by micro-
scopic examination. Retouched artifacts and debitage
had had a series of small, contiguous chips intentionally
removed from one or more margins, the scars of which
were typically uniform in size and shape. Utilized flakes
were often recognized by marginally detached chip scars
that were less regular in size, shape, and placement and
that occasionally resulted in a rather jagged contour. Uti-
lized edges commonly assumed one or more of four basic
forms: 1) nibbling, consisting of diminutive scars with an
ovate configuration that should have resulted from cut-
ting or scraping relatively soft materials (Binford 1963:
207); 2) hinge fractures with square or rectangular config-
urations and abrupt terminations, probably resulting
from contact with more resistant material; 3) attrition, the
gradual dulling or wearing away of formerly sharp edges
by the sustained removal of tiny chips or pieces of the tool
margin(s); and 4) polish, which succeeds attrition as the
functional portions of tools undergo progressive modifi-
cation with use against relatively soft materials. Polish is
recognized as a gloss or sheen under magnification.
74
KENTD. VICKERY
No. 39
Debitage: definitions
An unmodified piece of flint is reduced to a finished
tool through stages. The debris created at each stage is
reasonably distinctive, which facilitates a classification of
waste material for the sequence of stages. One component
of debitage analysis involves the recognition and defini-
tion of various debitage types which reflect the reduction
sequence.
For the Harness Mound collection only, each of several
such debitage types is described below in a logical knap-
ping order following the off-site selection or importation
of raw material for artifact production.
Checked pebbles. The first step is determining whether
or not the raw material acquired is suitable for knapping.
The quality of unmodified pebble chert can be checked by
removing one or more flakes to expose the sub-cortical
matrix. This activity results in a checked pebble. Com-
parable treatment of bedded material, which often exhib-
its thinly weathered surfaces from frost action, was not
observed in the sample.
Primary decortication flakes. The reduction sequence
is initiated with the systematic detachment of flakes from
a weathered piece of flint, which creates flakes retaining
cortex over the entire outer (dorsal) face. The removal of
these primary decortication flakes (White 1963:5) achieves
partial decortication of the objective raw material and
produces a core tool blank.
Secondary decortication flakes. Additional flake de-
tachment generates secondary decortication flakes (White
1 963:5). These flakes often retain the scars of one or more
previously detached primary decortication flakes and
some cortex on the dorsal face, but may be wedge-shaped
with cortex remaining only on the thickened edge. The
result of their removal is a blank in a more advanced stage
or an early stage of a preform.
Primary flakes. Wilmsen (1970:25) refers to post-
decortication flakes associated with the shaping of a
blank or preform by reducing its mass as primary flakes
and defines them as “those which were struck from a de-
corticated core.” Primary flakes retain the scars of pre-
viously detached flakes over most or all of the dorsal face,
with small amounts of cortex sometimes remaining
(commonly in the center). They are characteristically
thick and often triangular in section. Their removal pro-
duces, or reduces the mass of, a preform.
Thinning flakes. Crabtree (1972:94) defines thinning
flakes as “flakes removed from a preform by pressure or
percussion to thin the piece for artifact manufacture.
Thinning flakes are also removed to thin a biface or a
uniface. ” Because the process of bifacial reduction is con-
tinuous, the distinction between thinning flakes and pri-
mary flakes detached in the advanced stages of reducing
the mass of blank or a preform is often arbitrary. The
former are typically thinner than primary flakes and re-
tain scars over the entire dorsal face. Their removal pro-
duces an advanced stage preform or a finished bifacial
tool or weapon. Not all bifacial artifacts reach an ad-
vanced stage of reduction because their suitability as tools
for certain kinds of tasks (e.g., heavy cutting, scraping, or
chopping) may have been achieved by removing only de-
cortication or primary flakes.
As the preform approaches the morphology of the fin-
ished tool, sharpening flakes may be detached in an effort
to strengthen, straighten, and/or sharpen the edges or to
shape them by creating indented or projecting contours
(Vickery and Lambert 1977). Representing the final stage
of the reduction sequence, these tiny chips may not be
encountered unless fine mesh screening or flotation tech-
niques are employed during site excavation.
For the analyzed sample, flake type is used for all flake
debitage from the primary decortication stage through
the thinning stage, no sharpening flakes having been rec-
ognized in the Harness Mound debitage; debitage types
include all flake types plus checked pebbles and unclassifi-
able core fragments (exclusive of bladelet and bipolar
cores). The debitage types used for this analysis pertain to
a bifacial reduction sequence. A different set of debitage
types may characterize unifacial artifact production (in-
cluding bladelets) and such other specialized knapping as
the use of bipolar techniques.
Another dimension of debitage analysis involves at-
tempts to infer various techniques of flake removal. The
following categories were used to analyze the Harness
Mound flake debitage according to removal technique: 1 )
hard hammer percussion, 2) soft hammer (“billet”) per-
cussion, and 3) indirect percussion (“punch”) or pressure.
Stone hammers were probably used for hard hammer
percussion, deer or elk antler beam segments for soft
hammer percussion, and antler tines for pressure or indi-
rect percussion. These techniques are recognizd among
waste flakes mainly on the basis of shape and platform
configurations. Shape is influenced in part by the force
used to detach flakes while platform remnants on flakes
should correspond in size and shape with the contacted
portion of the percussion or the pressure tool used for
flake detachment.
Hard hammer percussion flakes characteristically have
a pronounced bulb of percussion and a relatively large
platform remnant. They are usually thick and may either
hinge or feather at their terminations. The incidence of
hinging is greater than with other percussion techniques.
Hard hammer percussion is the predominant technique of
flake removal in the initial stages of tool manufacture.
As the blank or preform was reduced, a stage would
have been reached where continued use of hard hammer
likely would have resulted in breakage from overly stress-
ful shock. At this point, the shift to a soft hammer for
further thinning would be expected.
Soft hammer percussion flakes generally have a less
pronounced bulb of percussion and a platform remnant
that is narrower (as measured from the dorsal face to the
1983
EDWIN HARNESS MOUND
75
ventral face) than those detached by hard hammer per-
cussion. Platforms are often of an elongated lenticular
shape but may be recurvate from the previous detachment
of a flake from the dorsal face at one end of the platform
remnant. Soft hammer percussion flakes are characteris-
tically thinner than flakes detached by a hard hammer
and usually feather at their terminations.
No attempt was made to distinguish indirect percussion
from pressure as flake detachment techniques. In both
cases, a tool such as the tip of an antler tine would have
been used to apply force on a point of the core platform
rather than an area. Thus, flakes removed by indirect per-
cussion and pressure are recognized by the presence of a
small circular to oval platform remnant in addition to a
bulb of applied force that is only subtly swollen. Such
flakes are characteristically thin and have feathered
terminations.
When correlated, flake type and technique of flake de-
tachment determinations provide a basis for inferring the
points in the reduction sequence at which shifts occurred
from hard to soft hammer percussion, and, if employed,
to indirect percussion or pressure. This is one way of de-
scribing a lithic waste flake assemblage. More impor-
tantly, this method monitors the results of prehistoric cul-
tural conditioning with the attendant potential of reveal-
ing ethnically diagnostic similarities and differences from
one assemblage to the next.
Raw Materials
Flint raw material identifications were attempted for
all four collections. At least 9 varieties are present among
the excavated artifacts and debitage from the Harness
Mound, while 1 5 were identified in the larger sample that
includes Harness’s surface collected material. Figure 9.1
identifies the outcrop locations of these varieties in rela-
tion to the Harness site complex.
Fig. 9. 1. Areas of outcrop of flint varieties present at Harness. ▲ Edwin Harness Mound (33Ro22);
1-Columbus; 2-Delaware; 3-Prout; 4-Plum Run; 5-Boggs; 6-Flint Ridge; 7-Upper Mercer; 8-Za-
leski; 9-Brassfield; IO-Cedarville-Guelph; 1 1— Bisher; 12-Brush Creek; 13-Harrison County;
I4-Kanawha Black.
76
KENT D. VICKERY
No. 39
Geological documentation of all Ohio varieties is pro-
vided by Stout and Schoenlaub (1945); archaeological
occurrences of two or more varieties are discussed by
Fowke (1894, 1902, 1928), Holmes (1919), Prufer and
Baby ( 1963), Converse (1972), and Morton and Carskad-
den (1972). Each is summarized here.
Flint Ridge. This well-known variety is described by
Smith (1885), Moorehead (1892:30-48), Mills (1921),
Crawford (1967), and Patterson (1979), among others. It
is variegated and fine grained with a vitreous luster.
Delaware. Delaware flint is tan to dark grayish brown
with sparsely scattered, tiny ostracod inclusions that are
white in the non-heat altered state. It is described by Con-
verse (1972:37).
Brush Creek. Described by Carskadden and Donald-
son (1973), Brush Creek flint is tan to light gray to dark
brown mottled with diagnostic small orange spots on its
cortex.
Zaleski. Zaleski would have been the closest of the
bedded varieties available to the Harness site inhabitants.
This material is brownish black to black and fine grained
with a vitreous luster. Kramer (1953) and Morton and
Carskadden (1972) discuss its characteristics and prehis-
toric utilization.
Upper Mercer. Carskadden (1971) and Morton and
Carskadden (1972) describe this flint variety, which
ranges from light to dark gray and from bluish black to
black with white inclusions and thin streaks. The darker
material is often fine grained with a vitreous luster.
Boggs. Boggs flint is described by Morton and Car-
skadden (1972). It is medium-dark gray to black, slightly
grainy in texture with a rather dull luster, and may have
numerous white fossil (commonly fragmented crinoid
stem), quartz, and pyrite inclusions. An orangish, ferru-
ginous patination is distinctive. This variety may be un-
derrepresented because of the non-availability of com-
parative samples when the Harness surface collections
were examined. The excavated collection from the Har-
ness Mound was reexamined, however, for Boggs chert.
Knife River. Clayton. Bickley, and Stone (1970) de-
scribe this fine-grained, honey-colored flint, and Braun,
Griffin, and Titterington ( 1 982:65—89) summarize mid-
western occurrences in Middle Woodland contexts. Knife
River flint has a color range from light to rather dark, a
diagnostic thin white cortex, and relatively large opaque
inclusions in a translucent matrix. That the presence of
Knife River represents long-distance acquisition, proba-
bly from west-central North Dakota, is noteworthy.
Harrison County. Harrison County flint (also known
as “Indiana hornstone”) from extreme south-central In-
diana also represents long-distance acquisition. Collett
(1878:421-423), Shaver et al. (1970:146), Lilly (1937:
101-104), Guernsey (1937), Seeman (1975), and Tomak
(1980:110; 1982:37-38) are among those who document
the archaeological utilization or geological context of this
well-known flint variety. It formed in nodules and typi-
cally ranges from dark bluish gray at the center to light
gray near the cortex. Harrison County flint is homogene-
ous and fine grained, slightly translucent at the edges of
thin flakes, and quickly weathers to a uniformly light to
medium blue-gray.
Cedarville-Guelph. Stout and Schoenlaub ( 1945:20-21)
document geological occurrences of this flint variety, and
Converse (1972:36-37) mentions its archaeological utili-
zation. It is grainy, pinkish tan or light gray, and has nu-
merous darker gray spots throughout the matrix which
are diagnostic for the variety among those identified in
the analyzed sample.
Columbus. Stout and Schoenlaub (1945:21-24) and
Prufer and Baby (1963:44) discuss this flint variety. It is
light and dark brown mottled and rather coarse grained.
Prout. Comparative samples of Prout chert upon
which identifications were made were collected from out-
crop on the Plum Brook NASA base near Sandusky,
Ohio. They are fossiliferous with small pyrite inclusions
and range from cream to tan and from light to dark gray.
Occurrences of chert in the Prout limestone member doc-
umented by Stauffer (1909, 1916) and Stumm (1942) in
this area probably refer to this variety. James L. Murphy
of Ohio State University (letter to K. D. Vickery, April 22,
1982) believes that the chert from Perkins Township in
Erie County mentioned by Stout and Schoenlaub (1945:
31) and the “Pipe Creek” chert mentioned by Stothersand
Rutter (1978) are also of the Prout variety.
Brassfield. Hastings (1969) gives a description and
summary of the prehistoric utilization of Brassfield chert.
It is white to flesh pink and is abundantly fossiliferous.
Bisher. Bisher flint is grainy and homogeneous, tan to
light purplish, and occasionally banded. It is described by
Hastings (1969).
Plum Run. The prehistoric use of Plum Run flint as
represented by quarries is discussed by Murphy and
Blank (1970); Converse (1972:38) provides a description.
Fresh comparative samples are fine grained and light to
dark gray and blue mottled, while weathered specimens
are tan, orange, red, brown, and green mottled with
streaks and patches of white-light gray inclusions.
Kanawha Black. The source of Kanawha Black flint is
central and southwestern West Virginia. It is bluish black
to black, non-vitreous, homogeneous, and very grainy.
White (1903:328-332, 1908:487-488), Krebs (1914:255-
266, 643-644), and Price (1921) document geological oc-
currences, Olafson (1964, 1972) archaeological occur-
rences.
Hastings (1971) discusses the use of pebble raw mate-
rial, which is ubiquitous in distribution.
Results of Analysis and Interpretations
Presented below is the distribution of thermally altered
1983
EDWIN HARNESS MOUND
77
specimens according to debitage type in the sample of 91
pieces of debitage from Harness Mound:
Debitage Type
Heat
altered
Heat
damaged Total
Core fragments
2
0
2
Primary decortication flakes
3
2
5
Secondary decortication flakes
2
2
4
Primary flakes
2
2
4
Thinning flakes
0
2
2
Uncertain
1
4
5
Total
10
12
22
Twenty-two specimens (24% of the total debitage) are
heat altered or damaged, which exceeds the 20% figure
that Collins and Fenwick (1974: 143) represent as “larger
than would be expected by chance” in their search for heat
treating technologies among the chipped stone assem-
blages of 12 archaeological sites in Kentucky. More than
half of the heated specimens from the Harness Mound,
however, were so damaged that the functional effective-
ness of the flakes (or the knapping suitability of the cores
from which they were struck) would have been impaired.
These specimens may simply represent post-detachment
damage unrelated to intentional heat treatment. None of
the 10 heat altered pieces of debitage exhibits the fre-
quently subtle changes associated with properly heat
treated material. Furthermore, heat altered or damaged
specimens occur throughout the reduction sequence
rather than being clustered at any particular stage. These
observations suggest that intentional heat treating of flint
and chert was not incorporated into the bifacial reduction
sequence. By contrast, the bladelet industry apparently
did feature the heat pretreatment of cores (Greber, Davis
and DuFresne 1981:513-514).
Table 9. 1 gives the proveniences of the Harness Mound
debitage. The frequency distributions of debitage and
flake types are presented in Tables 9.2 and 9.3. Among the
flake debitage, 14 (19.2%) are either utilized or retouched
(Table 9.2). One core fragment exhibits a utilized edge
(Table 9.3). This indicates a degree of chert resource con-
servation on the part of the people who produced the
sample. The ratio of retouched to utilized to non-utilized
flakes is 1 : 1.8: 14.6.
Among the flake debitage, it is interesting to note an
increase in the proportion of utilized and retouched
flakes from the early to the late stages of the reduction
sequence: 1 0% of the primary decortication flakes are uti-
lized or retouched; of the secondary decortication flakes,
18% are utilized or retouched, and the percentages of uti-
lized and retouched primary flakes and thinning flakes are
50% and 60%, respectively (Table 9.2). If the sample is
representative, selection of relatively thin flakes for spe-
TABLE 9.1
Provenience of Debitage from Edwin Harness Mound
Fragments
Checked
Pebbles
Primary
Decort.
Flakes
Secondary
Decort.
Flakes
Primary
Flakes
Thinning
Flakes
Uncertain
Total
1 I
1 1 1
1 1 1
4
1
5
15
1 1
1
2
3
1 1
1 1
1 1
1 3 2
14
3
2
1
5
36
5
13
29
1 1
10
5
18
91
5.5
14.3
31.9
12.1
11.0
5.5
19.8
100.1
78
KENT D. VICKERY
No. 39
TABLE 9.2
Correlation of Flake Type and Technique of Flake Removal, Edwin Harness Mound
Technique of
'"^flake Removal
Flake Type
Hard
Hammer
Percussion
Soft
Hammer
Percussion
Indeterminate
Percussion
Indirect
Percussion
Pressure
Platform
missing
Uncertain
Total
No. %
Primary Decort.
T
10
1
1
15
2
29
U
1
1
39.7
R
2
2
Secondary Decort.
T
4
7
11
U
1
1
15.1
R
1
1
Primary
T
1
2
7
10
U
1
2
2
5
13.7
R
Thinning
T
4
1
5
U
1
1
6.8
R
2
2
Uncertain
T
3
1
1
7
6
18
U
1
1
24.7
R
Total
T
18
7
2
1
37
8
73
100.0
U
3
3
3
9
12.3
R
2
3
5
6.8
%
24.7
9.6
2.7
1.4
50.7
11.0
100.1
T = Totals
U = Utilized
R = Retouched
cific or perhaps multiple tasks is suggested. Primary and
thinning flakes should have more acute edge angles and
sharper edges than decortication flakes. Hence, if the
sample reflects intentional selection of flake debitage for
tool use, marginal sharpness seems to have been a more
important consideration than strength. This in turn sug-
gests a greater association with cutting tasks than with
scraping tasks, which parallels the apparent utilization of
bladelets in the Edwin Harness Mound and vicinity.
Technique of flake removal is correlated with flake type
in Table 9.2 and with raw material in Table 9.4. The total
number of flakes in the sample is quite small and the often
diagnostic platforms of half are missing. Nevertheless, a
few observations can be made on the data bearing on dif-
ferential tool use in the knapping sequence.
Decortication waste flakes are dominated by hard
hammer percussion, with a noticeable reduction in the
number of hard hammer percussion detached flakes from
the primary to the secondary decortication stages. Only
one primary flake and no thinning flakes in the sample
were detached by this technique. There is a corresponding
increase in the use of soft hammer percussion from the
stage of primary flake detachment to the thinning stage.
Only one soft hammer percussion detached flake is rep-
resented in the early stages of decortication. Thus, hard
hammers were apparently employed in the initial stages of
knapping at least cortical raw material and the shift to a
soft hammer occurred when the mass of the blank or pre-
form was being reduced by the removal of primary flakes.
Indirect percussion or pressure is represented by a single
flake of indeterminate type. However, this flake removal
technique is often employed in the final stages of the re-
duction sequence when sharpening flakes are detached.
“Soft” cortex, such as is commonly present on chert nod-
ules that have not been transported far from their outcrop
areas by natural agencies, was not observed in the debit-
age sample. Thus, the presence of cortex on the dorsal
faces of flakes and on other debitage is presumed to repre-
sent pebble raw material.
In the debitage sample, a reduction in the proportion of
specimens with remnant pebble cortex is observable from
the checked pebble stage (100.0% pebble material) to the
primary decortication stage (79.3%) to the secondary de-
cortication stage (72.7%) to the primary flake stage
(10.0%) to the final thinning stage (0.0%). Seemingly con-
tradictory is the lack of a 100% representation of pebble
material in the decortication stages. Decortication , how-
ever, refers not only to the detachment of pebble cortex
but to the removal of portions of very thin patinas result-
ing from frost wedging as well. Such frost-cracked mate-
rial may have occurred as bedded chert in outcrop loca-
tions that served as prehistoric quarries. Otherwise, the
1983
EDWIN HARNESS MOUND
79
TABLE 9.3
Correlation of Debitage Type and Flint/Chert Variety, Edwin Harness Mound
Flint /
Chert
Variety
Debitage
Type \
Delaware
Brush Creek
Flint Ridge
Cedarville- Guelph
Columbus
Brassfield
Upper Mercer
or Boggs
Local Pebble Chert
Unknown
Total
No. %
Core Fragments
T
2
1
1
1
5
U
1
1
5.5
R
Checked Pebbles
T
3
1
2
1
6
13
U
14.3
R
Primary Decort. Flakes
T
7
2
1
1
1
14
3
29
U
1
1
31.9
R
1
1
2
Secondary Decort. Flakes
T
2
2
3
3
1
11
U
1
1
12.1
R
1
1
Primary Flakes
T
2
1
2
1
1
3
10
U
1
1
2
1
5
11.0
R
Thinning
T
2
2
1
5
U
1
1
5.5
R
1
1
2
Uncertain
T
1
2
5
5
5
18
U
1
1
19.8
R
Total
T
17
10
10
4
3
2
1
30
14
91
100.1
U
1
2
3
1
1
1
1
10
11.0
R
2
2
1
5
5.5
%
18.7
11.0
11.0
4.4
3.3
2.2
1.1
33.0
15.4
100.1
T = Totals
U = Utilized
R = Retouched
debitage assemblage reflects a bifacial artifact industry
that was dominated by pebble raw material. This is evi-
dent in the distribution of pebble material throughout the
knapping sequence and in reduced numbers of flakes
from the beginning to the end of the sequence, suggesting
that not a great deal of bifacial thinning occurred beyond
the decortication stage. Furthermore, no sharpening
flakes were recognized in the analyzed sample even
though feature contents were consistently subjected to
flotation appropriate to their recovery from a soil matrix.
Their absence (or scarcity) suggests either that this ad-
vanced stage of artifact manufacture occurred at one or
more off-site loci or that bifacial tools were considered
finished products after the removal of thinning flakes.
With this possible exception, the debitage reflects a rather
homogeneous assemblage in that the same raw material
apparently passed through the entire reduction sequence
with neither significant inputs of additional raw material
at any one stage in this sequence nor removal of blanks or
preforms on their way to becoming finished artifacts.
Flint/ chert raw material excavated from the Edwin
Harness Mound is correlated with debitage type in Table
9.3 and with technique of flake removal in Table 9.4. Ta-
ble 9.5 presents the frequency distribution of identified
and unidentifiable flint varieties among the four analyzed
collections, including questionable raw material and cor-
tex identifications. Items of Flint Ridge flint were exam-
ined for the presence or absence of cortex for the Harness
Mound sample only. Therefore, Flint Ridge pebble is un-
derrepresented in Table 9.5. At least one specimen in Rob-
ert Harness’s surface collection was observed with pebble
cortex, but none were present in the excavated Harness
Mound collection.
One striking feature of flint raw material utilization is
80
KENT D. VICKERY
No. 39
TABLE 9.4
Correlation of Technique of Flake Removal and Flint/ Chert Variety, Edwin Harness Mound
Flint /Chert
Variety
Technique
of Flake
Removal
Delaware
Brush Creek
Flint Ridge
Cedarville- Guelph
Columbus
Upper Mercer
or Boggs
Local Pebble Chert
Unknown
Total
No. %
Hard Hammer Percussion
T
3
2
1
1
6
5
18
U
2
1
3
24.7
R
Soft Hammer Percussion
T
1
2
2
1
1
7
U
1
1
1
3
9.6
R
1
1
2
Indeterminate Percussion
T
1
1
2
U
2.7
R
Indirect Percussion/ Pressure
T
1
1
U
1.4
R
Platform missing
T
7
5
6
1
2
9
7
37
U
1
2
3
50.7
R
1
1
1
3
Uncertain
T
I
7
8
U
11.0
R
Total
T
12
9
10
2
3
1
23
13
73
100.1
U
1
2
3
1
1
1
9
12.3
R
2
2
1
5
6.8
%
16.4
12.3
13.7
2.7
4.1
1.4
31.5
17.8
99.9
T = Totals
U = Utilized
R = Retouched
the dominance of Flint Ridge in all but the Harness
Mound collection. Flint Ridge flint accounts for at least
93% of Harness’s surface bladelets and bladelet cores. For
Sites 18 and 25, 81% and 89% of the items, respectively,
are of Flint Ridge flint.
This high-grade material also accounts for half of the
excavated artifacts from the Harness Mound; all (17
bladelets and 1 bladelet core) are related to the bladelet in-
dustry. Only 2 Flint Ridge flakes (that are not necessarily
associated with the bladelet industry) were recovered
from the Harness Mound. Seven of the remaining 8 flakes
are virtually identical in appearance. They may have been
struck from the same core, which seems also to have been
the parent core for 4 of the 23 bladelets and bladelet frag-
ments recovered. At least 6 of these 7 were from the same
feature (Feature 30). The single primary decortication
flake of Flint Ridge probably represents the acquisition of
bedded material without any (or with very little) knapping
prior to transport. Two thinning and primary flakes each
are not necessarily part of a bifacial reduction sequence;
rather, they may have resulted from bladelet core prepa-
ration of a suitable preform. Five Flint Ridge flakes, “un-
certain” as to flake type, may represent snapped frag-
ments from attempted bladelet removal or flake segments
detached in an effort to reshape one or more cores.
The association of Flint Ridge raw material with blade-
let production is also evident, although not quantified,
in the Site 1 8 and Site 25 collections. It is possible that the
remarkably high proportion of Flint Ridge flint (and/or
items related to the bladelet industry) in the surface mate-
rial is due to selective collecting. Utilitarian artifacts and
associated debitage of non-Flint Ridge materials are
present, however, in all three collections. It therefore
seems unlikely that selective collecting, had it occurred,
significantly inflated the proportional representation of
Flint Ridge.
The possibility exists that Flint Ridge flint occurred
naturally in the vicinity of the site complex, having been
transported there by natural agencies. As noted, at least
one specimen in the Harness surface collection is indeed
1983
EDWIN HARNESS MOUND
81
pebble material. James L. Murphy (letter to K. D.
Vickery, April 22, 1982) cautions that some flint of the
Vanport limestone member in Jackson County to the
south and east of the Harness site complex is high-grade
and resembles material from the famous Flint Ridge de-
posits in Licking County and vicinity, to the north and
east of Chillicothe. Being located in the ancient Teays val-
ley, this more southern variety may have been transported
naturally into the area. If the flint being exploited at Har-
ness had been carried to the area in this manner, however,
one would expect to find at least some low-grade, weath-
ered material, or at least a range of variation in quality.
Because nearly all of the flint is high-grade, its source area
for at least most of the Flint Ridge flint from the Harness
site complex was probably the ridge proper or its envi-
rons, and it was likely transported by human agency.
Of interest in the investigation of flint raw material
procurement strategies is the representation of varieties
that would have been available locally as redeposited
pebble material. This problem was approached in three
ways: 1) each archaeological specimen was tabulated ac-
cording to whether or not pebble cortex was present; 2)
pebbles collected from a gravel bar along the Scioto River
close to the Harness Mound were examined for identifia-
ble varieties; and 3) non-artifactual chert pebbles from the
Harness Mound excavation in broken and unbroken
conditions were identified according to variety.
Table 9.5 indicates the incidence of pebble chert in the
archaeological collections. This form of raw material
would have been available locally, and most or all of the
pebble specimens were likely collected in the immediate
vicinity of the Harness site complex. The present analysis
would not have detected the acquisition of pebble raw
material from distant locales, if such procurement had in-
deed occurred. Debitage and artifacts lacking cortical
remnants are ambiguous with respect to context of acqui-
sition (in situ or redeposited). The lack of cortex may
simply represent decorticated but locally acquired pebble
chert rather than “exotic” raw material obtained directly
or indirectly from outcrop locations at varying distances
from the site complex. No more than 28% of the unidenti-
fied flint retains pebble cortex, suggesting at least some
decortication at the site(s) of acquisition.
The impression remains that much unidentified “Local
Pebble” and “Unknown” chert is of local origin, having
been dislodged from distant outcrop areas to the north,
east, and south and then transported by glaciers and an-
cient and modern drainage systems. Such pebble chert
presently occurs in gravel bars along rivers and in Wis-
consin and Illinoian till and outwash in the site complex
vicinity.
Included with unidentified flint are 6 flakes with a thin,
smooth, white-tan cortex and 10 decorticated flakes of
identical raw material from Site 25 that is dark brown,
homogeneous, very fine grained and occasionally pitted.
A bladelet core in one of Harness’s surface collections
may be of the same Hint variety. Although direct compar-
isons were not made, this high-grade material seems to be
the same as several bladelets and debitage apparently as-
sociated with bladelet production excavated from the
Hopewell site of Mound City in the Chillicothe vicinity
and present among surface collected material in Clermont
and Hamilton counties in southwestern Ohio. One speci-
men from a workshop near the Turner site in Hamilton
County was thin sectioned by Timothy S. Dalbey, who
identified it with the Delaware Formation based in part
on similarities between its photomicrograph and that
from the Kuenzli Quarry at Delaware illustrated by Stout
and Schoenlaub (1945: PI. III).
The possible link between three “classic” Hopewell sites
through the co-occurrence of this distinctive flint variety
merits further investigation, particularly in view of its ap-
parently exclusive use for bladelet production. If the Del-
aware identification is correct, it is not necessarily exotic
to the area. It is represented as unidentified because geo-
logically documented comparative samples are not cur-
rently available.
Also from Site 25 are one pebble flake and three flakes
without pebble cortex of identical, high-grade but uni-
dentified material that is translucent and variegated (with
reddish tinges as part of the coloration). These are likely
exotic to at least the region from which most of the Har-
ness site complex flint/chert was acquired, as is at least
one flake in the excavated Harness Mound collection.
A sample of 100 chert pebbles was randomly selected
from among several hundred collected at a gravel bar near
the Harness Mound during the summer of 1976. The re-
sults of identifying the flint / chert varieties represented are
given in Table 9.5, which shows Delaware present in
abundance (46%), followed by Cedarville-Guelph (8%),
Bisher (3%), and Columbus, Brassfield, and Upper Mer-
cer (l%each). Forty percent were unidentifiable. Identifi-
cations of non-artifactual chert pebbles excavated from
the Harness Mound are also presented in Table 9.5.
Apart from Flint Ridge and unidentifiable varieties,
the most abundantly represented raw material is Dela-
ware. Delaware dominates the excavated debitage and
the non-bladelet artifacts and is prominently repre-
sented among the debitage from Sites 1 8 and 25. This sug-
gests that it was the most important raw material ex-
ploited for tool production unassociated with the bladelet
industry, although small amounts were exploited for the
latter.
Delaware, Columbus, and Cedarville-Guelph all out-
crop in areas presently drained by the Scioto River and its
tributaries north and northwest of the Harness site com-
plex. It is likely that they were dislodged by the river and
carried southward where they were then available locally
as pebble material in the site vicinity. All three varieties
were present in the pebbles collected along the Scioto
River, with Delaware accounting for almost half of the
sample. The non-cultural pebbles recovered from the
TABLE 9.5
Frequency Distribution of Flint /Chert Varieties for Cultural Collection and Comparative Local Scioto River Gravel Pebbles
ED WIN HA RNESS
MOUND
Artifacts Debitage Pebbles
Artifacts
ROBERT HARNESS
Surface Collection
Debitage
Bladelets sites It 1 8
& Cores Q( & tt25
et
TOTA LS FOR CUE TURA L
MATERIALS
% Total
non-flint
Subtotals Q* Total Ridge
Scioto
River
Gravels
(Pebbles)
Flint Ridge
18
10
7
1,103
2
1,151
2,289
2 2,291**
Flint Ridge (Pebble)
1
1
1
Delaware
2
3
1
18
2
7
i
31
3 34
7.8
Delaware (Pebble)
4
14
8
1
1*
2*
27
3 30
6.8
46
Brush Creek
2
7
1
1
4
3
15
3 18
4.1
Brush Creek (Pebble)
1
3
2
2
8
8
1.8
Zaleski
1
1
1
1
4
7
1 8
1.8
Zaleski (Pebble)
1
4
1*
5
1 6
1.4
Upper Mercer
2
1
4
1
6
2 8
1.8
Upper Mercer (Pebble)
1
1*
1
1 2
0.5
1
Upper Mercer or Zaleski
1
2
2
5
5
1.1
Upper Mercer or Zaleski
(Pebble)
1
1
1
0.2
Boggs
5
5 5
1.1
Boggs (Pebble)
1
1 1
0.2
Upper Mercer or Boggs
1
1
1
0.2
Knife River
2
3
2
3
4 7
1.6
Harrison County
1
2
1
1
2
3 5
1.1
Cedarville-Guelph
1
1
2
2
0.5
Cedarville-Guelph (Pebble)
3
4
7
7
1.6
8
Columbus
1
1
1
0.2
Columbus (Pebble)
2
2
2
0.5
1
Prout (Pebble)
1
1
1
2
1 3
0.7
Brassfield (Pebble)
2
2
2
0.5
1
Bisher
1
1
1
2
1 3
0.7
Bisher (Pebble)
3
Plum Run (Pebble)
1
1
1
0.2
Kanawha Black (Pebble)
1
1
1
0.2
Unknown
4
14
8
36
111
173
173
39.5
Local Pebble
1
30
40
1
2
29
1*
103
1 104
23.7
40
Total
36
91
54
22
1,167
13
1,328
19
2,698
32 2,730
99.8
100
♦Probably pebble; flint identification definite
♦♦Flint Ridge variety forms 83.9% of total identified cultural objects
tQuestionable
1983
EDWIN HARNESS MOUND
83
Harness Mound excavation included both Delaware and
Cedarville-Guelph. In the excavated debitage sample, the
proportions of Delaware, Cedarville-Guelph, and Co-
lumbus that are demonstrably pebble-derived are 82.4%,
75.0%, and 66.7%, respectively. These data indicate that
predominantly pebble material was used for bifacial arti-
fact manufacture, although the presence of bladelets and
bladelet cores of Delaware in the surface collections may
represent direct acquisition from outcrop in order to ex-
ercise some selectivity regarding quality and perhaps size.
Three flint varieties outcrop to the east and south of the
Harness site complex, in the path of the Pleistocene Teays
River and its tributaries. Brush Creek and Zaleski in Ohio
and Kanawha Black in West Virginia were all subject to
displacement by the westward and — in the site vicinity
and immediately south of it — northward flowing Teays,
which undoubtedly carried at least these three varieties to
and through the area. The most important of these varie-
ties for chipped stone tool manufacture was apparently
Brush Creek. Like Delaware, Brush Creek was used for
limited bladelet production, as suggested from the occur-
rence of one Brush Creek bladelet each in the excavated
sample and in Harness’s surface collection. However, it is
more abundant in the form of debitage not necessarily
associated with bladelet production and ranks second to
Delaware among the identified varieties in the total sam-
ple. Ranking third is Zaleski, which is the closest source of
bedded flint to the Harness site complex.
None of these three flint varieties is represented in the
Scioto River pebble chert sample, but nearly half of the
Zaleski flint in the total sample is demonstrably pebble
material and the only specimen of Kanawha Black also
retains pebble cortex.
The presence of Brush Creek in the mound as unmodi-
fied pebbles indicates that it was naturally occurring and
locally available. It is represented by one checked pebble
in the excavated debitage sample, suggesting that it was
exploited in pebble form. Nevertheless, the lowest pro-
portion of demonstrably pebble material of all the locally
available varieties is Brush Creek. Only 30% of the exca-
vated Brush Creek debitage and 26% of this variety in the
total sample retains pebble cortex. This may represent the
transportation of pebble material to the mound and vicin-
ity in partially decorticated conditions, probably in con-
junction with at least some direct acquisition of tabular
material from the source area (perhaps destined for blade-
let production). If such were the case, the procurement
strategies for both Brush Creek and Delaware were very
similar.
Relatively close to the Harness site complex are source
areas of Bisher and Brassfield to the west and Upper
Mercer to the east. The presence of all of these varieties in
the Scioto River pebble chert sample suggests that they
were carried into the area naturally. The Harness surface
collection contains a bladelet core of pebble Upper
Mercer and a bladelet of questionably identified Upper
Mercer in addition to two bladelets of either Upper
Mercer or Zaleski. Thus, Upper Mercer is yet another
flint variety that was used for bladelet production in the
Harness Mound vicinity, although apparently in very
small quantities. Even though available in the local area
in pebble form, bedded material from Upper Mercer out-
crops would have been accessible to groups traveling to or
from Flint Ridge deposits yet farther north.
Also present in small quantities are Bisher and Brass-
field. One of the Harness surface collected bladelet cores
is probably Bisher, but no bladelets of this variety were
recognized. Bisher is noticeably grainier than such high-
quality varieties as Harrison County, Knife River, Flint
Ridge, and most Upper Mercer and Zaleski. Brassfield
has many fossil inclusions that would have deflected
shock waves traveling through a core, causing bladelets
and flakes to detach unpredictably. These may have been
prehistoric considerations in selecting predominantly
other varieties for the bladelet industry.
With the possible exception of one stemmed knife of
either Upper Mercer or Zaleski, neither Upper Mercer
nor Bisher is represented among the excavated artifacts
and debitage, although both are present in the Scioto
River pebble sample. Given their low density in this sam-
ple and the small quantity of excavated debitage, how-
ever, this lack of representation is probably due to sam-
pling error. By the same token, the occurrence of Bisher,
Upper Mercer, and Zaleski artifacts unrepresented by
debitage in the excavated collection may also be due to
sampling problems because the Harness surface material
contains debitage of each. Brassfield, Columbus, and
Upper Mercer or Boggs are represented by excavated deb-
itage and no artifacts, but again the small amounts pres-
ent suggest sampling error.
Outcropping far to the north and present in the sample
in such small quantities that intentional procurement
seems remote are Plum Run (in northeastern Ohio) and
Prout (in the extreme north-central part of the state). The
mechanism or mechanisms by which these varieties were
transported to south-central Ohio is unknown. Glacial
action is plausible in the case of Prout but cannot account
for the presence of Plum Run. Unless misidentified, this
item may have been brought to the Harness site complex
by visitors carrying indigenous items with them for trad-
ing along the way. This may also account for the presence
of one Plum Run specimen at nearby Mound City
(Vickery 1983).
Exotic to Ohio are Harrison County and Knife River
flint, both of which are present in the Site 25 collection
and, for Harrison County only, among Harness’s surface
collected bladelets and bladelet cores. With its source area
in extreme south-central Indiana, Harrison County flint
was available ca. 200 miles away, while Knife River Hint,
if obtained from deposits in west-central North Dakota,
was at least 1,100 miles distant from the Harness site
complex. Unless these varieties were brought here by vis-
84
KENT D. VICKERY
No. 39
iting groups, their occurrence in south-central Ohio re-
sulted from direct acquisition or long-distance trade. If
the former, they were apparently imported in small
amounts (perhaps each representing a single episode of
acquisition) and likely specifically for bladelet, cache
blade, and/or ceremonial spear point production.
Summary of Flint/ Chert Raw Material Utilization
In the vicinity of the Harness Mound, Flint Ridge flint
was the most heavily relied upon raw material for bladelet
production. Though some utilitarian items of Flint Ridge
were manufactured — as exemplified by projectile points,
a scraper, and unidentifiable biface fragments in the Site
18 and Site 25 surface collections — the main purpose of
its importation seems to have been to maintain a rela-
tively large supply of high quality raw material on hand
for bladelet production. Some of this material may have
been available locally, the Teays River perhaps having
transported it northward. Nearly all of the Flint Ridge
flint observed in the collections, however, was high-grade
material, which, in combination with its general lack of
pebble cortex, suggests that procurement was predomi-
nantly or exclusively from bedded deposits, probably to
the north and slightly east of the site complex. Its abun-
dance and quality suggest that procurement was both sys-
tematic and selective.
Supplementing Flint Ridge for bladelet production
were Harrison County and Knife River flint, along with
small quantities of flint varieties available locally or a rel-
atively short distance away (e.g., Delaware, Brush Creek,
Upper Mercer). It is likely that some Delaware and Brush
Creek were acquired from outcrop.
The bladelet industry is represented in the excavated
sample from the Harness Mound, but this collection
mainly reflects bifacial reduction associated with the man-
ufacture of utilitarian artifacts.
There is a general correspondence between the debit-
age, artifacts, and naturally occurring pebbles from the
Harness Mound and the Scioto River pebble chert sam-
ples. There is also a fairly even distribution of flint varie-
ties throughout the reduction sequence (as represented by
the various debitage types). This suggests that locally
available pebble chert was exploited for non-bladelet and
very limited bladelet manufacture.
The territory of systematic flint and chert resource ex-
ploitation was probably a linear, north-south one that as-
sumed the configuration of an ellipse roughly 150 miles
north-south by 80 miles east-west, with the Harness site
complex located near the west-central periphery (see Fig.
9. 1). In addition, trade arrangements and/or forays were
undertaken to acquire certain high-quality varieties from
greater distances away. Visiting groups bringing indige-
nous flint with them likely supplemented the raw material
acquired in this manner.
Acknowledgments
I wish to thank James F. Murphy of Ohio State Univer-
sity for his critical reading of the manuscript and helpful
suggestions. Any errors or omissions remain the respon-
sibility of the author.
References
Binford, Lewis R.
1963 A proposed attribute list for the description and clas-
sification of projectile points. In Miscellaneous stud-
ies in typology and classification, Anta M. White, Lew-
is R. Binford, and Mark L. Papworth, pp. 193-221.
University of Michigan, Museum of Anthropology,
Anthropological Papers No. 19, Ann Arbor.
Braun, David P., James B. Griffin, and Paul F. Titterington
1982 The Snyders mounds and five other mound groups in
Calhoun County, Illinois. University of Michigan,
Museum of Anthropology, Technical Reports No. 13;
Research Reports in Archaeology Contribution 8,
Ann Arbor.
Carskadden, Jeff
1971 Upper Mercer flint quarries in Muskingum County,
Ohio. Ohio Archaeologist 2 1 ( 1 ):3 1 5—3 1 6.
Carskadden, Jeff, and Gerald Donaldson
1973 Brush Creek flint quarrying in Perry and Morgan
counties, Ohio. Ohio Archaeologist 23( 1):20— 2 1 .
Clayton, Lee, W. B. Bickley, Jr., and W. J. Stone
1970 Knife River flint. Plains Anthropologist 15(50):
282-290.
Collett, John
1878 Geological report on Harrison and Crawford coun-
ties, Indiana, 1878. Eighth, Ninth, and Tenth Annual
Reports of the Geological Survey of Indiana, pp.
291-522, Indianapolis.
Collins, Michael B., and Jason M. Fenwick
1974 Heat treating of chert: methods of interpretation and
their application. Plains Anthropologist 19(64):
34-145.
Converse, Robert N.
1972 Flints used by Ohio’s prehistoric Indians. Ohio Ar-
chaeologist 22(2): 36-39.
Crabtree, Don E.
1972 An introduction to flintworking. Occasional Papers
of the Idaho State University Museum No. 28, Poca-
tello.
Crawford, Betty J.
1967 Famous Flint Ridge flint. Earth Science 20(3): 1 1 8 —
1 19.
Fowke, Gerald
1894 Material for aboriginal stone implements. The Ar-
chaeologist 2:328-335, Waterloo, Indiana.
1 902 Archaeological history of Ohio. Ohio State Archaeo-
logical and Historical Society, Columbus.
1928 Archaeological investigations — II. Bureau of Ameri-
can Ethnology, 44th Annual Report , pp. 399-540,
Washington, D.C.
Greber, N’omi, Richard S. Davis, and Ann S. DuFresne
1981 The micro component of the Ohio Hopewell lithic
1983
EDWIN HARNESS MOUND
85
technology: bladelets. Annals of the New York Acad-
emy of Sciences 376:489-528.
Guernsey, E. Y.
1937 Certain southern Indiana sources of lithic artifact
material. Proceedings of the Indiana Academy of
Science 46:47-52, Indianapolis.
Hastings, Jerry
1969 Brassfield and Bisher Hints: their use by prehistoric
Indians in Adams County, Ohio and adjacent areas.
Ohio Archaeologist 19( 1 ): 1 1 .
1971 Nodular chert and Hint: their use by prehistoric Indi-
ans in southern Ohio. Ohio Archaeologist 21(1):319.
Holmes, William H.
1919 Handbook of aboriginal American antiquities: part I,
introductory: the lithic industries. Bureau of Ameri-
can Ethnology, Bulletin 60, Washington, D.C.
Kramer, Leon
1953 Is there a jet black flint and may it be found in Ohio?
Ohio Archaeologist 3(4):29— 3 1 .
Krebs, Charles E.
1914 Kanawha County. West Virginia Geological Survey,
County Reports , Wheeling.
Lilly, Eli
1937 Prehistoric antiquities of Indiana. Indiana Historical
Society, Indianapolis.
Mills, William C.
1921 Flint Ridge. Ohio Archaeological and Historical Pub-
lications 30:91-161, Columbus.
Moorehead, Warren K.
1892 Primitive man in Ohio. G. P. Putnam’s Sons, New
York.
Morton, James, and Jeff Carskadden
1972 Aboriginal flint quarrying activities in the Muskin-
gum County area. Ohio Archaeologist 22(2): 15-21.
Murphy, James L., and John E. Blank
1970 The Plum Run flint quarries. Ohio Archaeologist
20(2): 198-199, 217.
Olafson, Sigfus
1964 West Virginia flints used in Ohio. Ohio Archaeologist
14(2):37— 39.
1972 Some sources of lithic materials in West Virginia.
West Virginia Archaeological Society, Newsletter
14(3):3— 9.
Patterson, L. W.
1979 Additional data on thermal alteration of Flint Ridge
materials. Ohio Archaeologist 29( 1 ):33— 34.
Price, W. Armstrong
1921 Chert deposits of West Virginia. In Nicolas County,
edited by David B. Reger, pp. 221-240. West Virginia
Geological Survey, County Reports , Wheeling.
Prufer, Olaf H., and Raymond S. Baby
1963 Paleo- Indians of Ohio. Ohio Historical Society, Co-
lumbus.
Seeman, Mark F.
1 975 The prehistoric chert quarries and workshops of Har-
rison County, Indiana. Indiana Archaeological Bul-
letin 1(3):47— 6 1 .
Shaver, Robert H., et al.
1970 Compendium of rock-unit stratigraphy in Indiana.
Indiana Department of Natural Resources, Geologi-
cal Survey, Bulletin 43, Bloomington.
Smith, Charles M.
1885 A sketch of Flint Ridge, Licking County, Ohio.
Smithsonian Institution, Annual Report for 1884 , pp.
851-873, Washington, D.C.
Stauffer, Clinton R.
1909 The Middle Devonian of Ohio. Geological Survey of
Ohio, Fourth Series, Bulletin 10, Columbus.
1916 The relationships of the Olentangy shale and asso-
ciated Devonian deposits of northern Ohio. Journal
of Geology 24(5):476-487.
Stothers, David M., and William Rutter
1978 Pipe Creek chert: newly discovered aboriginal quarry
source. Ohio Archaeologist 28(3): 13.
Stout, Wilbur, and R. A. Schoenlaub
1945 The occurrence of flint in Ohio. Ohio Department of
Natural Resources, Division of Geological Survey,
Fourth Series, Bulletin 46, Columbus.
Stumm, Erwin C.
1942 Fauna and stratigraphic relations of the Prout lime-
stone and Plum Brook shale of northern Ohio. Jour-
nal of Paleontology 1 6(5):549— 563.
Tomak, Curtis H.
1980 Schershel: a Late Archaic occupation in southern In-
diana with appended chert descriptions. Central States
Archaeological Journal 27(3): 104-1 1 1.
1982 Selected Indiana chert types. Ohio Archaeologist
32( 1 ):36— 39.
Vickery, Kent D.
1983 Harness and Mound City: a flint raw material com-
parison. Paper presented at the 48th Annual Meeting
of the Society for American Archaeology, Pittsburgh,
Pennsylvania, April 27-30.
Vickery, Kent D., and Lois J. Lambert
1977 Analysis of chert debitage sample from the Nashport
mound, 33Mul5, Muskingum County, Ohio. Appen-
dix F to Report of the 1975 Excavations at the Nash-
port mound (33Mul5), Dillon Lake, Ohio, N’omi
Greber, pp. 87-109. National Park Service report,
contracts cx-6000-5-0186 and cx-5000-6-0301.
White, Anta M.
1963 Analytic description of the chipped-stone industry
from Snyders site, Calhoun County, Illinois. In Mis-
cellaneous Studies in Typology and Classification,
edited by Anta M. White, Lewis R. Binford, and
MarkL. Papworth,pp. 1-70. University of Michigan,
Museum of Anthropology, Anthropological Papers
No. 19, Ann Arbor.
White, I. C.
1903 Coal Report, Vol. 2. West Virginia Geological Sur-
vey, Reports , Wheeling.
1908 Supplementary Coal Report, Vol. 2(A). West Virginia
Geological Survey, Reports , Wheeling.
Wilmsen, Edwin N.
1970 Lithic analysis and cultural inference: a Paleo-Indian
case. University of Arizona. Anthropological Papers
No. 16, University of Arizona Press, Tucson.
10 DISCUSSION AND CONCLUSIONS
Environment and Subsistence
The botanical, faunal, molluscan, and soils analyses are
all consistent. The environment of the Central Scioto
when the Harness Big House was used and first covered
over was very similar to that found in the area in the
eighteenth century, a forest cover with some relatively
open areas and pocket prairies. Within 3 km of the site on
Gordon’s (1966) map of vegetation at contact are bottom-
land hardwoods, oak-sugar maple, oak savannas, and
mixed mesophytic forests. Although the debris found at
the site reflects specialized and public activities rather
than everyday subsistence, it does contain evidence for
the use of the several parts of the relatively local
environments: scale fish, shell fish, and building materials
from the river and the hills; deer, raccoon, turkey and
other birds, and small mammals from the woodland and
edge areas.
The evidence of plant foods comes both from skeletal
analysis and botanical samples. Skeletal analysis (Bender,
Baerreis, and Steventon 1981) demonstrates that, on the
basis of present evidence, corn became a regular part of
the diet of humans and of deer after the Middle Woodland
time period in the sites tested from Illinois, Ohio, and
Wisconsin. Three individuals from Edwin Harness and
five from the large Seip Mound were included in the
study. The skeletal evidence from Edwin Harness is
consistent with the amount of corn recovered. The Zea
mays found at Harness was in contexts in which it can
easily be interpreted as being a special purpose plant in
the same sense as tobacco or other known ceremonially
important plants. This is demonstrated within Feature 60
where fragments of Zea mays were found and the ratio of
Carbon- 13 to Carbon- 12 in the bones of the individual
interred was -22.7%. This value of the ratio is interpreted
as indicating no significant amount of corn in the diet
(ibid.).
Wild fruits and seeds were also identified in the Harness
botanical samples analyzed. Of particular interest are the
small unidentifiable seeds which composed high percent-
ages of some of the notation samples. The unavoidably
biased information from Harness does fit the best present
estimate of the subsistence pattern of Scioto Hopewell
peoples: gathering, hunting, and gardening (Ford 1979).
Additional work needs to be done, particularly with
respect to the position within the economy of the locally
available plants which bear starchy seeds.
Relative Intrasite Chronology
One of the primary tasks of our excavation was to seek
stratigraphic or other physical evidence for rebuilding or
multiple building stages of the structure at the base of the
mound. All the evidence which we found supports a single
stage of both construction and use of the major structure.
This is to be distinguished from the possible use through
time of the knoll itself and the various building stages of
the mound which were placed over the building.
There are two somewhat different sequences of con-
struction which can be interpreted from the known data.
They center on the time lapse between the preparation
and use of the sub-main floor area. Feature 33, and its
underlying base. Feature 50. No building posts were
found that originated on Feature 33; it appears that only
activities which would result in few if any subfloor
remains took place on the surface of this mixed clay floor.
This surface was used, if at all, before the Big House
was completed. Large areas of relatively featureless
“Floor” which had been separately covered by primary
mounds were recorded by Shetrone at either end of
Hopewell Mound 25 (Shetrone 1926). Within the Seip
Earthworks in Locality 23, which is immediately west of
the area excavated by the Ohio Historical Society 1971-
1977 (Baby and Langlois 1979), we have uncovered a
plaza-like area which was also relatively featureless and
had been separately covered (Greber and D. Griffin
1982). There were undoubtably other such places in
central Ohio. Thus the existence and use of such a space is
not unusual.
Feature 33 underlies the East Section and extends
approximately 13 m north and 10 m west of this section.
This is an area of particular interest because of the
somewhat anomolous character of the burial population
associated with the corresponding area in the Seip Big
House. At present we cannot identify the nature of the
population associated with the East Section at Harness.
Hopefully further work will aid in deciding whether the
location of Feature 33 with respect to the East Section is
of some significance.
With respect to possible construction phases of the
major structure, we did not find any evidence which
would indicate that the surface of the knoll had been
cleared in stages. Feature 3C, which underlay the entire
structure, was the same everywhere it was found. It is
possible that evidence of differences existed originally in
the heavily disturbed areas, and that the area under
Features 50 and 33 had been cleared first, the remainder
later. However, in addition to the physical characteristics
of Feature 3C, there is a design feature which makes me
tend not to accept two separate clearings. The heavy
gravel outer mantle had been placed essentially at the
edges of Feature 3C, which would not have been visible
when the wall was placed. This and the repetition of the
pattern at Seip do appear to indicate there was some basic
preconceived design for the total complex.
1983
EDWIN HARNESS MOUND
87
Mills (1907:137) states that the posts were placed and
the floor built around them. We did not have the
opportunity to see the join of Feature 3, a puddled clay
floor, and pristine building posts. However, the character
of this main upper floor and the constructions of tomb
remains and outer posts which we did find is consistent
with Mills’s observation. I do think it would have been
possible to have constructed post holes through the softer
clays of Features 33 and 50.
In sum, the surface of the knoll could have been
partially cleared, a special purpose ceremonial area
constructed, used, and left. Later, additional clearing and
cleaning of the first area, placement of major structural
posts (perhaps beginning construction on the east), and
laying of the main floor of the complete house followed.
Or, what appears more likely, the complete knoll was
cleared, and a special area made, perhaps with materials
of special significance from previous ceremonies. When
the appropriate ceremonies were completed, the major
structural posts were placed, the main floor layers put
down, and the floor features constructed, all in a rela-
tively short span of time. In considering this second as
more likely, I am regarding the radiocarbon assay DIC-
662 as a product of old trees and/or chance.
The Pattern of the Big House
The basic plan of the Big House, three nearly rectan-
gular and one circular section, is in contrast with the
design of the earthwork which contains three sections that
are parts of circles and one section that is square. Out of
both designs come the deliberate use of 3, 4, 7, and 48 (4 X
12) construction elements. Within the Big House itself the
two larger sections, which are similar in outline, differ in
the stylistic implementation of the basic design. The two
smaller sections differ in basic outline. Color appears to
have been used to distinguish structural posts on perim-
eters or entranceways in the three smaller sections and the
middle hallways. In the North Section three structural
posts in the southeast corner and one immediately west
of the geometric center were also marked with red stains.
The non-structural posts were distributed differently
among the sections. The North and Middle Sections each
had an irregular line just west of the building proper.
Other small, apparently colored posts were immediately
north of the North Section. A cluster of small stained
posts was on the east side of the Middle Section, while a
cluster of medium and small stained posts was on and
immediately west of the center of the North Section. In
contrast, at the center of the Middle Section was a long,
burning fire. There is also a contrast between the usual
small scale (total content) of the deposits found in many
places on the floors within and about the Big House and
those which were found by earlier excavators about the
center of the North Section. Comparison of the placement
of these deposits with respect to the central focus posts
suggests an east-west division in the design of the center.
Thus, there are differences in scale, design, and activity
remains within the four sections. These differences indi-
cate a likely difference in style if not in actual content of
some of the activities which took place in each of the four
sections. I have concluded, based on my analysis of burial
attributes, that each section of the Big House was the
social space of a sub-group within the total society which
supported the Big House (Greber 1979). The data given
here is separate from that of individual burial attributes.
This data, on the design of the complex, is consistent with
my additional conclusion that each sub-group represented
in the Big House had separate social responsibilities
within the whole society.
The Pattern at Seip
The Harness building design shown in Figure 3.2 was
placed over the map of tombs and floor features found at
the base of the large Seip Mound (see Greber 1 979: Fig.
1A). It was necessary to rotate the Middle Section 90°.
The results of the superposition are given in Figure 10. 1. 1
consider this an excellent fit. This new estimated map
suggests a refinement of the original groupings which I
had used in studying the social sub-groupings associated
with this Seip mound. The section which corresponds to
the Harness East Section groups together a small number
of individuals who have stood out among the general
population associated with Lobe 1 (West Section) (Gre-
ber 1976:76). Using the ranksums calculated for the
individuals whose tombs were found on the floor of the
Seip Big House (Greber 1976:53, 237-253), I calculated
new median ranksums using now four separate units. The
small group of six individuals in the North Section of the
Seip Big House has the highest median ranksum but with
a large confidence interval (see Table 10.1). The field
notes of the Ohio Historical Society excavations at the
large Seip Mound stated very clearly that the primary
mound, which covered the largest section (Lobe 1), also
covered this small north extension (Shetrone et al.
1925-1928). If this were strictly true, then perhaps this
high ranking group has a special relationship with the
major group within the Lobe. Such a small group would
have a different social character from that of the larger
groups associated with the other sections of the Big
House.
A review of the descriptions of the non-perishable
artifacts found with the remains of these individuals
suggests that, as a group, these individuals may have been
associated with special rituals (Greber 1976:Individuals
37, 39, 44, 45, 46, 51). With the remains of Individual 39
were a miniature copper plate and earspools and four cop-
per covered stone buttons. Above the roof of the tomb
pieces of a small pottery human head effigy were found.
This tomb was unusual in construction. It also had been
covered by a primary with two gravel strata. The primary
RTH
^
O o
"O "O C
c c cd
g 2 o.
- -S £
_
3- • 3
00 ^ O
S 00 S
o. E o.
o
C V
£ x
\S c
3 C
tu £
T3
— w
C3
E T3
C
3
0J o
■ 2f S o
It o c
I
Harness Middle Structure has been rotated 90°.
1983
EDWIN HARNESS MOUND
89
TABLE 10.1
Values of Ranksum F for Seip Mound 1 (Pricer)
(Based on calculations in Greber 1976)
House
Section
No. of
Individuals
Median
Ranksum
97%
Confidence
Interval
Confidence
Interval
Length
Highest
Value
Lowest
Value
West (Lobe 1)
41
607
554 to 667
1 13.0
371.0
743.5
North (Lobe 1)
6
555
412 to 685
273.0
41 1.5
685.0
Middle (Lobe 2)
37
656
628 to 689.5
61.5
440.0
750.5
East (Lobe 3)
19
698
643 to 738.5
95.5
297.5
747.5
is described as covering two tombs, but which two is not
clear. If it is that which Shetrone and Greenman called
Grave 34, then the artifacts found with Individual 37 in
this grave complement those of Individual 39. A diminua-
tive copper crescent along with four pearl beads, one
copper hemisphere, and two copper objects, called “pos-
sible effigies of the praying mantis” by Shetrone (Shetrone
and Greenman 1931:407), were found. The latter may be
metal working tools. The multiple burial (Individuals 44,
45, 46) was covered by a large primary mound with two
sand strata, an unusual construction. Associated with
Individual 44 were two copper celts; with Individual 45,
approximately 20 pounds of galena, pearl beads, burned
fabric, a copper celt, and pulley-type stone earspools; with
Individual 46, nothing. Deposited with the bones of
Individual 5 1 were a small, “fine” ocean shell container, a
copper plate, earspools, and 50 fine pearl beads. Within
the charcoal platform on which the bones had been
deposited, three hollow 1-inch (2.54-cm) diameter copper
hemispheres were found. I suggest that the miniature
items, the copper hemispheres, the conch shell, and the
galena may have been primarily used in rituals.
The four sections of the Seip Big House may have, at
one time, been reflected in the design of the earthwork
walls at Seip. There are some variations in the design of
the five major earthwork complexes which I have con-
sidered as related: Seip and Baum in the Paint Creek
Valley, Frankfort on North Fork, and Works East and
Liberty in the main Scioto Valley. The most nearly
complete maps for all these come from Squier and Davis
( 1 848: Pis. XX, XXI). All five of these earthworks include
complete squares, if we assume the square at Works East
was intact before erosion by the lateral movement of the
Scioto. All the complexes also contain one large incom-
plete circle. The next smaller circle is relatively complete
except at Baum, which has an amorphous area joining
two partial circles. There is an additional smaller partial
circle at Works East and Liberty. An amorphous wall
joins the small circle and the square at Seip. There were
changes in building design and apparent social groupings
at Seip as seen in the structures under the large mound
and the smaller conjoined mound (Greber 1979, 1976).
Changes appear to have been made in the earthwork
walls. Perhaps these two changes are related. The order of
the building of the walls and the significance of the design
of the parts is currently being studied (Greber and
Jargiello 1982). Based on types of variation in designs, the
earthworks can be grouped by river valleys. On North
Fork the design contains a square and two nearly
complete circles; in the Paint Creek Valley each has a
square, two circular parts with less geometrically identi-
fiable joining walls; and in the Scioto Valley the designs
contain a square and three circular elements.
Absolute Chronology
The radiocarbon dates presented in Table 3.2 are
plotted in Figure 10.2 with comparative dates from the
Russell Brown Mounds, which are part of the Liberty
Earthwork Complex (Seeman and Soday 1980); from the
McGraw site, which is on the west bank of the Scioto 8.5
km north of Liberty (Prufer 1968; Prufer et al. 1965); and
from the Seip Earthwork Complex (Baby and Langlois
1979; Greber and D. Griffin 1982). Considering first the
dates from the base of the Edwin Harness Mound, we can
group several of these into reasonable stratigraphic units:
the four dates associated with the Big House itself
(Features 19, 30, and P. M. 32); those from areas outside
the house but on the main, heavy puddled clay floor
( Features 1 7, 3 1 , 62); and the single one from the northern
activity area (Feature 53A). Since the radiocarbon years
represent a statistical mean, a group average mean can be
calculated using the standard deviation associated with
each mean as its weighting factor. The weighted average
mean for the use of the Big House is 1641 ± 32
radiocarbon years b.p. (a.d. 309 ± 32). The outer areas
average to 1619 ± 35 radiocarbon years b.p. (a.d. 331 ±
35); and the single date for the north area is a.d. 450 ± 50.
There is of course no justification for computing such
averages if the features associated with the radiocarbon
assays are not judged to represent a reasonably discrete
single cultural time. If all activities on the heavy clay floor
are assumed to be culturally contemporaneous, the
corresponding weighted mean is 1632 ± 24 radiocarbon
years b.p. (a.d. 318 ± 24). It should be noted that the
90
N’OMI GREBER
No. 39
Fig. 10.2. Radiocarbon dates plotted as mean calendrical years with one and two stan-
dard deviations noted. (See Key to Fig. 10.2 next page.)
apparent increase in precision of this latter date is the
result of averaging dates judged to represent contempo-
raneous cultural events. If this assumption is accepted, the
odds are better than 2 to 1 that the actual date represented
by this weighted mean is between a.d. 294 and a.d. 342.
Additional dates will be added to the sample as dating
techniques and resources allow. This further work may
indicate whether or not the shift in weighted means for the
floor groupings is a coincidence. Based on the present
sample size the best estimate for the time interval which
encompassed the construction and use of the area at the
base of the Edwin Harness Mound is centered around
a.d. 300.
Unfortunately there are not enough dates to judge the
time which intervened between the use of the floor and the
final capping of the top of the mound. The dates obtained
from the outer strata placed against the lower edges of the
mound do appear to indicate that the site, as part of the
earthwork complex, was used for many years.
Seeman’s analysis of the available radiocarbon dates
from the Russell Brown Mounds (Seeman and Soday
1980) places the use of these mounds probably before
(Mounds 1 and 2) and after (Mound 3) the use of the main
Edwin Harness floor. The sample of dates from these
smaller mounds is limited, but the use of the earthwork
area over such a time span is quite reasonable. A number of
interesting questions are posed both by the dates obtained
from the outer features at Edwin Harness and by the type
of materials being placed against the mound. The dark,
rock laden soils in Features 69 and 69A do not correspond
to any other major mound strata found in the above four
mounds; however, they do appear to resemble feature fill
1983
EDWIN HARNESS MOUND
91
KEY
Figure 10.2
Site
Lab. No.
Provenience
Calendrical Year
Harness
a
D1C-662
Under Feature 3
200 b.c. ± 155
a'
DIC-662
Under Feature 3
30 b.c. ± 155
b
DIC-665
PM 32
a.d. 130± 70
c
DIC-663
Feature 19
a.d. 330 ± 65
d
DIC-664 rerun
Feature 30
a.d. 350 ± 65
e
DIC-664
Feature 30
a.d. 450 ± 60
f
DIC-802
Feature 3 1
a.d. 320 ± 70
g
DIC-661
Feature 17
a.d. 460 ± 65
h
DIC-860
Feature 53A
a.d. 450 ± 50
i
D1C-1187
Feature 62
a.d. 180 ± 50
J
D1C- 1635
Feature 56
A.D. 750 + 65
k
DIC-1 188
Feature 81
a.d. 810 ± 60
1
DIC-1 190
Feature 55
a.d. 840 ± 50
Russell Brown
m
UCLA-244B
Mound 1, Burned zone
140 b.c. ± 70
n
UCLA-244A
Mound 1, Feature 210
a.d. 200 ± 80
o
UCLA-245
Mound 2, Feature 73
a.d. 90 ± 90
P
UCLA-246C
Mound 3, Feature 13
a.d. 430 ± 90
q
UCLA-246B
Mound 3, Feature 13
a.d. 590 ± 70
r
UCLA-246A
Mound 3, Feature 13
a.d. 615 ± 70
McGraw
s
UCLA-685
Excavation unit B-l
230 b.c. ± 80
t
UCLA-679A
Excavation unit D-l
a.d. 140 ± 80
u
UCLA-679B
Excavation unit D-l
a.d. 190 ± 80
V
M-1558
Excavation unit D
a.d. 230 ± 140
w
UCLA-688
Excavation unit B-l
a.d. 280 ± 80
x
OWU-62
Excavation unit D-2
a.d. 435 ± 166
y
UCLA-679C
Excavation unit C-l
a.d. 440 ± 80
z
OWU-61
Excavation unit C-l
a.d. 481 ± 65
Seip
aa
UCLA-292
Mound 1 (Pricer)
a.d. 55 ± 100
bb
D1C- 1725
Locality 20, Feature 2
a.d. 350 ± 45
cc
DIC- 1724
Locality 20, Feature 4
a.d. 470 ± 55
dd
DAL-280
House 7, Feature 5
a.d. 90 ± 85
ee
DAL-116
Unit D, Midden Layer, Houses 1-3
a.d. 230 ± 80
ff
DIC-28I
House 4, Feature 7
a.d. 590 ± 105
gg
D1C-289
Unit F, External post. House 4
a.d. 350 ± 60
hh
DAL-281
House 5, Feature 3
a.d. 590 ± 105
ii
DAL-282
House 6, Feature 1 3
a.d. 1055 ± 1 10
from two of the smaller mounds. In Russell Brown
Mound 1, one or possibly two out of ten pit features
contained dark soil and burned rock, while there were five
out of seven pits with such fill in Mound 3. These soils
may represent a change in type or in intensity of activities
which were carried out in the earthwork area at this later
time.
Considering the overall archaeological evidence, Prufer
has placed the occupation of the McGraw site in the fifth
century (Prufer et al. 1965: 106). Accepting his interpreta-
tion of the site as a single cultural event, and using all the
dates except that on bone, which was an experimental
date (Prufer et al. 1965: 104), we can compute a weighted
mean for the site of a.d. 318 ± 32. This is well within the
range of dates from the Liberty Complex and close to the
estimate for the use of the main floor at Edwin Harness.
As is shown in Figure 1 0.2, the range of available dates
from the Seip Earthworks overlaps that of Liberty.
Unfortunately there are no series of dates from the same
context within Seip, so relationships among the various
parts of the complex are more difficult to evaluate. Two
dates were not available when Figure 10.2 was prepared.
The first date, a.d. 430 ± 70 (DIC-2471), was obtained
from charcoal left from a small fire built on the surface of
a plaza-like area in Seip Locality 23 (Greber and D.
Griffin 1982) which is immediately west of the Ohio
92
N’OMI GREBER
No. 39
Elistorical Society Houses 1-7 (Baby and Langlois 1979).
While it was burning, this fire was apparently covered
with pea gravels similar to those in Harness Feature 41.
These gravels were the first of at least two covers over the
entire plaza. A second date, a.d. 330 ± 40 (DIC-2473), has
been obtained from the floor of the Seip Big House from
charcoal mixed with the cremated bones in Shetrone and
Greenman Burial 32. The assay on bone from this same
feature is unfortunately unacceptable (a.d. 670 ± 55,
D1C-2472). The a.d. 330 mean value (likely range a.d. 290
to a.d. 370) is relatively close to the dates from the floor of
the Harness Big House. More dates are needed to
corroborate the series.
Two dates, a.d. 470± 55 (DIC-1724) and a.d. 350± 45
(DIC-1725) have been obtained from charcoal found in a
pit and in a midden-type deposit of burned rock, mica,
colored clays, and other materials in Locality 20 which is
just west of the earthwork wall extending south from the
small circle (Greber and D. Griffin 1982). Sherds which
have been identified as Connistee material (Bennie Keel,
Chief Consulting Archaeologist, U.S. Dept, of Interior,
personal communication 1980; Roy Dickens, University
of North Carolina, Chapel Hill; and Jefferson Chapman,
University of Tennessee, Knoxsville) were found in the
general vicinity of the dated charcoal.
In the dates from the Ohio Historical Society units, the
assays for Houses 4 and 5 do not follow the stratigraphy
as reported by Baby and Langlois (1979). There does
appear to be some time range in the use of the seven
structures but it is not yet possible to associate specific
events in various parts of the earthwork with each other.
The time span of use does overlap with that of the Liberty
Works; however, more dates, hopefully from a variety of
absolute dating techniques, are needed. There is a growing
data base of absolute dates which demonstrates, consis-
tent with the archaeological evidence itself, that these
earthworks are indeed complexes of sites. No single
chronological date can be used to represent “Seip” or
“Harness”; therefore, discussions of intersite relationships
should specify the specific sites within the complex which
are pertinent to the points being discussed.
Epilogue
The Harness Big House was a place full of symbols. We
can identify at least some of these symbols; however, we
may not be able to interpret their intended meaning. It is
very tempting to interpret the data we have found using
details from ethnographies of known Eastern North
American peoples who are separated by more than a
thousand years from Hopewell peoples. In such interpre-
tations we are looking for elements which are basic parts
of an Eastern Woodlands culture and which could have
stability through time; that is, we are looking for a proto-
culture as a linguist seeks a proto-language. Yet we wish
this exercise to yield something that is not so general that
it is uninformative. This search must be done with addi-
tional cautions, since, as is well documented, the meanings
of a symbol vary not only through time, but within the
same time and even within a given culture at a given time.
What we can be confident of is the existence of a symbol
and a class to which the symbol belongs. The recent
salvage work at Edwin Harness has added to our
knowledge of Hopewell symbols. The classes we have
found include numbers, directions, colors, shapes, opposi-
tion or binary contrasts, special trees and plants, and
special uses of fire or smoke. These symbols reflect a
system of thought and a way of life. One can find
examples of the meaning of each of these within historic
Eastern Woodlands peoples in their range of activities;
gathering, hunting, gardening, curing, giving birth, giving
names, marrying, praying, trading, achieving great deeds,
and dying. They are part of the oral traditions which
explain the origins of the world and of important cultural
elements (e.g., Callendar 1978; Swanton 1946).
Structural remains which have been found at the base
of Hopewell mounds have at times been called “Charnel
Houses.” I would consider the use of such a name for the
Big House inappropriate. It should not be assumed that
all evidence of activities found in such contexts must be
interpreted as funeral. Even within modern cultures there
are examples of the juxtaposition or superposition of
crypts and/or tombs with areas which are used for a
variety of civic, ceremonial, and religious purposes, as
colonial churches which have burials below the main
floor or in adjacent yards. Describing only the funeral
aspects can be misleading and then centrality of the Big
House within the life of these Hopewell people is lost. We
have not yet been fortunate enough to find the house of a
family; at least we have not recognized any structure as
such. I expect that such a house will lack much of the
symbolic detail which can be found in a major civic-
ceremonial center because it is the center which is the
appropriate place to show and care for the symbolic life of
a people.
References
Baby, Raymond S., and Suzanne M. Langlois
1979 Seip Mound State Memorial: nonmortuary aspects of
Hopewell. In Hopewell archaeology: the Chillicothe
conference , edited by David S. Brose and N’omi
Greber, pp. 16-18. Kent State University Press, Kent,
Ohio.
Bender, Margaret M., David A. Baerreis, and Raymond L.
Steventon
1981 Further light on carbon isotopes and Hopewell agri-
culture. American Antiquity 46(2):346-353.
Callender, Charles
1978 Shawnee. In Handbook of North American Indians
(Vol. 15), Northeast. Smithsonian Institution, Wash-
ington, D.C.
1983
EDWIN HARNESS MOUND
93
Ford, Richard I.
1979 Gathering and gardening: trends and consequences of
subsistence strategies. In Hopewell archaeology: the
Chillicothe conference, edited by David S. Brose and
N’omi Greber, pp. 234-238. Kent State University
Press, Kent, Ohio.
Gordon, Robert B.
1966 Natural vegetation of Ohio at the time of the earliest
land surveys. The Ohio Biological Survey, Columbus,
Ohio.
Greber, N’omi
1976 Within Ohio Hopewell: analysis of burial patterns
from several classic sites. Ph.D. dissertation. Depart-
ment of Anthropology, Case Western Reserve Univer-
sity, University Microfilms, Ann Arbor, Michigan.
1979 A comparative study of site morphology and burial
patterns at Edwin Harness mound and Seip mounds 1
and 2. In Hopewell archaeology: the Chillicothe
conference , edited by David S. Brose and N’omi
Greber, pp. 27-38. Kent State University Press, Kent,
Ohio.
Greber, N’omi, and Dennis P. Griffin
1982 Comparison of excavations and subsurface remote
sensing data from sections of the Seip earthworks
complex, Ross County, Ohio. Paper presented at the
1982 Annual Meeting of the Southeastern Archaeo-
logical Conference, Memphis, Tennessee, October
28-30.
Greber, N’omi, and David Jargiello
1 982 Possible astronomical orientations used in construct-
ing some Scioto Hopewell earthwork walls. Paper
presented at the Annual Meeting of the Midwest
Archaeological Conference, Cleveland, Ohio, Octo-
ber 1-3.
Mills, William C.
1907 Explorations of the Edwin Harness mound. Ohio
Archaeological and Historical Quarterly 16: 1 13-193.
Prufer, Olaf
1968 Ohio Hopewell ceramics: an analysis of the extant
collections. Anthropological Papers No. 33. Museum
of Anthropology, University of Michigan, Ann
Arbor.
Prufer, Olaf H., D. H. McKenzie, O. Pi-Sunyer, H. C. Cutler,
R. A. Yarnell, P. W. Parmalee, and D. H. Stansbery
1965 The McGraw site: a study in Hopewellian dynamics.
Cleveland Museum of Natural History, Scientific
Publications (n.s.), 4(1).
Seeman, Mark F., and Frank Soday
1980 The Russell Brown mounds: three Hopewell mounds
in Ross County, Ohio. Midcontinental Journal of
Archaeology 5( 1 ) : 7 3— 1 16.
Shetrone, Henry C.
1926 Exploration of the Hopewell group of prehistoric
earthworks. Ohio Archaeological and Historical Quar-
terly 35: 1-227.
Shetrone, Henry, Frank Setzler, and Robert Goslin
1925- Field notes of Ohio State Museum Archaeological
1928 Expedition, Seip mound 1. On file. Department of
Archaeology, Ohio Historical Center, Columbus,
Ohio.
Shetrone, Henry Clyde, and Emerson F. Greenman
1931 Explorations of the Seip group of prehistoric earth-
works. Ohio Archaeological and Historical Quarterly
40:343-509.
Squier, George Ephram, and E. H. Davis
1 848 Ancient monuments of the Mississippi Valley. Smith-
sonian Contributions to Knowledge 1. Washington,
D.C. Reprinted 1973 with introduction by James B.
Griffin as Antiquities of the new world: early explora-
tions in archaeology (Vol. 2). A.M.S. Press, New
York, for Peabody Museum, Harvard University.
Swanton, John R.
1 946 The Indians of the southeastern U nited States. Bureau
of American Ethnology, Bulletin 137.
■
CONTENTS
No. 39.
Recent Excavations at the Edwin Harness Mound, Liberty Works, Ross County,
Ohio— N’omi Greber
KIRTLANDIA
CLEVELAND, OHIO NUMBER 40
•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,
Copyright © 1983 by The Cleveland Museum of Natural History,
Kirtlandia is abstracted in Zoological Record and Biological Abstracts and indexed in
Bibliography and Index of Geology.
ISSN: 0075-6245
KIRTLANDIA
THE CLEVELAND MUSEUM OF NATURAL HISTORY
CLEVELAND. OHIO SEPTEMBER 1983
NUMBER 40
A STUDY OF THE EASTERN BLUEBIRD
AT THE HOLDEN ARBORETUM,
LAKE COUNTY, OHIO
JEAN EAKIN
Abstract
A study of the nesting of the Eastern Bluebird at the Holden Arboretum in
Lake County. Ohio, was conducted from 1965 through 1980. Emphasis was
placed on causes of nest failure and methods used tocontrol them The 16-vear
volunteer effort contributed to an increase in the Eastern Bluebird population
at the Arboretum. Some information on the nesting ofT ree Swallows wasalso
gathered during the study.
Kirtlandia No. 40
0075-6245/83/1983-0040 $2.00
Copyright c 1983 by The Cleveland Museum of Natural History
Fig. 1: Holden Arboretum in August 1980
Introduction
In 1965, a volunteer effort was begun to increase the population of the
Eastern Bluebird ( Siala sialis) at the Holden Arboretum (hereafter referred
to as the Arboretum), in Kirtland and Mentor Townships, Lake County.
Ohio.
A previous effort had been made at this location to provide additional
nesting sites for bluebirds by a local volunteer, Paul Smith, who built and
placed 230 bluebird boxes on the property. Only eight pairs of bluebirds,
however, nested in this area in 1965. At that time. House Wrens ( Troglo-
dytes aedon) occupied 49 boxes and House Sparrows ( Passer domestieus)
occupied 134.
The boxes, located in 14 areas either in meadows or plantings along
hedgerows or woods, were unpainted with side openings and were placed
approximately 4 to 5 ft above the ground on metal fence posts. ( A few were
placed 6 ft high; one was only 2 ft above the ground.)
Project Bluebird was begun in 1965 by Virginia Barrus, a local ornithol-
ogist and bird bander, as a volunteer program to study the Arboretum
bluebird nesting population in an effort to maintain and, if possible, in-
crease it. Prior to the 1 965 nesting season she cleaned and repaired the 230
original boxes.
Although there have been several reports of such efforts to increase
bluebird populations (Kibler, 1969; Laskey, 1939; Pinkowski, 1979;
Musselman, 1935) none covered such a long period of time, 16 years
(1965-80), as in this study.
Because of a significant number of Tree Swallows ( Iridoprocne Bicolor)
nesting in the boxes, they have been included in this study.
The Study Area
The Holden Arboretum consists of 5 sq mi (approximately 2,800 acres)
of rolling land in Lake and Geauga Counties (lat 4 1°37'N; long 81° !9'W).
Elevations (excluding the highest point. Little Mountain) range from 750
to 1,200 ft. (The areas included in this study range from 900 to 1,100 ft.)
1 he area is drained by Stebbins Gulch and Pierson Creek, both of which
empty into the East Branch of the Chagrin River which runs through the
property. The Arboretum is approximately 27 mi northeast of Cleveland
and 2 mi south of Lake Erie.
As the accompanying map ( Fig. 4) shows, the area is highly irregular in
outline due to private ownership of adjoining properties. Although most of
the Arboretum’s properties are contiguous, one large area and several
small ones are separate.
SU3aWflN
YEARS
Fig. 3 Tree Swallows
Nesting Success 1965-1980
- o) r oi
w 2 « 2
O) 4) 0)0)
□
□
□
□
I I
f!
°
O 'NX
ol
o °l
\
I VfcS
Jp/S7 I
> 1 ,j7 S
0 1 ^ I
o ! \ 1
\ >4fc
^ .L^
v*
|0o\
«>n)
£ /^AV
1°^
X
_CN
SHAMAN
YEARS
6
JEAN EAKIN
No. 40
Fig. 4: Eastern Bluebird nest box areas at Holden Arboretum. Areas active in 1980:
T — Thayer Center; CL — Corning Lake; HP — Heath Pond; CA — Chapin Acres;
BA Bicknell Acres; N — Norweb; GP — Goose Pond; PM — Paul Martin Farm;
P Pines; DMA- Upper Mather Acres; LBF — Lower Baldwin Farm; UBF —
Upper Baldwin Farm. Areas discontinued during the study period were: CB —
Cooper Building; I — Island in Corning Lake; PA — Park Acres; AC — Around
Crabapples; AB— Around Blueberry Pond; AL — Around the Lilacs; LMA —
Lower Mather Acres.
1983
EASTERN BLUEBIRD
7
At least 1,700 acres are covered by woodland —coniferous, mixed, and
deciduous woods, with the latter predominating. There are 18 lakes and
ponds, the largest of which is Corning Lake, approximately 20 acres.
The areas included in this study are located either in nurseries or plant-
ings with sufficient open space to meet the needs of bluebirds, or in un-
developed fields. On the map each area is defined by the point at which
field meets roadway, woodland, or. in a few cases, an extension of the field
which was unsuitable because of competition from other species or heavy
human use. Areas which had boxes at the start of the project but which
have been phased out, as well as the present! 1980) placement of boxes, are
shown.
The following is a detailed account of each study area.
Former Areas
Cooper Building (CB). The area in the immediate vicinity of the admin-
istrative building contained 10 boxes in 1965, all of which were inhabited
by House Sparrows. We removed these prior to the 1967 season.
Island in Corning Lake (1). This area had 10 boxes in 1965. All boxes
were removed after the 1967 season due to inaccessibility for monitoring
and occupation by House Sparrows.
Park Acres (PA). An area on the southeast corner of Sperry and
Kirt land -Chard on Roads contained 10 boxes in 1965 inhabited by House
Sparrows. They were removed prior to the 1967 season.
Around Blueberry Pond( AB). Most of thisarea iscovered by deciduous
woodland and mature coniferous plantings. Seventeen boxes were placed
in small open spaces. There was one successful bluebird nesting in 1967. In
1975 and 1976, bluebird attempts were interrupted by predators and
wrens. Otherwise there was no bluebird activity. Black-capped Chickadees
( Parus atricapillus) successful in 1967 and 1968, were interrupted by wrens
or human beings in subsequent attempts. There are several heavily used
trails. Boxes were removed gradually until none remained in 1979.
Lower Mather Acres ( LM A). This is a small triangular area with an en-
trance on Mitchell's Mills Road. It is surrounded by deciduous or mixed
woods and contains an old barn. To the southeast is Stebbins Gulch. This
area was heavily populated by House Wrens. Six bluebird attempts were
unsuccessful due to interference by wrens. The original nine boxes were
gradually reduced to six by 1975. Five of these were removed before the
1976 season. The last box was removed before the 1980 season.
Around the Lilacs (AL). This area stretches from the Arboretum re-
ception center south along Sperry Road for approximately 1,000 ft to a
JEAN EAK.IN
No. 40
property line where it is separated by a fence from an undeveloped field.
On the north a double row of white pines ( Pinus strobus ) separates it from
the parking area. It extends west from the road for approximately 150 ft
where it meets deciduous woods and groves of pine and aspen. A small
pond and brook separate the lilac collection in the larger northern portion
from an open area containing scattered trees and shrubs and continuing to
the west where there is another pond.
1 he original 17 boxes were gradually reduced to 14 by 1969. Nine were
removed prior to the 1970 season in an effort to reduce Elouse Sparrow and
House Wren competition. There were four boxes from 1971 through 1974.
In 1979 the last two boxes were removed due to heavy human use of the
areas where they were located and continuing House Sparrow activity.
During 10 of the first 1 2 years of the study one pair of bluebirds nested or
attempted to do so. In 1967 and 1969 there were two pairs. Eight broods
fledged, the last one in 1974. Of 14 nestingattempts byTree Swallowsonly
five succeeded, the last one in 1973.
The Crahapple Collection (AC). This area contained 24 boxes in 1965.
In an effort to reduce House Sparrow and wren competition, we reduced
the number to eight in 1970. As tree growth gradually eliminated open
space, this area failed to attract bluebirds except for an open triangle con-
taining three boxes at the extreme northwest bounded on two sides by de-
ciduous woods and on one by crabapples (Mains) and hawthornes
(Crataegus). By 1979 this remaining section (Thayer Center, T on map)
contained only one box.
The total AC area almost always attracted one pair of bluebirds, oc-
casionally two. However, prior to 1976, when all but three of the boxes
were removed, bluebirds were successful in only three years. From 1976
through 1980, with the exception of 1978 when the nesting was interrupted
by House Sparrows, one or two broods fledged each year from the re-
maining box in section T.
Areas of Continuous Use
Corning Lake (CL). This area surrounds Corning Lake except on the
Sperry Road side (Fig. 5). It is almost entirely surrounded by deciduous
woodland or hedgerow and contains collections of evergreens, birch
( Betula), alder (Alnus), and hornbeam ( Ostrya). The open space is broken
here and there by maple (Acer) and sassafras (Sassafras alhidum) and there
are willows (Salix) near the lake. Both islands are covered with dense de-
ciduous growth. Most of the grassy area is mowed.
1983
EASTERN BLUEBIRD
9
Fig. 5. Corning Lake Area. Holden Arboretum.
By 1967 the original 29 boxes had been increased to 32. In 1968, we re-
moved eight and relocated seven. Since then the number has been reduced
to 19. We eliminated five boxes which attracted House Sparrows and/or
wrens and relocated two others.
In this area the number of bluebird pairs varied from two to four with an
average of 3.3, and 56 broods fledged. From 1967 through 1980, with the
exception of 1977, at least one pair of Tree Swallows nested. In 1970 and
1974 there were four pairs. During one season there were three, and during
three seasons, two pairs.
Heath Pond (HP). This small area is almost surrounded by mixed
woods. It contains two ponds, two enormous old oaks ( Quercus alba and
Quercus rubra ) and the Arboretum’s rhododendron and azalea collec-
tions. In 1965 there were nine houses. By 1971 we had removed all but four
in an attempt to reduce wrencompetition, which appeared to have stopped
bluebird attempts to nest. After 1 976 there were only three boxes. The first
bluebird success in this area was in 1971. Since then at least one brood
fledged in 8 of 10 years.
Bicknell Acres ( BA). This is a strip of land approximately 500 ft wide
and 2,200 ft long on the east side of Sperry Road extending north-
northeast from Kirtland-Chardon Road to a low wooded area bordering a
10
JEAN EAKIN
No. 40
Fig. 6. Bicknell Acres, Flolden Arboretum
Fig 7. Chapin Acres, Holden Arboretum
1983
EASTERN BLUEBIRD
I I
pond (Fig. 6). The magnolia collection at the north end, which contains
two boxes, is surrounded by deciduous woodland except where it meets the
road. South of the magnolia collection, deciduous woodland nearly meets
the road in three places, leaving two fields which slope to the east for
1,700 ft before they meet the woods. There are six boxes in all.
There was one pair of bluebirds during two of the first six years. From
1971 through 1980 there have been two with the exception of 1976 when
there were three pairs. A pair of Tree Swallows nested in this area during
four years, including 1979 and 1980.
Chapin Acres(C A). Thisarea is on the west side of Sperry Road ( Fig. 7).
At the northeastern end is a pond. Deciduous woodland extends from the
road along the far edge of the pond, then southerly, coming within 250 ft of
the road at the southern end. The evergreen planting which occupies much
of the area continues south to Kirtland-Chardon Road. This southern part
of the area had boxes which were removed prior to the 1 976 season because
they attracted House Sparrows and wrens. The one bluebird attempt
(1974) failed because of wren interference. Since 1976, there have been
seven boxes in the northern section of CA.
There was one pair of bluebirds in 4 of the first 10 years. From 1975
through 1979 there was one pair. In 1980 there were two pairs. At least one
pair of Tree Swallows nested in this area every year since 1966.
The Paul Martin Farm (PM). Thisarea is bounded by deciduous or con-
iferous woodland and traversed by three hedgerows following old fence
lines (Fig. 8). There are three buildings near Kirtland-Chardon Road. On
the southern border is a small pond drained by a brook which crosses the
property in a north-northeasterly direction. Considerable open space re-
mains in the northern field (approximately 375 by 2,000 ft) where the Ar-
boretum’s maple collection is located. The southern portion adjacent to
Kirtland-Chardon Road is planted with mountain ash ( Sorbus ) and var-
ious species of horsechestnut ( Aesculus ), leaving little open space. The
remainder of the southern portion consists of an undeveloped and un-
mowed field approximately 275 by 1,100 ft.
From 1965 through 1975 there were 30 boxes, some of which were grad-
ually relocated away from Kirtland-Chardon Road and the buildings in
order to avoid House Sparrow occupancy. In 1976, after studying the his-
tory of each box, we removed all but 1 2, of which 6 were relocated. In 1980
there were 14 boxes.
Pines (P). In 1977 we placed three boxes in the lot on the corner of
Kirtland-Chardon and Booth Roads adjacent to the PM area. It is an un-
developed field bounded by coniferous hedgerow, deciduous hedgerow,
and woodland except where it meets the two roads. A pair of bluebirds
12
JEAN EAKIN
No. 40
Fig 8. Paul Martin Farm. Holden Arboretum,
nested in this area each year.
The number of bluebird pairs in PM gradually increased from one in
1965 to three in 1976. After the addition of pines (P) there were three pairs
in the combined areas (four in 1978). Usually there were one or two pairs of
Tree Swallows in PM.
Goose Pond (GP). Except for the Kirtland-Chardon exposure on the
northeast, this area is surrounded by deciduous woodland. It contains five
ponds and lakes. There are two dense plantings of mature conifers; one en-
tirely white pine; the other, white pine and spruce ( Picea). There are many
willows ( Salix). The area available for boxes is divided by a thick hedgerow
running north and south. There are six boxes. One pair of bluebirds nested
each year of the study. There were also usually one or two pairs of Tree
Swallows.
Upper Mather Jcra(lJMA). This is an unmowed field of coarse grass
and weeds (approximately 400 by 2,000 ft) entirely surrounded bv mixed
or deciduous woodland or thick, dense hedgerow. The northeastern half
contains a nursery of small conifers. There were 10 boxes from 1965
through 1974. In 1975, due to vandalism and deterioration, only five re-
mained. Before the 1976 season, we relocated these in an effort to reduce
1983
EASTERN BLUEBIRD
13
House Wren interference. Another box was installed before the 1980 nest-
ing season.
With three exceptions (all prior to 1976) there was a pair of bluebirds
each year. From 1965 through 1976, presumably due to wren competition,
the bluebirds failed in three of nine years. In three additional years, wrens
prevented a second nesting by building in the boxes. From 1977 through
1980, one or two broods of bluebirds fledged each year. One pair of Tree
Swallows nested in 12 of the last 14 years with only four failures.
Lower Baldwin Farm (LBF). Except where it is bordered by the road,
this area is almost entirely surrounded by deciduous or mixed woodland.
The higher eastern section contains mature nut trees and the younger nut
tree collection. The land slopes south and east to the lower field which con-
tains a mainly deciduous mixed nursery. Except on one steep grassy slope,
a patch of deciduous woods separates the two areas.
There were originally 20 boxes in this area. Six of these, south of the ex-
tension of Baldwin Road, were relocated in the present area by 197 1 . From
1968 through 1974, while the number of boxes remained almost constant,
many were relocated in an effort to lessen competition from wrens and
House Sparrows and attract more bluebirds. Prior to 1976, after studying
the history of each box, we reduced the number to nine, of which four were
in new locations.
During three of the first four years of the project, only one pair of blue-
birds attempted to nest in this area. Only two broods fledged. From 1970
through 1980 there have been two or three pairs every year and 29 broods
have fledged. In 1977 a Tree Swallow nesting was interrupted, presumably
by a predator. In 1979 and 1980 two pairs of Tree Swallows were suc-
cessful.
Post-1965 Areas
Upper Baldwin Farm (U BF). This “upper field,” (actually lower in ele-
vation than the lower field) is approximately 1,000 by 2,500 ft and is en-
tirely surrounded by mature deciduous and mixed woodland. It containsa
mixed deciduous nursery of small trees and shrubs. The grass is not
mowed. Four boxes were placed in this area in 1968. Three additional
boxes were added in 1969. One box was removed in 1975 after House
Sparrows had destroyed bluebird nestlings for two successive years.
During the 13 years this area always had one or two pairs of bluebirds
and 26 broods fledged. Since 1969 there was at least one pair of Tree
Swallows every year but one. Six Tree Swallow broods fledged.
14
JEAN EAK.IN
No. 40
Norweb (N). This area, added in 1972, is bounded by deciduous woods
and by two extensions of its own fields which have not attracted bluebirds.
The eastern portion contains a holly planting. The western part is an un-
developed, un mowed field. The number of houses varied from five to eight.
Three at the south end near buildings were eliminated or moved due to
House Sparrow competition. From 1978 through 1980 there were five
boxes.
During its nine years this area always had one pair of bluebirds. Nine
broods fledged. There were usually one or two pairs of Tree Swallows. Five
Tree Swallow broods fledged.
Methods
Monitoring of Boxes. The original plan in this study was to monitor the
contents of the bluebird boxes on a weekly basis, recording the contents
and removing House Sparrow nests. Sometimes during the first few years
the interval between visits to boxes was as long as two weeks. In 1966 there
were no visits after July 16. Since 1978 an effort was made to check the
boxes twice a week. Beginning in 1975 we tried to check boxes with nest-
lings every two days from approximately the fifth to the tenth day. After
the twelfth day we used extreme caution in checking. We did not disturb
birds which were engaged in house selection or nest building. We disturbed
females brooding eggs as little as possible.
Box Placement. During the 16 years of the project we removed some
boxes and added others. In 1967 we eliminated twoareas(23 boxes) which
had attracted only House Sparrows (see Fig. 4, Areas CB and PA). We re-
moved II boxes from an island (Area I) in another area for the same rea-
son. The number of boxes in two areas near the Arboretum Reception
Center was gradually reduced until none remained in 1979. One of these
sites ( AB, 17 boxes) was a trail through the woods which lacked sufficient
open space to attract bluebirds. The other (AF, 16 boxes)ceased toattract
either bluebirds or Tree Swallows after 1975 as it became more heavily
used by people and as House Sparrow competition increased. Only one of
28 boxes was left in a remote part of the crabapple area (AC) which was
heavily used by people and frequented by wrens and sparrows. Boxes were
removed from parts of three areas which attracted only House Sparrows
(CA, PM, FBF) and from another area which attracted only wrens
(FMA). Roadside boxes, 22 at most in 1969, which were productive for
several years, were eliminated gradually due to vandalism, predation and
House Sparrow competition. Three new areas with 15 boxes were added,
all of which attracted bluebirds (see Fig. 4, UBF, N and P).
1983
EASTERN BLUEBIRD
15
Fig. 9. Bluebird box on post with metal guards, winter
1980.
In 1966, we decided that the boxes were placed too close together to meet
the territorial needs of bluebirds. Superfluous houses only served to attract
wrens and House Sparrows. In an attempt to reduce interference by these
two species, we removed some boxes and relocated some, making certain
that they were at least 100 ft apart. This was not done during the nesting
season. In the winter of 1975 and 1976, after reviewing the history of each
box, we reduced the remaining 146 to 100. After this, however, it was still
necessary to remove or relocate certain boxes because of new planting or
hedgerow growth. In 1980 there were 86 boxes.
Since we found that bluebirds, when their nesting was interrupted, often
sought a different site in the same area, we provided several boxes as alter-
nate choices for pairs that might be disturbed ( Krieg, 1971 : 124). We had
enough boxes in every area to accomodate such other species as Tree
Swallows and Black-capped Chickadees.
16
JEAN EAKIN
No. 40
We gradually replaced the original fenceposts with metal poles and
guards to protect the boxes from predators (Fig. 9).
Banding of Young. Most of the young were banded (Table 3 and Table
14). The banding was done when the birds were no younger than 8 or older
than 12 days.
Results
An Analysis of Factors Causing Nest Failures
House Sparrow. The fJouse Sparrow is an aggressive competitor for
bluebird nesting boxes ( Kibler, 1969: 123; Zeleny, 1976: 1 16-1 19). In the 16-
year study evidence indicated that House Sparrows destroyed 3 adult
bluebirds, at least 33 young and 55 eggs, and interrupted 89 additional at-
tempts of the bluebird to nest. There is also evidence that they destroyed 9 adult
Tree Swallows, 29 young and 24 eggs, and interrupted 12 additional nest-
ings (Table 4). Sometimes we actually witnessed House Sparrows entering
boxes and attacking the occupants. In the other cases, we found the House
Sparrow nest on top of the bluebird or Tree Swallow nest, often with bro-
ken eggs or dead young. Young small enough to be removed were some-
times found on the ground under the box. When they were too large, the
House Sparrows usually delayed nest building until we removed them.
Two of the bluebird adults were buried under House Sparrow nesting ma-
terial. The other adult bluebird and the adult Tree Swallows were found in
empty boxes early in the nesting season. All of the birds listed as destroyed
by House Sparrows were severely pecked on the head.
We always found House Sparrow nests when we first checked the boxes
in mid-March. Eggs appeared after the middle of April. Nesting activity
peaked in May, continued into June and decreased in July. There were
some attempts in August. Most bluebird casualties occurred in May and
June. We found House Sparrow nests in boxes less than fourfeet from the
ground and in a box of smaller than average dimensions. In 1980 House
Sparrows built in an experimental box with a screentop designed to be
“sparrow-proof.”
We routinely removed House Sparrow nests from our boxes. More
often than not the House Sparrow returned to build again. In 1972, out of
75 cases of nest removal, 9 did not return; I 1 came back once; 27 returned
twice; 5, three times; 3, four times; 2, five times; 4, six times; and 14, seven
times.
In 1970 a bluebird nest was completed on 22 April. On 27 April there was
a House Sparrow nest on top of a dead, severely pecked female bluebird in
1983
EASTERN BLUEBIRD
17
the bluebird nest. (In this same box four bluebird young were pecked to
death by House Sparrows in June of the same year.) We removed this box.
On 10 April 1976, a dead, severely pecked male bluebird was found in an
empty box. House Sparrow nests were removed from this box on 23, 27,
and 28 April.
On 9 May 1977, I found a House Sparrow nest with three eggs on top of
a dead male bluebird which had been pecked severely on the head. This
box was removed.
We found evidence that suggests that the H ouse Sparrow’s success is not
due to the unwillingness of the bluebird to fight or aggressively defend the
nesting site or young, but to the fact that the bluebird pairs leave the nest
together to seek food and return to the nest almost simultaneously. The
eggs and young, therefore, are intermittentlv left undefended (Goldman,
1975:800).
In 1965 a dead House Sparrow was found at the foot ofa box with blue-
bird young which subsequently fledged. A second brood was destroyed by
House Sparrows.
On 18 June 1970, 1 founda HouseSparrow pairata bluebird house. The
four approximately one-week-old bluebird young were alive but seemed
weak. All had been pecked on the head and one had been severely pecked
on the back. The bluebird pair returned at intervals, chased the House
Sparrows away, and fed the young. The sparrows waited until the blue-
birds left, at which time the male sparrow perched on the box and the fe-
male entered. I chased them away when the female entered and otherwise
waited for the bluebirds to return and defend the box.
When the bluebirds seemed to be staying near and defending the box, I
left for 20 minutes. When 1 returned the male sparrow was on the box, the
female inside. The young had been pecked more severely and were quiet.
The male bluebird returned with food and called softly and repeatedly for
some time but did not enter the box when he received no response from the
young. When the male House Sparrow returned and sat on the box, the
male bluebird did not challenge. The female bluebird returned and both
she and the male fluttered around the box calling but did not enter They
made no attempt to chase the male House Sparrow, which finally flew
away. The female bluebird perched on the box until the male House Spar-
row returned; then she flew to a nearby stake protesting but not fighting.
The male House Sparrow entered the box.
During the several hours that 1 observed them the bluebirds fought ag-
gressively every time they returned to feed the young until the young be-
came too weak to respond to their calls (Hartshorne, 1962:141). At our
next visit, three days later, there was a House Sparrow nest on top of the
bluebird nest and one dead bluebird young on the ground.
18
JEAN EAK.IN
No. 40
In 1975, a House Sparrow was observed several times leaving a box in
which a female bluebird was brooding. This nesting was not interrupted.
In 1976, a dead House Sparrow was removed from a bluebird nest. The
nesting was successful.
In 48 cases bluebirds nested successfully in boxes from which House
Sparrow nests had been removed ( Kibler, 1969: 123). In 21 cases Tree Swal-
lows were successful after the removal of House Sparrow nests (Table 4).
House Wren. The House Wren is a formidable competitor for nesting
cavities and may destroy eggs and young in boxes which it does not occupy
(Zeleny, 1976: 1 20). Since it is a protected native bird we simply included it
in our records during the first six years of the project. We attempted to pro-
tect the bluebirds by relocating boxes away from hedgerows to discourage
wrens (Kibler, 1969: 119).
Wrens occupied 49 of 230 boxes in 1965; 95 of 200 in 1970. In six years
approximately 900 young fledged. We believe that in this same period
wrens destroyed four bluebird young and 40 eggs and interrupted eight
other nestings. After 1970 we decided to remove wren nests from boxes
which had a bluebird history or were attracting bluebirds. In 1971, wrens
occupied or attempted to occupy 92 of 200 boxes; in 1972, 105 of 189. In
these two years at least 450 fledged (Tables 6 and 7).
In 1973 and 1974, with the approval of the Arboretum naturalist, weex-
perimented with nest removal and with removal of all but one or two eggs.
The majority of wrens kept returning after nest removal but deserted if
only one or two eggs remained. In 1973, 135 young fledged. In 1974, only
35 fledged but the destruction of bluebird eggs increased from an average
of 9 to 29.
In the 10 years, 1965 through 1974, approximately 1,300 wrens fledged.
During this period indications were that wrens destroyed 13 bluebird
young and approximately 92 eggs; 5 Tree Swallow young and 14 eggs; and
28 chickadee eggs. When the eggs were pierced and/ or the wren built on
top of the nest, I attributed the failure to wrens.
Beginning in 1975, we routinely removed wren nests. We occasionally
made an exception when bluebirds were nesting nearby. Wrens continued
to attempt to occupy about one-fourth of our boxes.
Wrens probably destroyed 21 bluebird eggs in five nests in 1975; 17 eggs
in four nests in 1976; 8 eggs in two nests in 1977; 1 1 in three in 1978; none in
1979; one in one nest and 2 in another in 1980 (Table 7).
The adult bluebird listed as destroyed by House Wrens in Table 7 was a
female found in a box in which bluebird nest failure had been followed by
two attempts by wrens to nest. Her head was pecked clean of feathers. The
wrens subsequently occupied the box.
Predators. Predators of several species destroyed 5 adult bluebirds, 55
1983
EASTERN BEUEBIRD
19
young, 1 10 eggs and 13 Tree Swallow young and 32 eggs(Table 8). Some-
times it was possible to determine that the destruction was caused by a rac-
coon or a red squirrel, while in other cases the evidence indicated a snake.
Often we could not determine the identity of the predator. Greasing the
metal fence posts was ineffective. In 1975, we began to replace these posts
with round metal poles which could be greased, and in 1977 we began to
place rectangular metal guards on the poles (as suggested by Zeleny,
1976: 107-109). In 1980 these guards were on all active houses. We have had
two cases of predation in boxes so equipped. One box was on a short pole,
since replaced. In 1980, one of three eggs disappeared and latertwo banded
young disappeared from a box 52 in. above the ground.
Human Interference. Human interference caused the loss of 9 bluebird
young and 49 eggs, and 14 Tree Swallow eggs. Two chickadee nests con-
taining five and six eggs respectively, were removed from boxes in sight of
a trail. Three young were found dead at the foot of another box, the door of
which was open. We removed these boxes. In 1976 we put Philips screws
on boxes in areas open to the public. More recent incidents of human dis-
turbance to the boxes in remote areas made it necessary to put such screws
on all boxes (Table 9).
We attempted to keep human interference at a minimum by making the
boxes inconspicuous and by removing ones that proved vulnerable.
Unknown Causes of Destruction or Disappearance. We could not al-
ways determine the cause of nesting failure. We could not rule out House
Sparrows, House Wrens, predators or human beings. Such cases ac-
counted for 19 bluebird young, 57 eggs and 25 nest failures. Three Tree
Swallow nests were destroyed by unknown causes.
Parasitism by Cowbirds. On 8 May, 1972, there were two Brown-headed
Cowbird (Molothrus ater) eggs in a bluebird nest with two eggs. On 16
May a third bluebird egg was in the nest. The two cowbird eggs were re-
moved. Three bluebirds fledged from this nest.
On 15 June, 1971, a young cowbird was found in a Tree Swallow nest
with two unhatched eggs (there had been five). In the same year one live
cowbird nestling was found in a nest with four dead T ree Swallow young.
Wasps. We removed many wasp nests from boxes. In May 1975, after a
bluebird nest with three eggs had been deserted, we found an active wasp
nest. In 1975 a Tree Swallow nest with five eggs was deserted after an old
wasp nest fell on top of it. In 1978 we began to coat the ceiling and upper
walls of the boxes with Vaseline. This reduced but did not eliminate the
problem.
Disappearance of Eggs or Young from Nest. Sometimes it was not pos-
sible to determine whether eggs or young had disappeared from a nest con-
taining live nestlings. When the nest was infested with fly larvae we as-
20
JEAN EAKIN
No. 40
sumed that dead young had been removed by the adults. It is unusual for
single eggs to disappear from clutches which are otherwise successful
( Pea kail. 1970:25 1 ). When in doubt as to whether an egg or a nestling had
disappeared 1 assumed that a dead nestling had been removed by a parent
(Hartshorne, 1962:145-146).
Failure of Eggs to Hatch. In 16 years, eight clutches of bluebird eggs
failed to hatch. In the case of one second clutch of two eggs the pair was
near the nest on the eighteenth day after completion. Four white eggs,
warm on the twelfth day, were cold on the eighteenth. One clutch, the third
of the season, was found covered with new nesting material five to eight
days after laying was completed. The pair did not return. In two cases one
egg was laid and then deserted. In 59 otherwise successful clutches of blue-
bird eggs, 70 failed to hatch.
Two clutches of Tree Swallow eggs failed to hatch. In 15 partially suc-
cessful clutches, 38 failed.
A paulina sialia. Parasitic fly larvae, A paulina sialia, first became a
problem in 1968. After two clutches of five young and nine other young
were found dead in infested nests we began to dust the nests ( with or with-
out eggs but not with young) with rotenone. There were no further casual-
ties in 1968. In 1969 and 1970 the only fatalities from parasites occurred in
nests which had not been treated. In 1971 six young disappeared and one
was found dead in nests on which rotenone had been used, and in 1972 two
clutches of banded young were found dead in infested boxes. One of three
was found dead in another nest. Three young fledged from another after fly
larvae had been removed by hand. Only one young Tree Swallow fledged
from four infested nests in which 18 were dead or missing. Rotenone had
been used in all of these nests. In 1973 and 1974, larvae continued to appear
in nests treated with rotenone, but 25 young fledged from infested nests af-
ter larvae had been removed manually.
In 1975, suspecting that some of the parasitic flies might have developed
immunity to rotenone, we decided to remove larvae by hand. Examination
of nest material spread out on white paper revealed the presence of many
larvae too small to be visible in the box. Since it was impossible to remove
them without destroying the nest, our solution was to substitute a man-
made nest of dry grass, as described by Johnson (1932:29) and Mason
(1944:232).
From the larvae and pupae which we collected. Dr. Sonja Teraguchi (en-
tomologist at the Cleveland Museum of Natural History) hatched adult
Apaulina sialia and the hyper-parasite Mormortiella. She advised us to
remove infested nests and place them at the foot of the boxes in wire mesh
bags which would confine adult Apaulina sialia but permit the smaller
Mormoniella to escape. We followed this procedure in most cases. We sub-
1983
EASTERN BLUEBIRD
21
stituted artificial nests of dry grass which the bluebirds never failed to
accept.
From 1975 through 1979, 60 bluebird and 9 Tree Swallow voungdied in
infested nests, most of which probably had not been checked soon or
thoroughly enough. During the same period 152 bluebird and 61 Tree
Swallow young fledged from nests that had been substituted for infested
ones. In 1980, although larvae were present in the nests of 1 8 broods, there
were no nest failures. Four bluebird young were found dead or dis-
appeared from four nests in which larvae were present. TwoTree Swallows
were found dead and one disappeared from two such nests.
Apaulina sialia appeared from late May through August. They were
present in every area. In some cases they were present in the first nesting
but not the second in the same box; in other cases the reverse was true. In
1979, in at least 10 cases larvae were present- even abundant one to
three days after the box had been cleaned. They were found afterall or part
of the young had Hedged in boxes that had been cleaned on the tenth to
twelfth days. During six seasons! 1975-1980) larvae werefound in the nests
of 103 of 155 broods in which young survived to the age of one week (see
Table 10).
Other Parasites. On 10 July 1975, two bluebirds, I or 2 days old, were
dead in the nest. Dr. Teraguchi found them covered with microscopic lar-
vae. (Of 4 eggs, 2 had disappeared, one at a time.)
In 1973 a Tree Swallow nest had six eggs due to hatch on 20 June. On 25
June there were four live young and one dead. On 2 July there were only
two young. The nest was infested with tiny Hies which a volunteer re-
moved. On 6 July there were a few flies which were removed. The young
were banded on 10 July, when they were approximately 20 days old. T wo
days later one of the young was seen leaving the nest. Both Hedged.
Young Dead in Nest — Cause Unknown. In June of 1976 a volunteerdes-
cribed one of four Tree Swallow young, 19 or 20 days old, as“sick. "There
had not been and were at no time any visible parasites in the box. Two days
later there were three dead young.
On 9 June, 1977 I banded four bluebirds approximately eight days old.
They were active and seemed healthy. No larvae were present. The adults
fed them as soon as I left. Two days later they were dead. I took these
young to Dr. Aaron Leash, a local veterinarian, who found them to be well
nourished. He dissected one and could find no parasites or abnormalities.
In a nearby area on 8 June, two young approximately two days old were
found in a nest which had had six eggs. One tiny nestling was dead on the
ground at the foot of the box. On the 1 I th, I found the two young ailing,
one on its back and one on its side. The adults were in a nearby tree. After
consulting the Arboretum naturalist, 1 gave each nestling two drops of vit-
22
JEAN EAKIN
No. 40
amins which seemed to revive them. The parents fed them before I left. We
administered vitamins again on 13 and 15 June. The young fledged at ap-
proximately 17 days.
In a nearby area three Tree Swallow young, hatched on 8 June, seemed
normal on 1 1 and 13 June but were dead on the 18th. No larvae were
found.
These three boxes were near ornamental trees and shrubs which had
been sprayed. After that date, the Arboretum kept me informed of any ne-
cessary spraying, which was always selective.
Weather. Kibler states, “That inclement weathercan be a significant fac-
tor in bluebird mortality is borne out by the result of my first nesting period
in 1968 in which 76% of 51 nestlings died in the nest” (1969:125).
On four occasions in the course of our study adverse weather conditions
appeared to be the cause of young mortality and/or failure of eggs to
hatch.
In 1973, from 4 to 6 May, the temperatures of the Chardon, Ohio,
weather bureau were recorded as follows:
There were six bluebird nests in five of which dead young were discovered
or from which young were missing:
1 ) On 6 May, four young, hatched approximately 20 April, were dead.
2) On 7 May, four young, hatched on 1 May, were dead.
3) On 9 May, a nest which had had fouryoung approximately four days
old on 3 May, was empty.
4) On 9 May, one of three young, newly hatched on 1 May, was missing.
5) One nest had four eggs on 29 April. On 5 and 1 1 May the incubating
female did not leave. On 20 May the female was dead on the nest with
two dead young several days old.
6) On 21 April, a female was incubating five eggs. On 29 April there
were four young and one egg. Four young were banded on 8 May and
Hedged.
No fly larvae were found in these boxes.
Of seven nests with eggs, one clutch of five and one of two failed to
4 May
5 May
6 May
Maximum
48
42
56
Minimum
32
29
30
1983
EASTERN BEUEBIRD
23
hatch. In three clutches, one of five, one of three, and one of four failed to
hatch.
On 27 April, 1976 daytime temperatures were in the low 30s; there was a
ground cover of snow on 3 May, and the temperature was in the low 30s in
the daytime of 4 May.
During this period there were six nests with young:
1 ) One brood of four hatched approximately 25 April had food in their
stomachs on 27 April. They were alive on 4 May, dead on 8 May.
2) One brood of four hatched approximately 2 May were alive on 6
May, dead on 8 May.
3) One brood of five alive and approximately one week old on 5 May
were dead on 10 May.
4) Two young alive and approximately three days old on 5 May were
dead on 10 May.
5) One nest which contained three eggs on 14 and 28 April contained
one egg and one young on 5 May. On 9 May there was only one egg.
6) One nest which had four eggs due to hatch on 3 May had three young
and one egg on 8 May. The three young fledged in spite of fly larvae
infestation after a new nest was substituted.
No fly larvae were found in the first five boxes.
Of six clutches of eggs, two failed to hatch.
On 8 May, 1977, a cold front moved across Lake Erie with wind gusts in
excess of 70 mph. The temperatures (° F) in Chardon for that day and the
ensuing four were as follows:
Maximum
8 May
56
9 May
65
10 May
46
11 May
55
12 May
66
Minimum
27
3 1 (45° during dav at Arboretum)
28
27
34
During this period there were five nests with young:
1 ) In a box at the highest elevation in the area, only one of five young
survived. These young hatched between 3 and 7 May. No fly larvae
were found.
24
JEAN EAK.1N
No. 40
2) One of a brood of five hatched between 5 and 7 May was alive on 22
May but dead in the box after the others fledged. Fly larvae were
present.
3) Two young hatched on 8 May and fledged.
4) One of five young hatched between 2 and 7 May disappeared on or
before 8 May. Another disappeared between 8 and 13 May. On 15
May the remaining three were dead. Fly larvae were removed on 13
May.
5) In one box were five eggs due to hatch on 5 May. There were only
four young on 8 May. One of these disappeared bv 23 May, when fly
larvae were found. The others fledged.
There were six nests with eggs:
1 ) In one clutch completed on 3 May, four of five hatched.
2) One clutch of four was completed on 8 May. The female was incu-
bating on 15 and 21 May. On 23 May the eggs were gone.
3) One nest contained five eggs on 23 April and 3 May. On 10 May, a
Tree Swallow had built on top. We do not know if the eggs failed to
hatch or if the Tree Swallow drove the bluebirds away.
4) A clutch of five eggs present on 24 April and 9 May had hatched by 16
May (probably on 9 or 10 May).
5) Of four eggs present on 9 May, one disappeared and three hatched.
6) On 9 May a female was brooding five eggs due to hatch 1 1 May. On
16 May there were two unhatched eggs and one dead nestling.
In 1979 the maximum and minimum temperatures (° F) and the precipi-
tation ( inches of rain) at Chardon were as follows from 24 through 27 May.
Maximum
Minimum
Precipitation
24 May
73
35
.00
25 May
53
36
2.00
26 May
52
38
.70
27 Mav
5!
35
.73
In five boxes, 24 healthy young free of fly larvae and banded or ready to
band were found dead on 26 and 27 May. They were from one to two weeks
old on 24 May. Four in another brood which hatched on 6 May and had
not developed normally, although free of fly larvae, were also dead on 26
1983
EASTERN BLUEBIRD
25
May. In one clutch of four dead on 26 May, Hv larvae were present. In two
cases in which three of five and four of five approximately two weeks old
were found dead, fly larvae were present. The three young which fledged
from these two nests were the only ones to survive this cold spell. Three
nests with eggs were not affected; all hatched.
The four Tree Swallow nests which contained eggs were deserted. In
every case the pairs made a new start in the same box.
(On 26 May there was no activity in the barn where Barn Swallows
( Hirundo rustica) were nesting except for two brooding adults. On 27 and
28 May, seven adults were dead on the floor. The nests containing eggs
were deserted. )
White Eggs
From 1973 through 1980 we found 44 white or extremely pale blue eggs
in 10 clutches, six of which were at least partially successful. Of the 18
banded young, none was recaptured. Two of the females producing white
eggs had hatched from blue eggs.
Clutch Size
Clutch size varied from 3.9 in 1968 to4.66 in 1965 and 1980. Theaverage
was 4.22. I used only completed clutches to make this calculation (Table
Breeding Season
The breeding season in terms of clutch completion extended from the
first week in April to mid-August. The 16-year average shows a peak in late
April and early May and a lower peak from 9 to 28 June (Table 12).
Interval Between Broods
In computing the interval between fledging or nesting interruption and
the laying of the first egg of the next clutch I have used only pairs which
have nested a second or third time in the same house. If 1 did not know the
exact dates I used 14 days as the incubation period and 17 days as the age of
26
JEAN EAKIN
No. 40
fledging (Clapp, 1974:15-19; Thomas, 1946:156-158). From fledging to
laying of first egg in 22 broods of one to three young the average interval
was 20 days. The range was from 42 days to 8 days. In 60 broods of four or
five the average interval was 24 days. The range was 42 to 9 days.
In five nest failures involving eggs the intervals were 65, 54, 32,21, and 2
or 3 days with an average of 25. In 13 nest failures involving young the
average interval was 19.4 days with a range of 45 to 8 days.
Banding Records
Most of our bluebird young were banded (Table 3). We were able to pick
up 10 brooding females and record their band numbers. Nine of these had
been banded as nestlings. Four were 6, 5, 4, and 3 years old respectively.
Two were 2 years old; three only 1 year old. One was found in two different
boxes, the second of which was the one from which she had fledged. One,
banded as a brooding female, was brooding in the same box the following
year.
We have a five-year record on our most productive box. The female in
this box in 1973 and 1974 had been banded as a nestling in 1969. The fe-
male in the box in 1975 had been banded as a nestling in 1972 in another
area. In 1976, the female in this box was unbanded. I banded herand was
able to recapture the same female in the same box in 1977.
On 30 June 1980, I caught a small flock of bluebirds in a net. The adult
pair was unbanded. Three juveniles had fledged between 9 and 12 May
from a nearby box; two had fledged from another box on 4 June.
Tree Swallows
Population. In 1966, only one pair of Tree Swallows nested in one of our
boxes. Since then there has been a gradual increase. The average number
of pairs for 1 5 years ( 1966-1980) is nine. In 1980, there were 13 pairs. They
laid 747 eggs, of which 567 hatched. The number of young fledged fluctu-
ated from four in 1972 to 5 1 in 1980, with the total at 396 (Table 14 and Fig.
3) and the average clutch size, 5.03 (Table 15).
The breeding season computed on the basis of clutch completion ex-
tended from 10 May to 22 June. Over 73% of the clutches were completed
between 20 May and 8 June (Table 16).
In seven cases a Tree Swallow female laid a second clutch in the same
1983
EASTERN BLUEBIRD
27
box after the first was interrupted. One female swallow laid eight eggs be-
tween 10 May and 1 June. On 8 June there were eight eggs. On 19 June the
box contained one intact egg, a broken egg, and a newly hatched dead
young. In one box, four of five eggs hatched after which three of the young
disappeared. Following this, two more eggs were laid, but failed to hatch.
The remaining young fledged.
There were two cases of second broods of Tree Swallows, both in 1980.
One brood of six, which hatched on 3 J une, was dead on 1 I June ( last seen
alive on 9 June). On 19 June, presumably the same pair of adults was build-
ing in a nearby box. On 22 June, there were four eggs; three young fledged.
In the other instance, five of seven young fledged after 14 June. On 22 June
a new nest with one egg was found on top of the old. Of foureggs laid, one
hatched. The bander discovered two dead young from the first clutch un-
der the new nest.
Tree Swallow broods often include one nestling smaller than the others.
Occasionally there are two. These birds often fail to fledge, particularly if
fly parasites are present. In our study, excluding cases of total or partial
nest failure due to other causes; of two broods of seven all fledged in one,
two were found dead in the other; of 20 broods of six all fledged in I 1 , one
disappeared from or was found dead in nine; of 38 broods of five all fledged
in 23, one disappeared or was found dead in 1 1, and two disappeared or
were found dead in four.
Banding Records. Tree Swallows were not present in every area each
year(Table 2). 1 banded 37 brooding adults in 1975-1980. During the same
period, 200 young were banded. From 1966 to 1974, approximately 100
were banded (Table 14). We have had only three recoveries of these banded
birds. Two adults banded in 1975 were brooding in different areas in 1976.
One adult found dead in a box in 1974 had been banded in a box one mile
away in 1973. Although certain boxes appeared to be preferred, we have no
evidence that the same swallows return to the same nest sites year after
year.
Bluebird-Tree Swallow Competition
Tree Swallows seldom interfered with the bluebirds. The instances of
competition were too rare to be significant. The two species appear to
share the same enemies and benefit from the same measures to protect
them from House Sparrows, House Wrens, predators, human interference
and parasitic fly larvae.
28
JEAN EAKIN
No. 40
Productiveness of Boxes — Bluebird and Tree Swallow
One box, occupied by bluebirds for 12 consecutive years, produced 45
young; another occupied 10 consecutive years, 54 young. From each of
three boxes 26 young fledged. Young fledged from all but one of these
boxes in 1980. Of our 12 most successful boxes, 1 1 were still in use in 1980.
The following table shows years of bluebird occupancy.
Number of Boxes
2
2
4
5
18
19
20
Number of Years
12
1 1
10
9
8
7
4
3
2
Of 82 boxes occupied more than once, 54 were still in use in 1980. The other
boxes were removed either because they had ceased to attract bluebirds or
because of repeated bluebird nesting failures. Of 86 boxes present in 1980,
bluebirds had occupied or attempted to occupy 65; Tree Swallows, 39.
Both species had occupied 28.
The following table shows occupancy of boxes by Tree Swallows:
Number of Boxes
4
I
3
4
18
Number of Years
6
5
4
3
2
1983
EASTERN BLUEBIRD
29
No box has been occupied by Tree Swallows for more than threeconsecu-
tive years. Of the 1 2 boxes occupied for three or more years, I I were still in
use in 1980.
Natural Cavities. On two occasions, we observed bluebirds nesting in
natural cavities. In 1966 a pair was seen on 26 April, 5 May, and 9 May en-
tering and leaving a hole in a dead stump projecting from the top of an old
white oak (Querus alba) in the center of a small bog (Fig. 4, HP). On 17
May, H ouse Wrens had occupied the cavity. During June 1980 a pair nest-
ed in a hole in a sassafras approximately 20 ft above the ground ( Fig. 4,
CL).
Black-capped Chickadees
Chickadees attempted 23 nestings from 1967 through 1980. There have
been 1 1 broods. Of 108 eggs laid. 58 hatched and 55 young fledged. House
Wrens destroyed 28 eggs and interrupted two other nesting attempts.
Human interference caused the loss of three young and I I eggs. Bees inter-
rupted one nesting (Table 17).
Bewick Wrens
In July 1969 a pair of Bewick W rens attempted to nest in a box in area
CA (Fig. 4). They completed a nest ( more grass than twigs) on the 2 1 st. Ten
days later. House Wrens were in the box.
Discussion
During the 16 years of the project, the minimum numberof nesting pairs
increased from 8 in 1965 to 18 in 1980 (Table 3). From 1972 through 1976,
the number fluctuated from 18 to 23. In 1972, when there were 19 pairs, 3
were in roadside houses. In 1973, 3 of the 22 pairs were in roadside houses,
one in an area which subsequently became unsuitable for bluebird boxes. In
1 974, 3 of the 23 pairs were in roadside boxes, 2 in areas which became un-
suitable. In 1976, 2 of the 21 pairs were in roadside houses. After 1976, we
removed the remaining roadside houses because of human interference
and House Sparrow competition.
The average clutch size during the 16 years was4.22. PeakalK 1970:249)
found an average of 4. 48 for Ohio based on 63 1 nest cards (97 prior to 1964;
30
JEAN EAK1N
No. 40
the remainder from 1964 to 1969). Since he used records from various parts
of Ohio and covered a different period, some differences could be expected
(Table 1 I ).
Following Peakall’s procedure, I have based my computation of
seasonal activity upon clutch completion. Table 12 shows the number of
clutches completed in each 10-day period of each season. It also shows the
percentage of the total number of clutches for 16 years completed during
each 10-day period. In Table 13 my figures are compared with those of
Peakall Since his study covers a larger part of Ohio and ends in 1969, it is
difficult to compare his data with the present study (Peakall, 1970:245-
251).
The number of young fledged fluctuated between a low of 31 in 1967 to a
high of I 10 in 1980. The 1966 figure is based on incomplete data. In spite of
fluctuations, the Hedging trend has been upward (Table 3 and Fig. 2). Nest
failures have been reduced by measures taken to deal with specific prob-
lems.
House Sparrows. Regular removal of Elouse Sparrow nests made more
boxes available for nesting bluebirds and Tree Swallows. Both species had
successful nestings in boxes from which House Sparrow nests were re-
moved (Table 4). We eliminated House Sparrow productivity in our
boxes. The number of boxes in which nesting attempts by House Sparrows
occurred decreased (Table 5) as we removed or relocated those which at-
tracted them, often near buildings (Zeleny, 1976:74). We were not able to
keep these aggressive birds from attempting to occupy some of our boxes
or to eliminate them as occasional causes of bluebird or Tree Swallow
nesting failure (Table 4).
House Wrens. The House Wren, a native bird and a natural competitor
for bluebird nest sites, posed a more complex problem than the House
Sparrow. During the first 10 years of the project, approximately 1,300
wren young fledged from our boxes. After 1975, when we began to remove
wren nests routinely, no more young fledged. We believe that wrens will
not tolerate competition for available insects within 50 to 100 ft of their
nesting cavity. They frequently pierce the eggs in boxes near their own even
if they do not subsequently occupy these boxes. To reduce such inter-
ference we placed our boxes more than 100 feet apart. Because wrens pre-
fer nest sites near shrubbery ( Kibler, 1969: 1 19), we placed our boxes as far
into the open as was possible to do and still provide a safe landing for the
bluebird young on theirfirst flight (Zeleny, 1976:74). As we removed boxes
attractive to wrens, the number which they attempted to occupy decreased
(Table 6). We were able to reduce but not eliminate wren interference with
nesting bluebirds (Table 7).
1983
EASTERN BLUEBIRD
31
Predation. Failure due to predation seemed to increase with the increase
in the bluebird population (Table 8). Regular monitoring of the boxes may
have been a factor. From 1977 on, rectangular sheet metal guards folded
around the poles seemed to keep most would-be predators out of the
boxes. It is probably not possible to eliminate predation by raccoons.
Human Interference (Table 9). Part of the losses due to human inter-
ference can be described as theft or vandalism. Some were d ue to curiosity
and carelessness. We found that boxes fastened with Philips screws were
protected from interference by human beings who wished to remove or
merely look at their contents. Occasionally an entire box was damaged or
removed. We kept this kind of destruction at a minimum by removing vul-
nerable houses and by keeping our boxes as inconspicuous as possible.
With the exception of one dark green box, the boxes were unpainted. The
metal guards were painted dark brown.
Apaulina sialia. These larvae were present in two-thirds of our broods of
bluebirds and 154 young died in infested nests (Table 10). Rotenone, ef-
fective in 1968, 1969, and 1970, lost its effectiveness from 1972 to 1974.
Beginning in 1975, we removed larvae manually, which necessitated re-
moving the infested nest and substituting one of grass made to simulate the
original nest as far as possible. We found larvae as early as the fifth day,
when we felt that the young could be handled safely. We found it necessary
to change the nest one or two more times at two-day intervals. When this
procedure was followed, there were few if any fatalities from these para-
sites. The infested nest was placed in a mesh bag at the foot of the box.
Adult Apaulina sialia flies could not escape. The smaller Monnoniella
which parasitizes the Apaulina sialia pupae could leave or enter the bag.
We do not know if larvae pupated or if adult Apaulina sialia emerged with-
in these bags. We have no proof of the presence of Monnoniella in the
bags. We followed this procedure in an effort to permit natural control of
Apaulina sialia by Monnoniella.
Removal of the young bluebirds and returning them to a substitute nest
did not disrupt the nesting process. One removal could be combined with
banding. The fly larvae, if unchecked, caused so many fatalities in our area
that a routine nest change on the fifth or sixth day seemed to us to be a wise
precaution.
Weather. It appeared that in certain years adverse weather conditions in
northeast Ohio caused nest failures. This occurred during four years of our
study. Fortunately, most bluebird pairs when interrupted go on to a sec-
ond or even a third nesting. Eight pairs attempted a second nesting in the
same box, and five pairs were successful One pair which lost four of five
young raised two more broods in the same box. Often bluebirds seek a dif-
32
JEAN EAK.1N
No. 40
ferent box near the one in which they have had a nest failure. We believe
that in eight cases of nest failure during cold weather the pairs had second
nestings in nearby boxes. One pair apparently moved toa nearby box and,
after an interruption by House Sparrows, moved to the original box and
raised a brood. One pair probably had two successful nestings in a nearby
box.
Since we did not dissect any of the young that died during periods of cold
weather, we do not know if starvation due to a temporary dearth of avail-
able insects was a factor. Of 23 survivors. 14 were less than a week old, 5
approximately one week old, and 4 approximately two weeks old. It may
be that birds young enough to be brooded have a better chance of survival.
The 82 young that died varied in age from one day to two weeks.
Because we cannot eliminate all the causes of bluebird nest failure, we
feel that in the future the number of fledglings will continue to fluctuate. It
seems clear that humans can help the bluebird population to maintain it-
self, if not to increase.
If every portion of our study area in use in 1980 were occupied according
to its demonstrated capacity, we might expect to have 21 nesting pairs.
Bluebirds nest on some properties adjacent to the Arboretum. In two in-
stances, brooding females banded as young at the Arboretum were cap-
tured on such properties. Other bluebirds nesting in the vicinity have been
observed to be banded. It seems reasonable to expect that any population
increase will take the form of territorial expansion rather than increased
concentration on Arboretum property.
Summary
During 16 years (1965-1980) a study was made of a volunteer effort to
increase the population of the Eastern Bluebird, Sialia sialis, at the Holden
Arboretum in Lake County, Ohio. In 1965 there were 230 boxes in 14
areas. Eight boxes were occupied by bluebirds, 49 by House Wrens (Tro-
glodytes aedon ), and 134 by House Sparrows ( Passer domesticus).
During the study, five areas with 64 boxes were eliminated and three
areas with 16 boxes were added. The total number of boxes was gradually
reduced to86. In 1980 bluebirds occupied 22; Tree Swallows ( Iridoprocne
bicolor) 15; and Black-capped Chickadees ( Parus atricapillus) 2. House
Sparrows attempted to occupy 15; House Wrens, 22.
The minimum number of nesting bluebird pairs increased from 8 to 18
or 19 (Table 1 ). The number of young which fledged fluctuated from 10 in
1983
EASTERN BLUEBIRD
33
1966 to 1 10 in 1980 (Table 3 and Fig. 2).
The following measures were taken to deal with specific causes of blue-
bird nest failure:
House Sparrows. Removal of boxes that attracted only sparrows re-
duced the number of their nesting attempts. Relocation of others dis-
couraged but did not stop sparrow competition. Regular removal of spar-
row nests eliminated their reproduction in the boxes and made more boxes
available for bluebirds.
House Wrens. Removal of boxes occupied only by wrens and relocation
of others as far into the open as possible reduced the number of attempts by
this species to nest in bluebird boxes. From 1975 on, wren nests were re-
moved from the boxes. These measures reduced but did not eliminate
bluebird nest failures due to wrens.
Predators. Rectangular metal guards folded around the posts were
found to be effective against most would-be predators.
Human Interference. The use of Philips screws in the doors of the boxes
prevented nest failures caused by humans.
Apaulina Sialia. Larvae of this species were present in two-thirds of the
boxes in which bluebird young survived to the age of one week, and 154
young died in infested nests. Rotenone, effective at first, became useless as
the flies apparently developed immunity to it. We found that nest failures
caused by these parasites could be prevented by removing nests when the
young were five days old and substituting nests of grass made to simulate
the original. This procedure did not interfere with the nesting process.
The following tables show what we believe to be causes of failure of eggs
to hatch and of young to fledge in order of numerical significance.
Bluebird Eggs
Causes of Failure
Number of Eggs
House Wrens
Predators
House Sparrows
Human Interference
Clutch Failed to Hatch or Abandoned
Weather
Wasps
152
110
55-57
49
36
21
3
34
JEAN EAK1N
No. 40
Bluebird Young
Causes of Failure
Number of Young
Apaulina sialia
154
Weather
82
Predators
55
House Sparrows
33-35
House Wrens
13
Human Interference
9
Causes of Failure
Predators
House Wrens
House Sparrows
Human Interference
Cowbirds
Wasps
Bluebirds
Weather
Tree Swallow Eggs
Number of Eggs
32
31
24
14
5
5
2
2
Tree Swallow Young
Causes of Failure
Apaulina sialia
House Sparrows
Weather
Predators
House Wrens
Cowbirds
Number of Young
46
29
21
13
7
4
35
Acknowledgments
1 am indebted to Virginia Barrus who initiated Project Bluebird and to
C. W. Eliot Paine who helped to coordinate it during the early years. 1
gratefully acknowledge the cooperation of R. Henry Norweb, Jr., Director
of The Holden Arboretum, and the Arboretum staff.
Dr. Harold Mahan, Director, and Dr. Sonja Teraguchi, Entomologist,
of the Cleveland Museum of Natural History gave valuable assistance in
an advisory capacity, and Dr. Mahan edited the manuscript and made use-
ful suggestions regarding the use and arrangement of data.
Without Elizabeth E. Walker and the Arboretum bluebird volunteers,
the study would not have been possible. The photographs were taken by
Patti Buchtel of the Arboretum staff and by Elizabeth Walker.
o
oo
— v^fNiCNl
~
oo
04
O'
04
m
s©
50
SO
in
Os
00
C/2
1—
C/2
"O
o
o
C/2
00
00
C/2
01}
OX}
C/2
OX}
ft n
oo
c
3
o
00
c
3
o
E
o
C/2
00
00
O
02
LU
lu
lu
>*
LU
£
sh-
OO o
'-H-
o
o ■S
<-*=.
o
73
E
V-
t—
!- X
J3
1—
00
ftfi
1
'c
E
3
z
C/2
JD
E
3
X
e
3
Laid
O
fl
c
CJ
l—
U
u,
70
u-
o
a
X
z
Z
CL
CL
Banded 36 10 20
Total Number of Eggs Laid: 2.004
Total Number of Eggs Hatched: 1.428-30
Total Number of Young Fledged: 1 .015
TABLE 4
Effects of House Sparrows on Eastern Bluebird Box Nesters
— — rn — Tt rf m — co
— — • C 4 P\J
i/~> (N
^ oo n oo
- Os
C\l o
CnJ \0
Tf
c*~,
^ C*->
CsJ
‘o^ot^oooso — co
sOvOvOvOsOi — r- j —
O' Os Os Os 0s Os Os Os
rfiTfi^v0h000'O
r-r-r-r-r-i^-r^oo
OsOsOsOsOsOsOsOs
TOTAL 3 33-35 55-57
40
JEAN EAKIN
No. 40
TABLE 5
Hpuse Sparrow Occupancy of Eastern Bluebird Boxes
Year
Total Boxes
Number of Boxes House
Sparrows Attempted
to Occupy
1965
230
134
1966
236
179
1967
228
150
1968
204
1 12
1969
201
lb
1970
200
91
1971
200
100
1972
189
99
1973
178
72
1974
17!
77
1975
146
69
1976
100
65
1977
97
50
1978
89
40
1979
84
26
1980
86
15
1983
EASTERN BLUEBIRD
41
TABLE 6
House Wren Occupancy of Eastern Bluebird Boxes
Year
Occupied or
A t tempted
to Occupy
Earliest and
Latest Dates
Found in Boxes
Period during
which Bluebird
Eggs Were Destroyed
1965
49 of 230
No Record
1966
53 of 236
8 May, 16 July
May and June
1967
92 of 228
5 May, 18 Aug.
May
1968
9! of 204
29 Apr., 23 Aug.
May through July
1969
89 of 201
26 Apr., 27 Aug.
April through July
1970
95 of 200
29 Apr., 12 Aug.
May and June
1971
92 of 200
1 May, 24 Aug.
May through July
1972
105 of 189
21 Apr., 17 Aug.
June and July
1973
77 of 178
29 Apr., 24 Aug.
April through June
1974
93 of 171
29 Apr., 18 Aug.
May through July
1975
72 of 146
2 May, 2 Sept.
April through August
1976
43 of 100
10 May, 18 Aug.
May through July
1977
29 of 97
16 May, 18 July
May
1978
21 of 89
16 May, 13 Aug.
June
1979
18 of 84
2 May, 13 Aug.
None
1980
22 of 86
3 May, 2 Aug.
June and July
TABLE 7
Effects of House Wrens on Other Eastern Bluebird Box Nesters
^ *3 ^
“Q
A
« So o
"J4 rSV3 £;
C Q
s
h.
£ ^
0 <$ -r-
1 £? 5
^ Q
t$
* b« 5;
=2 § o'
« o £
r S ^ 2
^ Q
5j
c ^
tb ~Q
% -L
.a -c,
wj ^
5
„ — — . r*1 .— — (N — (N — \o r-- oo
sD \Q vO vO
O' C?' O' O'
Qs O — CN
sO h h h
O' O' O' O'
n ^ 'O
r- r-
0N O' O' 0>
r- oo O' o
r^- r- r- o©
O' <0 O' O'
TOTAL 1 adult
13 young
TABLE 8
Numerical Summary of Predation on Eastern Bluebirds and Tree Swallows
Year
Eastern Bluebirds
Tree 5
wallows
Adults
Young
Eggs
Young
Eggs
1965
1966
1967
4
1968
1969
4
1970
4
7
1971
2
10
9
1972
4
17
1973
1
3
9
1974
18
3
1975
1
17
23
5
S3
1976
1
8
16
4
3
1977
4
9
1978
3
4
4
1979
9
1980
2
1
TOTAL
5
55
1 10+
13
32
TABLE 9
Detrimental Effect of Eluman Interference on Nesting Success
Year
Bluebird
Young Eggs
Tree Swallow
Young Eggs
Chickadee
Young Eggs
1969
6
1970
1971
12
1972
3
1973
2
1 1
1974
15
3
1975
5
2
1976
4
1
1977
2
1978
5
1979
2
1980
5
8
TOTAL
9
49
0 14
3 1 1
44
JEAN EAKIN
No. 40
TABLE 10
Incidence of Apaulina sialia in Eastern Bluebird Nests
Year
Number of Broods
in Which Young
Survived One Week
Number Infested
With Apaulina sialia
1975
27
23
1976
18
12
1977
30
23
1978
21
9
1979
32
18
1980
27
18
TOTAL
155
103
TABLE 1 1
Computation of Eastern Bluebird Clutch Size
Year
Completed Clutches
Eggs Laid
A verage Size
1965
12
56
4.66
1966
1 1
44
4.00
1967
17
68
4.00
1968
19
74
3.90
1969
25
105
4.20
1970
23
92
4.00
1971
34
142
4.17
1972
30
131
4.36
1973
41
164
4.00
1974
41
177
4.31
1975
41
161
3.92
1976
42
176
4 19
1977
34
143
4.20
1978
25
109
4.36
1979
35
160
4.57
1980
30
140
4.66
Average clutch
size for the 16 years: 4.22.
TABLE 12
Dates of Eastern Bluebird Breeding Season Based on Clutch Completion
Os
Os
"n o
05 Os
n ^ oo
cn— — r-'iosor^fNvor-fNOo- oo
— (N M ■“ (N rn tJ- fN f'' — fN OO
<
H
iOsO^OOO'O^-CNir^, ^ VS ^ h oo av O P
sOsOsOvOsOt^t-r^r'-r^r-'t^r'-r-'-r-'G© r-
0’s Qs O' O' O' O' O' O' O' O' O' O' O' O' O' O'
46
JEAN EAKIN
No. 40
TABLE 13
Breeding Season of Eastern Bluebird:
Percentage of Clutches Completed in Each 10-Day Period
Dates
Ohio
(Peakall, 1970:246)
Holden Arboretum
1965-80
1 1 Mar. -20 Mar.
0.20
21 Mar. -30 Mar.
0.17
0.00
3 1 Mar. -9 Apr.
0.33
0.81
10 Apr. -19 Apr.
9.52
5.69
20 Apr. -29 Apr.
18.03
15.85
30 Apr. -9 May
10.52
1 1.79
10 May-19 May
7.35
6.91
20 May-29 May
7.68
7.52
30 May-8 June
11.52
9.55
9 June-18 June
12.02
11.58
19 June-28 June
10.18
1 1.78
29 June-8 July
6.18
7.1 1
9 July-18 July
3.34
4.47
19 July-28 July
2.34
4.26
29 July-7 Aug.
0.33
2.03
8 Aug. -17 Aug.
0.406
TABLE 14
Tree Swallow Nesting Data
r\
VO
X
ro
r-
O
—
X
OO
OO
X
OO
X
X
O'
r^-
os
X
Tf
X
rt
X
OO
fN
X
fN
OO
'O
Os
0
fN
x
On)
fN
00
fN
X
X
X
r-
x
r-
X
X
X
"
>
<*—
UJ
>"
X
OX)
<-•_
0
<4—
O
0
•0
2:
1—
OJ
X)
E
3
umber
[.aid
0)
X)
S
3
X
CJ
<3
c
OJ
O
X
0
<3
3C
1—
X
E
3
C
‘ob
"O
Ll
X
0
c5
X
t—
dJ
X
£
3
X
E
3
z
z
2:
CL
2
0-
2:
2
Number of Young Fledged: 396
48
JEAN EAKIN
No. 40
TABLE 15
Computation of Tree Swallow Clutch Size
Year
Completed Clutches
Eggs Laid
Average Size
1966
1
5
5.00
1967
4
23
5.75
1968
6
25
4.16
1969
7
33
4.71
1970
9
44
4.99
1971
12
63
5.25
1972
1 1
57
5.09
1973
10
54
5.40
1974
12
57
4.66
1975
15
77
5.13
1976
10
48
4.80
1977
12
67
5.58
1978
8
41
5.12
1979
9
45
5.00
1980
17
82
4.82
Average clutch
size for 15 years: 5.03.
1983
EASTERN BLUEBIRD
49
TABLE 16
Tree Swallow Breeding Season Based on Clutch Completion
A/av 30-
Year
May 10-19
May 20-29
June 8
June 9-18 June 19-28
1966
1
1967
4
1968
3
2
1
1969
5
3
1970
3
4
1
1
1971
2
7
2
1
1972
7
2
2
1973
2
6
3
1974
2
8
1
3
1975
2
7
3
1
1976
1
5
3
2
1977
5
6
1
1
1978
2
2
4
2
1979
1
6
2
1980
3
10
1
TOTAL
20
67
39
17 3
Percentages of
Clutches
Completed in
Each 10-day
Period
13.81
46.53
27.08
11.8 .69
TABLE 17
Black-capped Chickadee Nesting Success
o
o
be 7?
s? &
O Cu
— — ti-
£
o
o
.be S3
UJ 53
5;
— — tj-
Tt
3,5
uj1 -3
o rn r-
Cl
3
O
-c
o
o
a*.
QQ
(N (N O “
O O >
Cl
■g g-
cs '-5
oo
o c
W -3
-O O
>.
r- oo & o
'O \C L-
O' O' O'
— rsif^Tfi^soi^ooa^o
Os 0s O'' 0s O' 0s 0s O'
<
H
o
(-
£ -
o Q
C/3 O
u
= 3.6; axial telson spine length, »2.6.
Remarks
Eller ( 1 935) based the species Echinocaris auricula on a
well-preserved left valve from the Upper Devonian “Che-
mung” Shale at Alfred Station, in southwestern New
York state. This species has not previously been photo-
graphically illustrated. The West Virginia specimen con-
forms closely to Eller’s specimen and his description and
sketch. Because the species is readily distinguishable from
all others, there can be little doubt of its placement.
The general outline of E. auricula is different from that
of other members of the genus, particularly because of the
pronounced anterior position of the hinge and concomi-
tant prolongation of the posterior of the carapace. Addi-
tionally, it is one of only two species that possesses three
carinae on the carapace. The most closely related species
is E. castorensis Copeland, 1 960, reported from the upper
Devonian rocks of the Alexo Formation, Alberta, Can-
ada. This species, as reconstructed by Copeland (1960:
Fig. 1), is also characterized by the development of three
carinae. Upon examining the type material we have been
able to discern the centroventral ridge and the postero-
dorsal ridge on these specimens, but have been able to
discern only a suggestion of a very short posterocentral
ridge. By contrast, all three ridges are well developed and
distinct on E. auricula. The nodes on the carapace of E.
auricula are generally similar to those of E. castorensis.
The most notable difference in the nodes of these two spe-
cies is the presence of a tubercle on the centroventral lobe
of E. auricula and the lack of such a feature on the corre-
sponding lobe of E. castorensis. The general outline of E.
castorensis is also different than that of E. auricula, being
less alate and having a more centrally located hinge.
Relatively few species of the genus in addition to E.
auricula and E. castorensis have a posterocentral ridge.
These include: E. pulchra Sturgeon, Hlavin, and Kesling,
1964; E. randalli Beecher, 1902; E. socialis Beecher, 1884;
and E. whidbornei Jones and Woodward, 1889. Among
these, however, there are also other differences, such as
the degree of development of lobes, presence or absence
of tubercles, and shape of the hinge that serve to distin-
guish these species from E. auricula.
The above redescription allows for some intraspecific
variation, especially in regard to the development of tub-
ercles on the anterodorsal node and the exact nature of
the carinae. Since Eller based the species E. auricula on a
single specimen, he could not discuss variation and, in
fact, little has been done regarding intraspecific variation
26
HANNIBAL AND FELDMANN
No. 41
among members of the genus. It is certainly clear that
some variation must, indeed, have existed and the devel-
opment of minor tubercles and minute elements of orna-
mentation on the carapace must be taken as well within
the range of individual variation.
Discovery and identification of this second specimen of
E. auricula lends credence to the validity of the taxon.
Ridges, grooves, carinae, and related structures on deli-
cate arthropod skeletons are often the result of distortion
during preservation. Very often it is difficult or impossi-
ble to distinguish between actual ridged structures and
artifacts of preservation. Further, different surface struc-
tures can often be discerned, depending upon the degree
of exfoliation of the exoskeleton. The discovery of a sec-
ond specimen conforming very closely to the morphology
of the holotype not only reinforces the original descrip-
tion but also confirms that the structures described were
structures actually present on the living organism. Addi-
tionally, the West Virginia specimen provides a view of
the abdomen and telson. Unfortunately, both are crushed
substantially and very little can be said about the mor-
phology of this region except that it seems to be more or
less like that seen in better preserved specimens of some
other Echinoearis species, notably E. punctata.
Acknowledgments
John L. Carter, of the Carnegie Museum, Pittsburgh, pro-
vided the type specimen and counterpart of Echinoearis auri-
cula. Murray J. Copeland, Geological Survey of Canada, Ot-
tawa, made arrangements for the loan of type specimens of
Echinoearis from Canada. Thomas Rammer, West Virginia
University, and Kathleen Farago, Cleveland, aided in an inves-
tigation of the West Virginia locality. Murray J. Copeland and
W. D. Ian Rolfe reviewed the manuscript. The assistance of
these colleagues is greatly appreciated. A portion of this work
was supported by NSF grant EAR8312798 to Feldmann. Con-
tribution 270, Department of Geology, Kent State University.
References
Beecher, C. E. 1884. Ceratiocaridae from the Chemung and
Waverly groups of Pennsylvania. Second Pennsylvania Geo-
logical Survey PPP:!-22.
1902. Revision of the Phyliocarida from the Chemung
and Waverly groups of Pennsylvania. Quarterly Journal of
the Geological Society of London 58:441-449.
Cardwell, D. H., et al. 1968. Geologic map of West Virginia.
West Virginia Geological and Economic Survey. 2 sheets.
Cooper, G. A., et al. 1942. Correlation of the Devonian sedi-
mentary formations of North America. Bulletin of the Geo-
logical Society of America 53:1729-1794.
Copeland, M. J. 1960. The occurrence of Echinoearis and Spath-
iocaris (Phyliocarida) in western Canada. In M. J. Copeland
and T. E. Bolton, Canadian fossil Arthropoda, Eurypterida,
Phyliocarida and Decapoda. Geological Survey of Canada
Bulletin 60:1-1 1.
Dennison, J. M. 1971. Petroleum related to Middle and Upper
Devonian deltaic facies in central Appalachians. American
Association of Petroleum Geologists Bulletin 55:1 179-1 193.
Eller, E. R. 1935. New species of Echinoearis from the Upper
Devonian, of Alfred Station, New York. Annals of the Car-
negie Museum 24:263-274.
1937. Echinoearis crosbyensis , a new species from the
Upper Devonian of New York. Annals of the Carnegie
Museum 25:257-259.
Hennen, R. V., and D. B. Reger. 1914. Preston County. West
Virginia Geological Survey. 566 pp.
Jones, T. R., and H. Woodward. 1889. Notes on phyllopodi-
form crustaceans, referable to the genus Echinoearis , from
the Palaeozoic rocks. Geological Magazine, Decade III
1:393-396.
Manspeizer, W. 1963. A restudy of the Chautauquan Series of
Allegany County, New York. In V. C. Shepps, ed.. Sympo-
sium on Middle and Upper Devonian stratigraphy of Pen-
nsylvania and adjacent states. Pennsylvania Geological Sur-
vey, 4th ser.. Bulletin G39, 4th ser., pp. 259-277.
Rickard, L. V. 1975. Correlation of the Silurian and Devonian
rocks in New York state. New York State Museum and
Science Service. Map and Chart Series No. 24.
Rickard, L. V., and D. W. Fisher. 1970. Geologic map of New
York. Niagara Sheet. New York State Museum and Science
Service. Map and Chart Series No. 15.
Sturgeon, M. T., W. J. Hlavin, and R. V. Kesling. 1964. Rare
crustaceans from the Upper Devonian Chagrin Shale in
northern Ohio. Contributions from the Museum of Paleon-
tology of the University of Michigan 19:47-64.
Williams, H. S., and E. M. Kindle. 1905. Contributions to De-
vonian paleontology. Part 1: Fossil faunas of the Devonian
and Mississippian (“Lower Carboniferous”) of Virginia, West
Virginia, and Kentucky. United States Geological Survey
Bulletin 247. 58 pp.
Woodward, H. P. 1943. Devonian system of West Virginia.
West Virginia Geological Survey [Reports], 15:1-655.
LIFE HABITS AND DISTRIBUTION OF RIVERINE
LAMP SI LIS RADI A TA LUTEOLA
(MOLLUSCA: BIVALVIA)
MICHAEL J. S. TEVESZ
DAVID W. CORNELIUS
Department of Geological Sciences
Cleveland State University
Cleveland, Ohio 44115
and
J. BERTON FISHER
Research Center
Amoco Production Corporation
Tulsa, Oklahoma 74102
Abstract
Lampsilis radiata luteola is an abundant, widely distributed, freshwater
bivalve whose autecology and habitat preferences are little known.
Analysis of populations from an Ohio stream reveal that it is most fre-
quently found on compositionally mixed substrata, near shore, where
current velocities are low. Although the life position of L. r. luteola is
extremely variable, the most common orientation observed was with the
shell approximately two-thirds buried in the substratum, the anteropos-
terior axis at a 45° angle to the substratum, and the siphonal areas nor-
mal to the direction of current flow. While L. r. luteola is an active
burrower and crawler, its punctuated and variable mode of locomotion
is in contrast to the more constant and predictable life habits of many
comparably sized marine bivalves.
Biometrical studies show that stream-dwelling L. r. luteola are larger
than lake-dwelling forms. Also, the relation of shell length to age is nearly
linear for individuals three to eight years old. The paucity of bivalves less
than three years old indicates the L. r. luteola experiences sporadic re-
cruitment and frequent reproductive failures.
Introduction
Lampsilis radiata luteola ( Lamarck, 1819) ( Unionidae)
is one of the most widely distributed freshwater bivalves
in North America. It occurs throughout the Mississippi
and Missouri river systems, much of Canada east of the
Rocky Mountains, the St. Lawrence drainage, and the
Atlantic slope south to South Carolina. In addition, L. r.
luteola is extremely widely distributed in aquatic envi-
ronments within its geographic range. It is found in
streams, ponds, lakes, and other wetland areas on sub-
strata ranging from muds to gravel. Moreover, L. r. lu-
teola is frequently the most abundant bivalve in these en-
vironments (Burch 1973; Clarke 1973) and is a dominant
organism in terms of standing crop and biomass. For ex-
ample, unionid bivalves have the largest standing crop
(82.5 g/m2 wet) and individual biomass (average individ-
ual wet weight is about 10.5 g) of any invertebrate group
in western Lake Erie, and L. radiata is the most abundant
unionid (Wood 1953).
It is therefore surprising that very little ecological in-
formation has been published concerning this remarkable
species. The literature concerning lentic populations
mainly focuses on behavioral characteristics and biogenic
modification of sediments (e.g., McCall et. al. 1979;
McCall and Tevesz 1982). And with the exception of
communications by Whittine (1969) and Salmon and
Green ( 1 983) there is very little published information on
the ecology and distribution of L. radiata in lotic
environments.
This paper is an attempt to begin building a body of
knowledge concerning the autecology and distribution of
lotic populations of L. r. luteola. In this paper, we provide
new information on the distribution, biometrics, age,
substratum preference, and life habits of L. r. luteola. The
Vermilion River, Ohio, was chosen as the study area be-
cause it is a relatively clean, accessible river, which, like
many others, contains large populations of Lampsilis.
The Vermilion River is 58.7 mi in length, has an average
fall of 7.8 ft/mi, and drains an area of 271.7 sq mi (Ohio
Division of Water 1954: Fig. 1). The river flows through
mostly rural countryside, where its chief pollutants are silt
and fertilizers from farms. The presence of a variety of
pollution-intolerant organisms (e.g., stonefly and mayfly
nymphs; caddis fly larvae) in the river over much of its
length indicates that the river is mostly reasonably clean
(Beck 1954; Gaufin and Tarzwell 1952, 1956).
Methods
Live and freshly dead L. r. luteola and associated
Unionidae were collected in summer months during
1 975-82 at thirteen stations along the Vermilion River by
bank-combing and wading (Fig. l;Stas. 1-7, 9-14). The
general characteristics of the substratum were recorded
for each station. In addition, four more stations (Fig. 1;
Stas. 8, 15-17) were sampled in detail by wading and
Fig. 1. Vermilion River. Scale (inset): ! in. = 5 mi.
1985
LAM PSILIS RADIATA LUTEOLA
29
Key to Figure 1
COLLECTING STATIONS
1 . Location: M ud Lake. At the intersection of
U.S. Rt. 250 and U.S. Rt. 224, turn SE and
travel 6.2 mi to the first right turn after
Crum Rd., then due E 0 I mi; Ashland Co.,
Oh.
2. From the intersection of U.S. Rt. 250 and
Crum Rd., go W on Crum 0.5 mi to St. Rt.
545. Turn right onto 545 heading to an
overpass 0.4 mi NE; Ashland Co., Oh.
3. From the intersection of U.S. Rt. 250 and
U.S. Rt. 224, take U.S. Rt. 250 SE for 5.0
mi to Clear Creek Rd. in Savannah, Oh. Go
W on Clear Creek Rd. 0.8 mi to an overpass;
Ashland Co.. Oh.
4. From the intersection of U.S. Rt. 250 and
U.S. Rt. 224, take U.S. Rt. 250 SE 3.0 mi to
Base Line Rd. Take Base Line Rd. W 0.5 mi
to a steel truss span bridge; Ashland Co.,
Oh.
5. From the intersection of U.S. Rt. 250 and
U.S. Rt. 224. take U.S. Rt. 250 SE 2. 1 mi to
river overpass; Ashland Co., Oh.
6. From the intersection of U.S. Rt. 250 and
U.S. Rt. 224, take U.S. Rt. 224 E 0.1 mi to
river overpass; Ashland Co., Oh.
7. From the intersection of U.S. Rt. 250 and
U.S. Rt. 224, take U.S. Rt. 250 N W 3.0 mi
to Town Line Rd., then E on Town Line
Rd. to a curve which parallels the river bed
below, then hike 0. 1 mi up river; Ashland
Co., Oh.
8. From the intersection of U.S. Rt. 250 and
U.S. Rt. 224, take U.S. Rt. 250 NW 2.8 mi
to steel truss span bridge, then upstream 0. 1
mi from bridge; Fluron Co., Oh.
9. From the intersection of U.S. Rt. 250 and
Fitchville River Rd., take U.S. Rt. 250 E 0. 1
mi to bridge, and from there to 200 m up-
stream; Huron County, Ohio.
10. From the intersection of U.S. Rt. 250 and
Fitchville River Rd., take Fitchville River
Rd. N 1.1 mi to Fayette Rd. Then take
Fayette Rd. E 0.6 mi to a steel truss span
bridge just W of Palmer Rd; Huron Co.,
Oh.
1 1. From the intersection of U.S. Rt. 250 and
Fitchville Rivei Rd., take Fitchville River
Rd. 5.3 mi N to Prospect Rd. Then take
Prospect Rd. E 0.5 mi to steel truss span
bridge; Huron Co., Oh.
12. From the intersection of U.S. Rt. 250 and
Fitchville River Rd., take Fitchville River
Rd. 8.2 mi N to Cook Rd. Take Cook Rd.
SE 0.4 mi to a steel arch-truss span bridge;
Huron Co., Oh.
13. From the intersection of Fitchville River
Rd. and Cook Rd.. take Fitchville River
Rd. 1.1 mi N to Zenobia Rd., then take Ze-
nobia Rd. E 0. 1 mi and turn S (right), fol-
lowing this road around a curve to a steel
span bridge; Huron Co., Oh.
14. From the intersection of Fitchville River
Rd. and Zenobia Rd., take Zenobia Rd. 0.7
mi E to a bridge; Huron Co., Oh.
15. From the intersection of St. Rt. 60 and St.
Rt. 116 in Birmingham, Oh., take St. Rt.
1 16 E 1.3 mi to Gore Orphanage Rd. Then
take Gore Orphanage Rd. N 1.8 mi to
bridge; Lorain Co., Oh.
16. From the intersection of St. Rt. 60 and St.
Rt. 1 16, take St. Rt. 1 16 E 1.3 mi to Gore
Orphanage Rd.. then take Gore Orphanage
Rd. N 2.4 mi to Morse Rd. Follow Morse
Rd. to Bank Rd , then take Bank Rd. S 0.9
mi until it ends. From the dead end, hike
due SW circa 0.5 mi to river; Lorain Co.,
Oh.
1 7. From the S end of Bank Rd., hike 0.8 mi NE
to river; Lorain Co., Oh.
SCUBA for live individuals only, and the following in-
formation was recorded where each individual was found:
substratum type (estimated as gravel, sand, or mud, or
some combination of these); distance from shore; current
velocity (measured with a submerged float and stop-
watch); inclination of the anteroposterior axis of the bi-
valve with respect to the substratum surface (estimated as
either 0, > 0< 45°, 45°, > 45° < 90°, or 90°); orientation
of posterior with respect to direction of current (measured
with protractor); and fraction of shell buried in substra-
tum (estimated as> 0. 15, 0.25, 0.33, 0.50, 0.67, or> 0.75).
Over 1 30 live L. r. luteola were collected and the follow-
ing data were taken in the laboratory: total length (great-
est linear dimension parallel the hinge); anterior length
(portion of total length anterior of the mid-point of the
umbo); height (greatest linear dimension normal to length
and in the same plane); and width (greatest linear dimen-
sion across both valves normal to length and width). Ab-
solute age determinations were made by cutting a valve in
two with a rock saw, polishing one of the sections, and
counting the annual markings under a microscope. Tax-
onomic identifications were provided by Dr. David H.
Stansbery, the Zoology Museum, Ohio State University,
Columbus, Ohio.
Distribution
Areal
The presence or absence, at 17 sampling stations, of
L. r. luteola and associated Unionidae, is represented in
Table 1 (see also Fig. 1). The predominant substratum
30
TEVESZ, CORNELIUS, AND FISHER
No. 41
TABLE 1
Occurrence of Lampsilis radiata luteola at 17 sampling stations
Station
No.
Predominant
Substratum Type
Sampled Bivalves
Lampsilis
radiata
luteola
(Lamarck, 1819)
Lampsilis
ventricosa
(Barnes, 1823)
Anodonta
grandis
grandis
(Say. 1829)
Strophitus
undulatus
undulatus
(Say, 1817)
Lasmigona
costata
( Rafinesque,
1820)
Fusconaia
flava
( Rafinesque,
1820)
Elliptio
dilitata
( Rafinesque,
1820)
Anodontoides
ferussacianus
(Lea, 1934)
I
mud
X*
2
mud /sand
X*
3
sand /mud
X
X*
4
mud/sand/gravel
X*
5
mud
X*
6
gravel
X*
X
7
sand/mud, with gravel
X
X
X*
X
X
8
gravel, with pockets of
X*
X
X
X
sand/ mud
9
bedrock/ gravel
X*
X
X
10
gravel
X*
X
X
11
gravel
X*
X
X
X
X
X
12
gravel
X*
X
13
sand /gravel
X*
X
X
X
X
14
bedrock/ gravel
X*
X
X
X
15
gravel, with pockets of
X*
X
X
sand/mud
16
gravel, with pockets of
X*
X
X
sand /mud
17
gravel, with pockets of
X*
X
X
X
sand /mud
♦Indicates most abundant species at particular site.
type found at each station is also listed there. The table
and figure demonstrate that Anodonra grandis grandis is
the numerically dominant unionid at the sampling sta-
tions in Mud Lake and at the source area of the river. But
below Station 5 the numerically dominant species at the
collecting locales is generally L. r. luteola , although Las-
migona costata is the most abundant species at Station 7.
Additionally, Table 1 reveals correlations between lacus-
trine-conditions/muddy-substrata and the presence of a
relatively species-poor fauna dominated by A. g. grandis
(one to two species). Typical fluvial conditions and
coarser substrata, on the other hand, appear to be asso-
ciated with a relatively species-rich fauna (two to six spe-
cies) generally dominated by L. r. luteola. Also correlated
with this downstream increase in species richness is an
increase in substratum heterogeneity.
The fact that we have yet to discover Lampsilis living
on soupy muds in the Vermilion River is interesting, be-
cause it occurs with A. g. grandis on similar bottoms in
Lake Erie (McCall et al. 1979; Tevesz and McCall 1979).
This difference in distribution may be related to the
smaller size and thinner shells of lake-dwelling forms (cf.
Clarke 1973; Harman 1970) which would make them less
prone to sinking into the substratum.
Within- Habitat
Table 2 provides more detailed information on the sub-
stratum distribution of L. r. luteola and the five unionid
species (combined) collected live at Stations 8, 15, 16, and
17. Of three “pure” substratum categories, L. r. luteola
was most abundant on sand, compared to mud or gravel,
was most frequently found on substrata containing a mix-
ture of particle sizes, and was particularly common on a
mud/sand mix and a mud/sand/gravel mix. The pattern
of substratum preference also obtains for the remainder
of the Unionidae observed at these stations.
Information on the distance from shore and associated
current velocities of all live-collected Unionidae at Sta-
tions 8,15,1 6, and 1 7 is presented in Tables 3 and 4. Over
60% of L. r. luteola and other Unionidae are found within
1 m of stream bank as shown in Table 3. Over 75% of the
live-collected Unionidae, including L. r. luteola, were
found where current velocities were less than 5 cm/sec
TABLE 2
Substratum Distribution (%)
Mud
Sand
Gravel
M/S
G/S M/S/G
Lampsilis
radiata
luteola
0
16
3
32
15
34
n = 134
Other
Unionidae1
0
18
15
25
13
29
n = 55
1 Lampsilis ventricosa; Elliptio dititata ; Lasmigona costata; Strophitus
undulatus urtdulatus; Anodontoides ferussacianus.
1985
LAM PS I LIS RADIATA LUTEOLA
31
TABLE 3
Distance from Shore
Distance
from shore
(m)
Lampsilis
radiata
luteola
(%)
n = 136
Other
Unionidae
(%)
n - 52
1
61.8
61.5
2
14.7
21.2
3
8.8
7.7
4
7.4
5.8
5
6.6
3.8
6
0.7
0.0
TABLE 4
Current Velocity Preference
Lampsilis radiata
Other
Current Velocity
luteola (%)
Unionidae (%)
(cm sec)
(n - 136)
(n = 52)
<5
77.9
75.0
>5 but <10
13.2
11.5
>10 but <15
2.2
3.8
>15 but <20
0.0
3.8
>20 but <25
0.0
0.0
>25 but <30
6.6
5.8
(Table 4). Thus,
L. r. luteola appears to show definite
preferences for certain microhabitats within the river at
the sampling sites
. It was most frequently found on com-
positionally mixed substrata near shore, where current
velocities are low.
The same statement is true for the other
live-collected Unionidae as a group. The within-habitat
distribution of all these live-collected Unionidae at the
sampling stations is highly similar.
Life Position and Locomotion
Table 5 presents information on the life position of L. r.
luleola collected in situ. While life position of L. r. luteola
is extremely variable, the most common orientation re-
corded was with the shell approximately two-thirds buried
in the substratum, the anteroposterior axis at a 45° angle
to the substratum, and the siphonal areas normal to the
direction of current flow.
The life position and locomotory behavior of L. r. lu-
teola from the Vermilion River and adjoining areas of
Lake Erie were also observed in laboratory microcosms
containing simulated native substrata and regulated to
ambient temperature. The burrowing sequence of L. r.
luteola observed for this laboratory study from both
stream and lake population is similar, and is described by
McCall et al. (1979). Individuals from both the stream
and lake are active semi-infaunal crawlers.
TABLE 5
Life Position of L. r. luteola
Orientation of posterior (siphonal) area with respect to current
(% of population). 0° — area oriented directly into current
0° 45° 90° 135° 180° n = 1 33
23 14 40 11 13
Orientation of anteroposterior axis with respect to sub-
stratum (% of population).
0° >0° < 45° 45° >45° <90° 90° n=136
9 18 29 20 24
Fraction of shell embedded in substratum (% of population).
n = 136
< 15 .25 .33 .50 .67 >.75
8 5 18 10 51 8
Burrowing rate indices (BRI’s) (Stanley 1970) for L. r.
luteola are presented in Table 6. These data show that the
BRI of L. r. luteola is variable, but by comparison to ma-
rine bivalves of roughly similar size and shape, Lampsilis,
over all, is a slow burrower (cf. Stanley 1 970). In addition,
the burrowing rate is dependent on the type of substratum
the clams inhabit. The rates on fine substrata are two to
four times higher than rates on coarse substrata. This re-
lation of BRI to substratum type is otherwise unreported
for freshwater bivalves.
The burrowing and crawling behavior of L. r. luteola
was frequently interrupted by long periods of stasis. Usu-
ally there was a lag time of several minutes to several days
before animals placed in the laboratory microcosms be-
gan locomotory behavior. The time involved to complete
a burrowing sequence also took several minutes to several
days, depending on the number of pauses, and duration
and number of interspersed semi-infaunal crawling epi-
sodes. This punctuated and variable mode of behavior is
in contrast to the more constant and predictable habits of
many shallow-water marine bivalves of comparable size
studied by Stanley (1970). Comparable information for
other freshwater species is scarce.
The meager burrowing ability of Lampsilis, coupled
TABLE 6
Burrowing Rate Indices
Vmass (g) \
7 : : : X 100
burrowing time (sec.) /
Bivalve 1 Bivalve 2 Bivalve 3 Bivalve 4
(Length - (L - 69.8 (L = 62.4 (L=92.9
Substratum 56.1 mm) mm) mm) mm)
mud 0.13; 0.13
coarse sand 0.057 0.049; 0.057
50% sand / 50% 0.031; 0.036;
gravel 0.062
33% mud/ 67% 0.13
sand
TABLE 7
Lampsilis radiata : Comparative Morphometries
bo
7
>,
o
>>
ON
A
~ +1 ^
5 Er
t: ^
On ~C)
° o
5
5
Uj
V)
5 3 So ^
2- Co
■ 3: '-J +1
§
■S E
^ +1
■5 E
^ +1
"? tjJ
C Co
^ +1
Co
g,
Co
o
o
+1
o
CM
o
o
©
+1
o
Co
SO
©
+1
SO
o
o
©
+1
SO
o
o
+1
l/~>
so
o
o
+1
so
Tt OC
© ©
+1 +1
r- —
+1
+1
0 i
c
.2
I
>
so
CM
©
©
+1
7t
©
©
+1
o
so
©
o
o
+1
OS
CM
o
©
©
+1
Os
o
o
+1
o
4-1
uo
sO
O
— t/7
4-1
o
ro —
O
4-1
o
o
4-1
o
m rn —
— U~)
4-1
o
tj-
o
-H
o
• — co
O
4-1
o
CM
oo —
St
cd
C
o
o rx
x: £
O fe
>
3
X>
o3
X
C/3
<
^
5 2
o
. .
(U 7
x 6
CO
o3
y o
C/3
03 o>
*J C
sO
O
■ — IC'J
sd rf
—
■Tf
r-
CM
—
CM
oo
o
O
m
Os
CM
4-1
4-1
4-1
4-1
4-1
K
sd
OO
—
o
o
o
o
oo
sO
V~)
»/4
r* 4
o
— ■
Os
Os
Os
Os
Os
7f
OO
Os
CM
7f
7t
CM
oo
V~)
CM
m
oo
sO
CM
O
r-
—
>r
o3
CQ
i—
O
X)
.2
’ C
j—
u.
c3
W
c5
UJ
cd
C/3
W
0)
>
X
(U
33
aj
0)
< L)
s
w
o3
j;
cd
U
03 .C3
s ~s
u
Sr
"O ^
C oo
03 rc
y Os
i3
U
c
£
o
CD
"-5 ^
5 1L
R 5
Os O
— 3
w Cr
a3 ^
t— •—
o3 ^
c ^
c3
>
u
II
1985
LAM PS I LIS RADI A TA LUTEOLA
33
with its preference for near-shore, shallow-water habitats
associated with low-flow velocities, suggests that this bi-
valve is not frequently physically removed from the sub-
stratum by erosion. Further evidence of this is the absence
of appreciable physical wear on almost all live-collected
L. r. luteola, compared to the extensively abraded ap-
pearance of loose valves collected on the sediment sur-
face. L. r. luteola does not selectively reduce the forces
exerted on it by currents. In fact, the preferred orientation
is normal to flow. This suggests that the organisms are
passively oriented by (low much like a flat rock.
Biometrics
Morphometrical information for L. r. luteola from
the Vermilion River and other environments is pre-
sented in Table 7. As shown here, L. r. luteola from the
Vermilion River are morphometrically similar to other
Lampsilis radiala riverine populations and those from ar-
tificial impoundments. Further, these riverine/ impound-
ment populations tend to have larger average dimensions
than those from naturally occurring lakes. These observa-
tions support the findings of several studies (cf. Clarke
1973; Harman 1970) who report that L. r. luteola and sev-
eral other Unionidae are typically larger in streams than
lakes.
Additionally, comparative figures on the upstream
(Sta. 8) versus downstream (Stas. 15-17) populations of
L. r. luteola support the findings of Ortmann (1920) in
that they show this species does not follow the “Law of
Stream Distribution.” It should be noted, however, that
the “Law of Stream Distribution” was mainly derived
from study of large rivers in which habitat changes from
shallow, rapid-flow environments to much deeper, slower
-flow environments. The Vermilion is not such a river.
Age Structure
The age distribution of live-collected L. r. luteola is
shown by the data in Table 8. Age data for all live-
collected Unionidae are given in Table 9. The mean age of
all live-collected L. r. luteola is 5.55 years, and few indi-
viduals less than three years of age were observed. It is
possible but doubtful that small individuals were missed
during this study, because the length of three-year-old in-
dividuals is only slightly less than 9 cm (Fig. 2) which is a
size readily collectable by our techniques.
Individuals between three and eight years of age show a
nearly linear relation of age to mean length, as shown in
Figure 2. Younger and older individuals are much smaller
than would be predicted by the curve in Fig. 2, which indi-
cates that the clams grow most during their second year of
life and grow much more slowly after reaching eight years
of age. The absence of young individuals has been noted
TABLE 8
Age Distribution (%) of Lampsilis radiala luteola
Aye (in years)
Percent of Individuals
(n = 133)
1
1.5
2
1.5
3
9.8
4
15.0
5
25.6
6
21.1
7
11.3
8
12.0
9
1.5
10
0.8
TABLE 9
Mean Age of Live-Collected Individuals
Species
Mean Aye
Number of
Individuals
L. r. luteola
5.55
129
L. ventricosa
5.87
23
E. dilitata
5.90
1 1
L. costata
5.72
1 1
A. ferussacianus
5.20
5
S. undulatus
4.67
3
for Elliptio complanata in lacustrine environments ( Fisher
and Tevesz 1976; Matteson 1948).
The observed age structure for L. r. luteola suggests
that recruitment is not continuous. This could be due to a
number of factors Matteson ( 1948) noted that environ-
mental stresses such as low dissolved oxygen and elevated
temperatures can result in the death of both newly settled
young and gravid females of E. complanata. as well as the
expulsion of any maturing glochidia from gravid females.
For L. r. luteola. males and females are present in approx-
imately equal numbers; there does not appear to be evi-
dence for differential mortality. It should also be noted
that since L. r. luteola depend upon fish for the dispersal
of their parasitic larvae, factors affecting the biology of
the host organisms may be important in regulating the age
structure of L. r. luteola. Detailed surveys covering a rea-
sonably long period of time would be required to resolve
this problem.
References
Beck, W. M., 1954. Studies in stream pollution biology. Quar-
terly Journal of the Academy of Sciences 17:21 1-227.
Brown, C. J. D., C. Clark, and B. Gleissner. 1938. The size of cer-
tain naiades from western Lake Erie in relation to shoal expo-
sure. American Midland Naturalist 19:682-701.
34
TEVESZ, CORNELIUS, AND FISHER
No. 41
Age (yrs.)
Fig. 2. Age/ mean-length relationships for 3- to 8-year-old L. r. luteola.
Burch, J. B. 1973. Freshwater unionacean clams of North Amer-
ica, biota of freshwater ecosystems. Identification Manual
No. 1 1. EPA, U.S. Government Printing Office. 176 pp.
Clarke, A. H. 1973. The freshwater molluscs of the Canadian in-
terior basin. Malacologia 13:1-509.
Cvancara, A. M. 1970. Mussels (Unionidae) of the Red River
Valley in North Dakota and Minnesota, U.S. A. Malacologia
10:57-92.
1972. Lake mussel distribution as determined with
SCUBA. Ecology 53:154-157.
Cvancara, A. M., and P. G. Freeman. 1978. Diversity and dis-
tribution of mussels (Bivalvia: Unionacea) in a eutrophic res-
ervoir, Lake Ashtabula, North Dakota. Nautilis 92: 1 -9.
Fisher, J. B., and M. J. S. Tevesz. 1976. Distribution and pop-
ulation density of Ellipio complanata (Mollusca) in Lake
Pocotopaug, Connecticut. Veliger 18:332-338.
Gaufin, A. R., and C. M. Tarzwell. 1952. Aquatic invertebrates
as indicators of stream pollution. Public Health Reports
67:57-64.
1956. Aquatic macro-invertebrate communities as indi-
cators of pollution in Lytle Creek. Sewage and Industrial
Wastes 28:906-924.
Harman, W. N. 1970. New distribution records and ecological
notes on central New York Unionacea. American Midland
Naturalist 84:46-58.
Matteson, M. R. 1948. Life history of Elliptio complanatus
(Dillwyn, 1817). American Midland Naturalist 40:690-723.
McCall, P. L., M. J. S. Tevesz. 1982. The effects of benthos on
physical properties of freshwater sediments. In P. L. McCall
and M. J. S Tevesz, Animal-sediment relations, 105-176.
New York: Plenum Press.
McCall, P. L., M. J. S. Tevesz, and S. F. Schwelgien. 1979.
Sediment mixing by Lampsilis radiata siliquoidea (Mollusca)
from western Lake Erie. Journal of Great Lakes Research
5:105-111.
Ohio Division of Water. 1954 Gazetteer of Ohio streams. Ohio
Department of Natural Resources, Division of Water. 1 75 pp.
Ortmann, A. E. 1920. Correlation of shape and station in fresh-
water mussels (naiades). Proceedings of the American Philo-
sophical Society 59:269-312.
Salmon, A., and R. H. Green. 1983. Environmental determi-
nants of unionid clam distribution in the Middle Thames
River, Ontario. Canadian Journal of Zoology 61:832-838.
Stanley, S. M. 1970. Relation of shell form to life habits of the
Bivalvia (Molluscs). Geological Society of America Memoir
125. 494 pp.
Tevesz, M. J. S., and P. L. McCall. 1979. Evolution of substra-
tum preference in bivalves (Molluscs). Journal of Paleontol-
ogy 53:112-120.
Whittine, A. H. 1969. Distribution of unionids in Hinkley
Creek, Portage County, Ohio. Compass 46:142-149.
Wood, K. G. 1953. Distribution and ecology of certain bottom
living invertebrates of the western basin of Lake Erie. Ph D.
diss., Ohio State University, Columbus.
THE PREHISTORIC OCCUPATION OF THE HALE FARM
BATH TOWNSHIP SUMMIT COUNTY OHIO
DAVID S. BROSE
Head Curator of Archaeology
The Cleveland Museum of Natural History
Abstract
A prehistoric occupation at the Hale Farm site (33Su 1 7) underlay the
1809-27 occupation of a broad tributary floodplain of the Cuyahoga
River. During the historical archaeology, four aboriginal pits/hearths
and possible houses represented by two single post arcs of 5 m diameter
were encountered in 36 m2 excavated over a 1 50 m2 area. These features,
and the undisturbed strata in places, yielded a wide range of lithic tools
and debitage with over 60 projectile points, intermediate between the
styles of the ninth and fourteenth centuries A.D.. At least 1 1 handmade
ceramic vessels were represented by over 500 grit-tempered, cord-
marked and smooth sherds. These simple ceramic jars had rims with
interior cordmarking, and flat lips which were either plain or notched.
The ceramics are ancestral to the earliest Whittlesey ceramic types of the
thirteenth century. Faunal and floral analyses indicate a late summer
campsite with some evidence for maize and squash agriculture.
Introduction
In 1971 and again in 1980, archaeological investiga-
tions were undertaken at the Western Reserve Historical
Society’s Jonathan Hale Farm and Village in Bath Town-
ship, Summit County, Ohio.
The 1971 archaeological excavations, between June 9
and August 14, were supported by a grant from the Na-
tional Endowment for the Humanities under my direction
(H-4982). Work in the field was supervised by David
Frayer and George Miller. Full-time crew members in-
cluded Bert Barnard, Bruce Palmer, and Larry Ruben-
stein. Part-time crew included A1 Hugley, Meg Conley,
Elizabeth Coppedge and Harvey Yates (see Plate I).
The limited 1980 excavation was performed by Donna
L. Benson as part of her Ph.D. research. As Benson’s dis-
Pl. I. 1971 excavations of the Hale Farm site view to east from 1827
house to Oak Hill Road.
sertation committee chairman, this work, too, was done
under my overall direction.
The focus of these archaeological investigations was
the historic occupation of the early nineteenth century. It
was my aim to study the structural and material differ-
ences between the Connecticut Land Company “settlers”
and the non-landowning “squatters” whom the settlers
frequently encountered occupying the best homesite in
their often vast properties. We also hoped to learn some-
thing about the rates of the economic effects and the
material consequences due to opening the Ohio-Erie
Canal through the Western Reserve in the 1820s. The
Hale Farm had been occupied for several years by Abra-
ham Miller when Jonathan Hale arrived from Connecti-
cut in 1810. Hale continued to live in Miller’s cabin until
1826/1827. With the opening of the canal. Hale pros-
pered. After solidifying his economic position, he built
and occupied the large brick house now on the property,
but continued using the cabin as an outbuilding until after
1 840. Excavation of the original cabin and analyses of the
artifacts it contained seemed to offer an ideal opportunity
to begin the study.
The expectation concerning what archaeology could
add to the history of the early Western Reserve era largely
met with success. Although the final report of the
1971/ 1980 excavations has not yet been published, sev-
eral articles have detailed various aspects of that archaeo-
logical excavation (Benson 1978, 1980; Brose 1973a;
Brose and Benson, 1985; Brose et al. 1981; Miller 1980,
1983).
An unexpected result of the investigation was the dis-
covery that not even Abraham Miller had been the first
inhabitant. In 26 units which were excavated, we encoun-
tered the remains of a prehistoric Indian campsite which
had been abandoned and buried nearly 500 years before
Columbus had set sail. To a degree. Miller and Hale, in
their continual historic occupations of the area, had re-
moved or disturbed much of this archaeological site
which lay beneath their feet. Those storage pits, brick
kilns, roads, cellars, and foundations which we had been
expecting to find had often been dug through the thin
prehistoric stratum.
Of course, our digging for historical information was
carried deeply to recover all of the prehistoric remains in
any area where we worked. Yet our excavations were lim-
ited; indeed, we dug into about 20% of the area over which
prehistoric remains were found. Nonetheless, enough of
the aboriginal camp remained relatively undisturbed that
36
DAVID S. BROSE
No. 41
it was possible to carefully reconstruct many aspects of
the economy, the technology, and the social lifestyles of
these ancient Americans. These I have described in this
report, and 1 will attempt to show when and with whom
early people of Hale’s Farm were in contact and how these
small self-reliant villages stood at a threshold: their ances-
tors were the scattered hunting-gathering-gardening
bands who had built the mounds and earthworks across
Ohio in the first centuries of our era. Their descendants
built the agricultural tribal confederations encountered
by the European explorers of the seventeenth century.
The Environment
The prehistoric Hale Farm site (33Su 1 7) is located in
the front yard of the present Jonathan Hale Farm on Oak
Hill Road, lot I 1, township 3, range 12, in Bath Town-
ship, Summit County, Ohio. The site area was estimated
from the recovery of prehistoric artifacts and from the
thin, discontinuous patches of prehistoric strata and pits.
From the north bank of Hale Run the prehistoric site ex-
tended about 30 m northward to what had been a smaller
east-flowing branch of Hale Run. East of the present
1826/27 Hale House the prehistoric materials appeared
concentrated in a roughly 30 to 45 m zone centering about
25 m west of modern Oak Hill Road. A few prehistoric
ceramic sherds and one broken chipped stone projectile
point were recovered washing out along the north bank of
Hale Run about 30 m east of the road but the intervening
area had been disturbed by the clay pits and brick burning
activities of Hale himself in 1826. Within these 1378 m2,
between the branches of Hale Run, the topography is rel-
atively level (see Fig. 1), sloping genty to an old terrace
east of the road, and further leveling out. The geological
deposits in this portion of the Cuyahoga Valley have been
actively reworked by glacial ice, by lakes of ponded melt
water from retreating glaciers, and for the past 10,000
years the region has been controlled by the Cuyahoga
River which flowed first south and then north along sev-
eral different channels (Brose et al. 1981; Hall 1980; Rau
1968; White 1953a, 1979; Whiteand Totten 1982; Wittine
1970).
The slopes behind the Hale Farm are composed of
rather level Mississippian and Pennsylvanian shales and
sandstone formations, deposited in near-tropical seas
about 200 million years ago. Over the past million years,
the area was scoured by glacial ice, and filled by clays,
sands, and gravels as the ice melted. The last glacial ice left
large gravel ridges called “kames” along its valley mar-
gins. Just southeast of the Hale Farm a series of hills were
formed when a channel of the lower Cuyahoga flowed
south along the western valley wall to join the upper river
at Akron about 1 4,000 years ago. With lowered water lev-
els in Fake Erie about 3,000 years later, the lower Cuya-
hoga flowed north along the east side of the present val-
ley. It captured the upper Cuyahoga at Cuyahoga Falls.
The old Oak Hill Road channel, filled with clays, silts,
and gravels, was covered by a veneer of sandy soils
washed from the western valley slopes. Furnace Run and
Yellow Creek began to build a level floodplain by sea-
sonal flooding, while they, with some even smaller tribu-
taries, began to dissect the old glacial deposits into
smaller and less steep hills and terrace remnants. By 4000
b.c. Fake Erie and its tributary valleys had reached their
modern configuration. By that time also the modern post-
Glacial climatic patterns and the developed soil associa-
tions had established the mixed hardwoods forest com-
munities which were recorded by the first surveyors
(Andreas 1980; Gordon 1966, 1969).
The Hale Farm site is located at the juncture of several
different soil associations. The prehistoric site itself lies
upon nearly level areas of Fitchville silt loams. Small
areas of Orrville silt loams and moderately eroded
Glenford-Geeburg silt loam and Furay loam soils occur
on the slightly steeper knolls and terrace formations to the
north and south along Oak Hill Road. On the steep slopes
to the west of Hale Farm, soils are predominantly those of
the Ellsworth-Mahoning association. Along the Hale
Run valley east of the Kame Terrace and knolls east of
Oak Hill Road are the poorly drained soils of the
Conotton-Oshtemo complex. Over most of the level por-
tions of the floodplain between Hale Run and Furnace
Run, east to the Cuyahoga, the soils are Glenford and
Chagrin silt or Chili silt loam, while between Hale Run
and Yellow Creek most floodplain soils are either gravelly
Chagrin silt loam or the slightly alkaline Chili silt loam
( Ritchie and Steiger 1 974). J’hese soils are of a markedly
different agriculture potential. In an unimproved state
their yield in bushels of corn per acre would run from over
90 for the Chagrin loam to 70 to 75 for the Fitchville-
Orrville silt loam, to less than 50 for the Ellsworth soils
(Ritchie and Steiger 1974). Differences in the quality of
soils were certainly important to Jonathan Hale, and may
also have been significant for the location of the prehis-
toric site itself.
The typical soil profile of the 2% to 6% Fitchville soils
which underlie the Hale Farm site is described as having a
18 cm dark grayish-brown plow zone layer of slightly
acid, well-drained silt loam; about an 8 cm layer of light
yellowish-brown, strongly acid silt loam; about 20 cm of
yellowish-brown and iron-stained acidic, poorly drained,
silty clay loam; and a final soil layer of as much as 80 cm of
brown, moderately acid silt loam. Below this 150 cm
depth lay the sands and gravels of the Pleistocene lake and
river deposits. (Ritchie and Steiger 1974:78). In general
this soil is seldom flooded and moderately well-drained. It
presents little problem for occupation.
At the time of its first prehistoric habitation, the Hale
Farm site would probably have represented a small,
rather open and grassy area with scattered yellow poplar,
white oak, and hickory. To the west the slopes would have
Fig. 1. Location of the Hale Farm site (33Sul7).
38
DAVID S. BROSE
No. 41
been covered with a beech-maple forest with hemlock in
the ravines. To the east the lower flat floodplains would
have had a mixed forest with elm, ash, yellow poplar, lo-
cust, walnut, and basswood, while the old Kame Terrace
hills would have been covered with black oak, red oak,
beech, and maple. (Braun 1950, 1955; Gordon 1966; Hor-
ton 1961; Williams 1936, 1949). Numerous flowers,
shrubs, and vines covered much of the forest floor save on
the open oak lands. Along the runs and rivers, reeds and
rushes grew thickly.
Within this wilderness lived a variety of animals includ-
ing bears, wolves, and cougars. The largest mammal in
these forests was the wapiti or American elk, but it is
likely that elk would have been found only scattered
widely within the valley during the warmer seasons (see
Murie 1951). The most common big game in the site area
would have been the whitetail deer, while cottontail rab-
bit, squirrel, and raccoon would have been far more nu-
merous (ODNR 1975).
The Excavations
Initially, the discovery of the prehistoric occupation at
the Hale Farm was fortuitous. The location and nature of
the excavation units had naturally been planned to re-
cover information about the nineteenth, rather than the
ninth century a.d. Although prehistoric materials were
recovered at the base of even the first excavation unit, the
investigation of this early occupation was always limited
by the historical focus of the archaeological investigations
undertaken for the Western Reserve Historical Society. It
was further constrained in part by the nature of distur-
bance due to the subsequent occupations of the area by
Miller and Hale. Thus prehistoric structural data recov-
erable were also limited. The upper surfaces of the four
aboriginal features encountered along the north side of
Hale Run had all been truncated by nineteenth-century
activities. Few continuous stratigraphic profiles connect
the bases of these shallow storage pits or cooking hearths.
Most of the aboriginal materials were recovered from the
fill of storage pits, foundations, and cellars built between
1810 and 1826.
We also discovered that during the continuous histori-
cal occupation of that area from 1826 to as late as 1957,
there had been occasional excavations for new fences, for
cisterns, for driveways, and to rebuttress the slumping
banks for Hale Run. Not only had these more recent ac-
tivities disturbed the early historic and prehistoric levels,
they had brought to the surface some of the buried mate-
rials which were thus reincorporated into the developing
topsoil. It was clear enough that a chipped stone projectile
point, an 1835 half-dime, and an aluminum wire-drawn
nail did not belong to a single occupation. While those
artifacts could be assigned to an appropriate period, that
was not possible for fragments of charred bone, or for the
empty small postholes which we encountered in disturbed
areas of the site. Since we cannot know to which period
such items pertain, they have been excluded from the fol-
lowing discussions of the prehistoric component of the
site.
All archaeological excavations at the Hale Farm site
were based on a 5 ft square grid system (1.52m) oriented
north-south, with the front of the 1 826/ 27 Hale House as
its western boundary. The actual units chosen for excava-
tion were determined by a combination of three types of
information. First, various historical sources suggested
that the original Miller cabin was located just north of
Hale Run and due west of Oak Hill Road. Therefore a
series of fifteen excavation units in that area were ran-
domly chosen for testing. Secondly, there was a vague leg-
end that the 1870 barn which had been located north of
the present driveway, was built over the original occupa-
tion. In that area ten excavation units for testing were
chosen in order to avoid obvious recent disturbance. Fi-
nally, a series of ten test units were chosen to investigate
otherwise unaccountable depressions, knolls, or topo-
graphic irregularities and to “fill in” the coverage of the
area west of Oak Hill Road.
Excavation began by stripping the recent sod from
within each test unit. Below that level the excavations
proceeded with shovel and trowel to remove soils by a 10
cm depth or to the point where we noted any changes in
soil color or texture which might represent a feature such
as the top of a pit or the edge of a wall. At that point a
measured drawing of the floor of the test unit was made,
photographs and samples were taken, and the new feature
was given a field provenience number which was assigned
to all material recovered in association. Any concentra-
tions of artifacts or biotic remains were treated as separ-
ate features. All soils after removal by trowel were
screened through 1/4 in or 3/32 in hardware cloth. The
decision to expose all features at the level they were first
encountered often required excavation of adjacent units
and resulted in a rather strange pattern of excavation un-
its (Fig. 2).
Prehistoric archaeological materials were encountered
in 26 of the units excavated (Fig. 3), as well as along the
northern bank of Hale Run. Yet only three areas of test
units revealed prehistoric archaeological features. To the
immediate north of the driveway, below the twentieth-
century disturbance resulting from the removal of the
1870 barn, a small shallow pit had been dug into the silty
clays. This pit, about 61 cm in diameter and 20 cm deep,
contained a broken chipped stone point and a single turtle
carapace. Possibly the turtle shell had once been used as a
bowl.
In parts of three adjacent excavation units just west of
Oak Hill Road, another aboriginal pit was located. It had
been cut into by one of the posts from the original fence
line erected by Hale prior to 1826. Even more shallow
than the first pit, it was about 76 cm in diameter, while less
than 1 5 cm at its deepest. In it we found a concentration of
Fig. 2. Excavation units at the Hale Farm site.
Fig. 3. Areas of prehistoric materials at the Hale Farm site.
40
DAVID S. BROSE
No. 41
26 prehistoric ceramic fragments from two different
vessels.
The greatest concentration of prehistoric remains was
encountered along the edge of Hale Run between Oak
Hill Road and the 1826/27 house. We had excavated two
sets of contiguous trenches which crossed at the location
of the 1810-27 root cellar below the original cabin. We
found a refilled aboriginal storage pit, an aboriginal
hearth, eight postholes, and portions of a sheet midden (a
scatter of artifacts and charred remains) upon the ground
surface where Hale Run had flowed some ten centuries
ago. The scale drawings of the north-south trench side-
wall stratigraphy, reproduced as Fig. 4, show clearly that
these postholes and the associated midden underlay, and
thus predate, the early nineteenth-century occupation.
However, that nineteenth-century activity had removed
large portions of the original site. We were able to expose
and record the earlier prehistoric levels only along the
floor of our east-west trench (Fig. 5).
The western set of five postholes and one pit may repre-
sent a small, more or less round structure about 2.4 to
3.5 m in diameter. If the eastern set of three postholes
were the remains of another similar structure, then most
of it appears to have been destroyed by Miller when he
built his cabin. In the middle of the 1971 digging season,
there did not appear to be any significance to the distribu-
tion of different types of prehistoric material within this
area of the site. And yet, we were primarily concerned
with the historic remains, and even at that time we had
found as much or more prehistoric material redeposited
in historic features. It is nonetheless clear that this area,
just north of Hale Run, was the core of the prehistoric site
occupation area — just as it was to be the core of Miller’s
occupation and of Hale himself until 1827.
Prehistoric Material Recovered
During the excavations at the Hale Farm a variety of
prehistoric artifacts and waste was recovered. These came
from thin undisturbed aboriginal floor levels, and from
the few pits dug for storage or as fire hearths and later
filled with trash. We also encountered a larger number of
similar prehistoric materials from soils which were dis-
turbed and redeposited by Miller or Hale in the early
nineteenth century. Altogether there were 153 chipped
stone tools, along with over 4,000 waste flakes, many of
which were used; ground and polished stone tools and
ornaments; several drilled stone and clay beads; over 500
fragments of aboriginal pottery; and nearly 1 70 floral and
faunal remains.
The detailed analyses of these materials have taken the
Fig. 4. Stratigraphic profile of trench at the Hale Farm site.
1985
PREHISTORIC HALE FARM OCCUPATION
41
545 E 255
^) POST HOLE
ABORIGINAL MIDDEN
Fig. 5. Floor plan of prehistoric stratum in East West trench at the Hale
Farm site.
better part of four years by staff members and volunteers
of the Cleveland Museum of Natural History, and while
these details may seem rather uninteresting and overspe-
cialized to the general reader, such studies represent the
basic data from which archaeologists have been able to
reconstruct much of the economy, technology, and social
geography of these early inhabitants of the Hale Farm
site. And through the stylistic comparison of artifact de-
sign and decoration with similar materials from sites
across the Great Lakes/Ohio Valley region, it is possible
to assign rather closely a date to this occupation. This is in
spite of the fact that the nature of the nineteenth-century
activities seem to preclude recovering any amount of
charred aboriginal organic material which could be sub-
mitted for a direct radiocarbon determination of the date.
Floral am! Faunal Remains
Altogether, over 2,700 fragments of broken, butchered,
or burned animal bone, and 56 pieces of charred plant
material were recovered from the 1971 excavation. Most
of these were either found in historical or in disturbed
contexts. Thus they can have no significance for the un-
derstanding of the prehistoric site occupation. From what
appeared to be stratigraphically intact prehistoric fea-
tures such as postholes or refilled storage/cooking pits,
we recovered fragments of only 160 animal bones and
only 9 charred plant remains.
The animal bone is described in Table I. Although
most of the available meat comes from the two elk, it is
clear that a wide range of animals was being used for food.
The elk, represented by skull fragments and limb bones
primarily, may have been killed and butchered away from
this location. The young bear must have been taken be-
tween spring and late fall. The deer were represented by
the jaw of a yearling and by parts of the skull and several
ribs and proximal limbs of an adult. The antlers of the
adult deer and one elk had been shed, suggesting that they
were taken between late winter and summer. Tooth erup-
tion and wear on the yearling deer indicate an age of
about 15 fnonths (Severinghaus 1949). Most fawns are
dropped between May and late June (ODNR 1975). The
turtle and the fish remains found, along with the pas-
senger pigeon bones, are frequently indications of early
spring or late summer hunting. All in all, analysis of the
animal bones (viz Cleland 1966) reveals that most of the
prey were probably hunted between July and September
(see Cleland 1966, 1976; Ziegler 1973).
This interpretation is complemented by analysis of the
small number of plant remains. These consisted of a single
charred half of a hickory nut shell, three wild grape seeds,
three charred squash or pumpkin seeds, and two frag-
ments of what might have been a single cut cob of ten- or
twelve-row maize. While all of these plant products could
have been harvested between late August and early No-
vember, the hickory nuts could have easily been stored for
a long time. The maize cob could have been discarded just
after harvest or could also have been dried. However,
based upon the historic accounts (Ford 1977); one might
expect only the kernels to have been stored. The grapes,
normally ripe in late September, could not have been kept
long given the damp winters of the valley.
The plant and animal remains together suggest that the
site was occupied in the late summer. During that season
hunting parties brought back two elk, taken from some
other part of the valley. All of the plants and most of the
animals remains recovered should have been available
within the varied environment represented by the Cuya-
hoga River itself, its western valley slopes, and the tribu-
tary floodplains of Furnace Run and Yellow Creek, bi-
sected by Hale Run along which the site was built.
42
DAVID S. BROSE
No. 41
TABLE 1
Animal Bone From Prehistoric Context At The Hale Farm Site
Species
Number of
bones present
Minimum Number
of Individuals
Usable Meat
for Individuals
Total Pounds Usable
Meat Per Species
Percentage of
Usable Meat at Site
American Black Bear
immature
4
1
75 lb
75
10.5%
Whitetail Deer
adult
40
2
80
160
22.0%
immature
13
I
40
40
5.5%
American Elk
adult
21
2
200
400
56.0%
Racoon
13
3
1.5
4.5
0.5%
Beaver
2
2
7
14
2.0%
Passenger Pigeon
5
2
0.5
1
—
Turkey
6
2
10
20
3.0%
Turtle
17
1
0.5
0.5
—
Suckers
39
7
1
7
1.0%
Total
160
23
—
712.
100.5%
Stone Artifacts and Debitage
There were several different sources of stone from
which Hale Farm tools were made. Most of the unfin-
ished chips and cores are debitage representing the knap-
ping of the Bois Blanc, Columbus /Delaware, or Onon-
daga cherts which occur as nodules of chert in the Silurian
and Devonian limestone and dolomite strata forming the
Niagara escarpment. The outcrops nearest to the Hale
Farm would be those in Monroe County, Michigan, in
Essex County, Ontario, or in Erie County, New York. Yet
cobbles and pebbles of these types of chert are quite
common throughout northeast Ohio as stream gravels
washed from the glacial deposits pushed south during the
late Pleistocene.
A number of finished tools and a surprising amount of
the unused debitage recovered at the Hale Farm site came
from working tabular Plum Run, Pipe Creek, and Upper
Mercer flint which occurs in bands and as lenses inter-
bedded with the Mississippian strata of shale and coal in
eastern Ohio. A number of these outcrops occur in Stark,
Erie, and Coshocton counties, Ohio, at distances from 30
to 75 km from Hale Farm. Many of these outcrops show
signs of aboriginal quarrying and several have yielded the
waste from the initial stages of knapping (Stout and
Schoenlaub 1945).
The lithic categories employed in the analysis of the 153
tools recovered at Hale Farm, follow those outlined by
Brose ( 1967b, 1978c), with the aim to define functional arti-
facts on the basis of selective micro-wear studies and the
reconstruction of manufacturing sequences. Table 2 illus-
trates the frequencies of formal chipped stone tools recov-
ered from the Hale Farm site, and indicates whether tabu-
lar flint or nodular cherts were the source. A similar
frequency illustration of the debitage, or technological
byproducts of the knapping activities are presented on
Table 3. By combining both sets of information, it has
been possible to consider the alternative technological se-
quence which shows major differences from later sites of
the general period. Alternative sequences of lithic manu-
facture have been recognized at functionally differing site
types within this same temporal phase (Brose 1976a,
1978a, 1978c). These sequences involve alternative
sources of raw materials, differing initial core preparation
TABLE 2
Absolute Frequencies of Formal Lithic Tools at Hale Farm
Functional Tool
Category
Irreg.
Blades
Bifacial
Scrapers
Drills
Triangular
Projectile
Points
Triangular
Knives
Flake
Gravers
Bifacial
Knives
Wedges
Gouges
Unifacial
Scrapers
Subtotal
Total
Lithic Source
TF
TF NC
TF NC
TF NC
TF NC
TF NC
TF NC
NC
NC
TF NC
TF NC
7
2 5
1 2
24 43
5 14
0 0
7 19
2
6
3 7
53 100
153
Key
TF: Tabular Flint
NC: Nodular Cherts
1985
PREHISTORIC HALE FARM OCCUPATION
43
TABLE 3
Lithic Debitage From All Units at Hale Farm Site
Reduction Stage
Tabular Flint
Nodular Cherts
Total
N
%
N
%
N
%
Pebble cores
used
2
0.05
0
0
2
0.05
unused
9
0.22
0
0
9
0.22
Block cores
used
0
0
3
0.07
3
0.07
Decortication
unused
0
0
2
0.05
2
0.05
Oakes
used
1
0.02
99
2.44
100
2.46
unused
282
6.95
400
9.85
682
16.80
Primary shatter
used
6
0.15
14
0.35
20
0.50
Secondary flakes
unused
283
6.97
286
7.05
569
14.02
with platform
used
13
0.32
25
0.62
38
0.94
w/platform
unused
576
14.19
875
21.56
1451
35.75
without platform
used
1 1
0.27
30
0.74
41
1.01
unused
545
13.43
596
14.68
1141
28.1 1
TOTALS
1728
42.6
2330
57.4
4059
100%
SUBTOTAL
used
33
0.81
171
4.22
204
5.03
unused
1695
41.76
2159
53.19
3854
94.95
and utilization, differing application of heat treatment,
differing reconstructed preform, blank, and finished arti-
fact sequences with different functions for morphologi-
cally similar artifacts.
Some attempt was made to identify, in quasi-ethno-
graphic terms, the inferred function of these “finished ar-
tifacts” and the quantities of utilized debitage. The degree
of detailed investigation needed for much statistical con-
fidence in such identification is unfortunately not present
in the sample used in this study. Nonetheless it has been
possible to reconstruct a model of the lithic reduction
strategy used at the Hale Farm (Fig. 6). Hard hammer
and soft hammer percussion were the major manufactur-
ing technique at all stages, although it was more com-
monly employed in the earlier portions. Pressure retouch
occurred as the final technique for many of the final tool
types, although it occurred earlier for specific artifacts
and was altogether absent for some. Thermal alteration,
never common in any Whittlesey tradition lithic assem-
blage (viz Brose 1973b, 1980, 1985a, 1985b), occurred at
several different times in the reduction sequence, and
again for some tools seemed to be absent entirely. The
products of this lithic strategy, whether unfinished pre-
forms, flake debitage, or finished formal artifacts, dis-
played variations in their indication of utilization such
that less than 20% of the functional stone tools at the Hale
Farm consist of utilized debitage and utilized interme-
diate reduction sequence forms (Brose 1973b), while in
smAll
FLAKES
LARGE
FLAKES
TABULAR
(QUARRIED)
FLINT
TABULAR
CORE
I
NODULAR
(GLACIAL)
CHERT
I
PEBBLE
CORE
DECORTICATION
FLAKES
SMALL
BLANK
IRREGULAR
BLADES
LARGE
BLANK
SMALL
FLAKES
LARGE DECOR-
TICATION FLAKES
SMALL
FLAKES
Fig. 6. Reduction sequence as reconstructed.
44
DAVID S. BROSE
No. 41
other coeval sites these categories may represent over 75%
of the functional lithic assemblage (e.g., Brose and Scarry
1 976). Many of the intermediate reduction forms also ap-
pear to have been used to duplicate the functions of some
of the heavy-duty formal tools (e.g., triangular preforms
were utilized as knives, large decortication flakes were
utilized as gravers). There is considerable redundancy
within this system: however inefficient such a system may
appear, the possible alternative sequences made it rather
stable and very flexible.
Known archaeological sites of this early Late Wood-
land period show little demographic differences between
what are seen as summer multifamily horticultural vil-
lages (along alluvial floodplains in mixed mesophytic
forests) and spring or fall fishing-fowling sites (along the
lakeshore at river mouths) (Brose 1976b, 1978c). Mid-
winter occupation sites are known in upland rock shelters
or at spring locations. These appear to represent smaller
family segments engaged in a rather focused hunting econ-
omy. Summer villages such as the Hale Farm site reveal a
mixed economy. The site shows a utilization of lithic
sources which approximates their relative local availabil-
ity. At winter hunting camps, however, glacially derived
local chert nodules were selected over quarried sources.
These differences in the lithic tool assemblages and in the
reduction strategies between summer villages and winter
campsites seem relatively minor, although in the winter
campsites most cores from tabular quarried flint received
thermal alteration and appear to have been used predom-
inantly for irregular blade production. The limited ther-
mal alteration of nodular glacial cherts appears late in the
reduction sequence at many winter campsites and, where
present, seem restricted to immediate pre-pressure re-
touch of triangular preforms into projectile points and
hafted knives. At the Hale Farm site thermal alteration
appears to have been applied to about 60% of all quarried
flint prior to any knapping and to about 35% of all nodu-
lar chert following its reduction to a large rough blank.
The relative frequencies of functional tool categories
are significantly different from those at winter sites. At
this summer occupation there are several varieties of
small speical purpose tools, numerous projectile points,
and a number of small scrapers. However, at winter occu-
pations (viz Brose and Scarry 1976; Lee, Brose, and
Weisman 1983; McKenzie et al. 1974) the tool assemblage
was characterized by fewer different tool types. There was
also a lower frequency of projectile points, and a large
number of knives and heavy scrapers. In general, the lithic
assemblage at summer villages reflects greater evidence
for what Binford (in Whallon 1 974) has termed curatorial
activity, and for a much wider range of extractive activi-
ties than do winter campsites. Summer sites also reflect an
industrial sequence with more carefully chosen raw mate-
rials, more evidence for thermal alteration and at earlier
and less artifact-specific production stages, and a higher
frequency of nonspecific utilized debitage than in winter
campsites (Brose 1978c).
TABLE 4
Metric Attributes (mm). Unifacial Scrapers,
All Units Hale Farm Site ( 33Su 1 7)
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Shape
H73BD
30.65
16.55
5.35
Lunate
9576
HI 16S
29.55
22.85
12.85
Thumb-nail
9618
H170HG
19.45
14.25
5.10
Triangular
H78AC
21.45
14.65
4.30
Triangular
9590
H136AK
19.0
14.45
5.40
Rhombridal
9506
H20EE
21.85
10.95
4.95
Ovate
9508
H23DK
13.25
8.30
3.35
Thumb-nail
9561
H 101 AP
16.2
14.65
2.65
Rhombridal
9586
H 1 3 1 A
22.75
13.65
5.75
Ovate
9584
H135KY
20.0
15.65
8.85
Convergent
Beyond this there is little in the assemblage of chipped
stone tools from Hale Farm of great interest (PI. II). The
unifacial scrapers (Table 4), bifacial scrapers (Table 5),
triangular preforms and knives (Tables 6 and 7), or drills
(Table 8) all seem similar to those from contemporary
sites such as the Bass Lake and Columbia Road sites, and
to those from components representing the early phases
of the Whittlesey tradition after a.d. 1200 (Brose 1973b,
1976b, 1978a, 1980). Assuming that drills were for drilling
and scrapers were for scraping, and recognizing that the
differing frequencies of different tools probably reflect
TABLE 5
Metric Attributes (mm)
Bifacial Scrapers from Hale Farm Site
(x = Broken)
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Shape
9523
H44C
25.45
24.35
9.95
Triangular
*9535
H65B
18.60
19.65
3.95
Sub-triangular
9549
H86AV
35.60
30.0
6.85
Sub-triangular
9536
H67AQ
28.0
20.25
10.35
Ovate
9500
H12S
26.20
23.10
10.0
Ovate
*9605
H134BB
26.20
12.85
4.0
Lunate
9505
H 1 38 Y
20.55x
15.85x
4.45
Triangular
PI. II. Chipped stone artifacts from the Late Woodland component at the Hale Farm site.
46
DAVID S. BROSE
No. 4!
TABLE 6
Attributes (mm) of Triangular Preforms/ Bifacial Knives
(x = Broken)
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Striae,
Grinding,
Beveling or
Edge Damage
Present?
9536
H67AH
33.65
29.35
9.85
Yes
9536
H67AQ
28. Ox
20.45x
11.0
None
9618
H170HG
20.35x
I4.35x
5.15
None
9618
H170AG
23.20x
30.25x
11.40
None
9530
H58U
28.90
21.25
7.85
None
9538
H69AW
44.75
22.15
7.45
None
9618
H170AC
21.95
23.85
10.10
None
9618
HI70AG
35.35
22.45
11.55
None
*9555
H94C
35.95
22.75
7.10
Yes
9656
H214KT
29.0
20.55
8.55
None
9656
H214KR
32.0
29.9
7.75
None
9626
H180D1
23.25
14.55
8.10
None
9578
H120N
28.85x
20.0
9.0
None
9519
H33S
31.55
15.85
6.75
None
9519
H33S
31.55
15.85
6.75
None
9519
H38CK
21.45
22.50
8.65
Yes
9596
H I44DI
l!.35x
17.15
7.65
None
9508
H23BV
41.35
30.55
9.35
None
9584
H27A
38.1
15.4
7.85
None
9617
H16BJ
17. Ox
1 1.45x
4.85
None
9574
HI 10AC
13. Ox
9.10
3.60
None
9644
lO.lOx
17.85
5.55
None
9618
H170AF
17.25
25.45x
7.0
None
9623
HI77GR
22.0x
14.45
6.45
None
9547
H82E1
59.10
34.55
12.55
None
9541
H172AF
23.35x
23.20
9.45
None
TABL E 7
Attributes (mm) of Triangular Knives
(x = Broken)
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Striae,
Grinding,
Beveling or
Edge Damage
Present
9560
H100B
26.75
20.6
4.45
Yes
9561
H 101 AD
25.85
18.85
3.25
Yes
9560
H100A
27.05
16.25x
3.65
None
9560
H100G
30.95
21.65
5.55
Yes
9560
H100D
28.0
19.65
6.75
None
9560
H100E
43.25
28.75
10.0
Yes
9679
H175A
18. Ox
14.0
4.65
None
9545
H78AB
14.65x
15.35
5.90
None
9642
H204CP
20.55x
18.85
7.65
None
9618
H170AD
25.65
23.25
7.55
None
9614
H 166 AH
18.55x
15.35
4.25
None
9647
H204EQ
13.75x
14.85
4.45
Yes
9622
H176HI
13.75x
16.20
5.75
Yes
9530
18.95
13.65x
3.75
None
9565
HI 05 A
35.25x
21.65x
4.55
Yes
9569
H 1091
26.55x
21.45
7.35
Yes
9564
H104H
22. 8x
I3.55x
4.0
None
9563
H103FY
14.9x
14.7
3.35
None
9566
H106DI
12.25x
13.4
3.25
Yes
differing site-to-site task emphases, there does not seem to
be much major difference in the tools themselves during
the early Late Woodland in this area except for the pro-
jectile points (Table 9).
Traditionally, in late prehistoric sites, the most diag-
nostic chipped stone artifacts have been the triangular
projectile points, and variations in width and base shape
have been used to order assemblages in time. The single
broadest triangular projectile point does have a concave
base, and the single narrowest triangular projectile point
has a straight base. However, the statistical analysis of
1985
PREHISTORIC HALE FARM OCCUPATION
47
TABLE 8
Metric Attributes (mm) of Drills From Hale Farm
Catalogue
Location
Point
Type
Max.
Length
Max.
Width
Max.
Thickness
Shoulder
To Base
Base
Width
Min.
Haft
Width
Haft
To
Base
9499
HI 105
46.5
19.7
8.0
16.1
19.7
6.6
30.4
9532
H61AL
18.55
8.25
3.20
—
—
—
—
9532
H61AM
32.15
10.85
10.10
TABLE 9
Attributes (mm) of Triangular
(x = Broken)
Projectile Points
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Base Type
Catalogue
Maximum
Maximum .
Maximum
9530
Location
Length
Width
Thickness
Base Type
H-58
19. lOx
14.0
3.65
Concave
9644
9644
H201
lO.Ox
17.65
5.50
Straight
H201GA
14.95x
10.25
3.20
Concave
9679
9643
H 175
17.85
14.15
4.74
Concave
H200HE
12.65x
1 1.35
3.15
Straight
9622
*9618
H176HI
13.1 5x
16.25
5.75
Straight
HI70C1
30.1
19.9
6.6
Straight
9570
*9612
H-136AR
14. 1 5x
19.25
5.35
Concave
HI63AE
26.2
17.7
4.2
Straight
9545
9609
H78AB
14.55x
15.55
6.0
Straight
H160BC
21.6
14.4
4.6
Straight
9545
9612
H78AD
30.65
11.45x
5.65
Straight
H163AF
1 l.Ox
9.2x
2.1 x
X
9536
9617
H67AH
33.45
25.40
9.85
Concave
H 169
19.8
15.7
2.8
Straight
9614
9629
H 166 AH
18.65
15.1 5x
3.95
Straight
H183FU
24.6
15.7
4.4
Straight
9618
9623
H170AF
1 7. 1 5x
25.35x
6.95
Straight
H177GD
12.4x
17.7
4.1
Concave
9574
9629
HI 10AC
12.70
9.20x
3.45
Concave
H183MN
19.1
1 !.5x
3.4
Straight
9647
*9635
21.9
19.8
3.7
Straight
H204CP
25.35x
14. Ox
5.10
Straight
9629
9618
H187AA
18. 3x
17.9
6.3
Concave
H170AC
16.0x
26. lOx
9.65
Straight
9595
9623
H143D
27.3
18.1
5.6
Straight
H177GR
1 3.55x
22.35x
7.65
Concave (?)
9586
9647
H131D
21.75
23.75
5.5
Concave
H204EO
13.75x
13.65x
4.35
Concave
9589
9651
H135LU
16.25x
13.5
4.75
Straight
H208GR
16.85x
15.90
5.95
Straight
9647
9656
H204CR
25.20
12.40x
5.55
Straight
H214KP
17.95
12.85x
3.65
Straight
9545
*9656
H78AD
29.25
1 I.OOx
5.45
Straight
H214K.U
16.75
12.60
3.85
Straight
9548
9552
H840
24.10
15.10
4.70
Straight
H90D
20.20x
16.20
3.40
Concave
9597
9537
H145DG
24.45
13.70
4.55
Straight
H65?
19.15
10.35
4.45
Straight
48
DAVID S. BROSE
No. 41
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Base Type
9509
H24A
12.35.x
17.0
2.75
Concave
H20ED
9.80x
22.70
4.35
Concave
9644
H20IGP
1 5.65x
I5.35x
4.60
Concave
9644
H20IGA
14.95x
10.25
3.20
Straight
9643
H200HE
1 2.65x
11.35
3.15
Concave
*9618
H170C1
30.1
19.9
6.6
Straight
*9612
H163AE
26.2
17.7
4.2
Straight
*9619
H172AE
36.2
29.8
9.1
Straight
9609
H160BC
21.6
14.4
4.6
Straight
9612
H163AF
1 l.Ox
9.2x
2.7x
X
9617
H 169
19.8
15.7
2.8
Concave
9545
H78AC
18.85x
I4.45x
3.95
Straight
*9592
H145DG
23.75
13.80
4.40
Straight
*953 1
H57A
25.75
22.35
3.65
Concave
*9535
H65A
26.95
20.75
5.50
Straight
9530
H158EX
(9608)
18.10
14.9
5.55
Concave
9512
H30FF
20.55x
12.55
4.35
Straight
9511
H28 1
18.55
14.50
4.75
Concave
9519
H38CZ
17.0
14.75
3.45
Straight
9513
H31AR
25. Ox
21.45
6.75
Concave
9574
HI ION
30.25x
30.10
5.10
Straight
9574
HI I0P
22.20x
22.0
4.80
Straight
*9574
HI I0Q
30.85
28.80
4.55
Straight
*9574
H110R
28.10
23.65
4.90
Straight
9577
HI 19Z
19.25x
1 5.85x
5.0
X
9656
H214K.Q
15.25
9.85
5.10
Straight
9656
H214KV
27.25
6.65
4.45
Straight
9656
H214KS
22.35
15.0
6.20
Straight
length/width ratios show that the complete straight base
points (n = 23; x = 1.53; s.d. = 59) and the complete con-
cave base points (n = 7; x = 1.20; s.d. = 1.08) both lie within
the normal length/ width distribution for all 30 unbroken
points (x = 1.45; s.d. = 0.63). That is, it is statistically
proper to consider these all as representing the normal
variation within a single population. Although virtually
identical projectile points were first formally described as
Madison points by Scully (1951), the most detailed typo-
logical analysis of such projectile points in the Great
Lakes area is that by W. Ritchie ( 1 97 1:13—14, PI. XVI; see
also Converse 1964; Luedtke 1978). Ritchie had described
the Madison point in New York as having replaced the
somewhat larger and broader Levanna point with their
more concave bases. Levanna points were common from
a.d. 900 to a.d. 1300 (Ritchie 1971:31). Morphologically
all of these Late Woodland projectile points from the
Hale Farm site are transitional but are best seen as an
early varient of the New York Madison points rather than
as a late variant of Levanna type (cf. Fitting et al.
n.d.TO). This suggests the Hale Farm points (PI. Ill)
should date to between a.d. 1000 and a.d. 1200.
As at most late archaeological sites in northeast Ohio,
not all of the chipped stone knives/points were as recent
as the site itself (Brose 1973b, 1975, 1976a, 1980). A
number of notched Archaic projectile points (some
broken) were recovered from Hale Farm (Table 10). Most
of these are not very diagnostic in terms of the specific
temporal periods they represent, but the resemblance to
the types Otter Creek (PI. Ill, bottom row, second) to
Normanskill (PI. Ill, bottom row, third) or Lamoka (PL
III, bottom row, fourth) suggest they may all have been
made between 4000 and 2500 b.c. (Ritchie 1971:29-30,
37-38, 40-4 1 , Pis. XIII, XIV, XVIII, XX, and XXI). Cer-
tainly there are numerous Middle/ Late Archaic sites in
this portion of the Cuyahoga Valley which were occupied
at that time (Brose 1975; Brose etal. 1981). A final class of
chipped stone tools recovered from the Hale Farm are
represented by nine whole and fragmentary split shale or
slate knives (Table 11). All display crudely chipped
curved sides with battered and often scratched edges, pre-
sumably as the result of heavy cutting use of some sort
(Keeley 1974, 1980). Similar split shale knives were noted
at the Boston Ledges shelters (Brose and Scarry 1976;
Read 1880) which yielded ceramics similar to the Hale
Farm site.
In addition to the chipped stone materials, the Hale
Farm site excavations yielded four drilled limestone and
fired clay beads; three fragmentary drilled gorgets, two of
slate and one of sandstone; and a single celt (axe or adze)
of granodiorite (Table 1 2). Not one of these artifacts is par-
ticularly diagnostic; all have a wide distribution in time
and space within the aboriginal cultures of the eastern
Woodlands. Other than the celt these artifacts were found
in somewhat disturbed contexts. The celt was found em-
PI. III. Chipped stone projectile points from the Hale Farm site.
50
DAVID S. BROSE
No. 41
TABLE 10
Metric Attributes (mm) of Archaic Points, All Units
Point Attribute Chart
Min.
Depth of
Catalogue
Maximum
Maximum
Maximum
Shoulder
Base
Haft
Haft to
Bifurcation
Location
Length
Width
Thickness
to Base
Width
Width
Base
or Indent
Grinding or Beveling Present?
9521
H42CW
9547
51.35x
41.95
6.85
7.20
20.55
7.55
6.2
No
H82EK
9547
20.0
18.60
8.45
10.5
3.35
3.35
0
No
H82EN
17.40x
23.45
8.55
6.10
20.90
13.60
8.0
Shoulder margins and haft edges
ground
9500
H21H
13.85x
15.95
3.45
9.55
9.55
3.10
Base and edges ground
9640
H195A
9558
48.25
14.20
9.00
5.5
14.2
9.85
3.85
No
H97A
39.90x
38.10
8.70
11.95
30.10
21.95
14.90
2.95
Side notches ground
9578
H120M
34.65x
23.85
7.10
5.20
17.10
14.0
5.65
Side notches ground
bedded within the thin aboriginal midden level encoun-
tered just northwest of the cabin foundations along Hale
Run (see Fig. 5).
Aboriginal Ceramics
Previous Whittlesey ceramic typology has been both
confusingly inclusive, inconsistent, and imprecise. This
TABLE 11
Split Slate Knives
(in mm)
Striae,
Grinding,
Beveling or
Catalogue
Location
Maximum
Length
Maximum
Width
Maximum
Thickness
Edge Damage
Present?
9545
H78A
45.0
16.25
9.25
Yes
9578
H120P
21.65
23.0
6.85
Yes
9578
H120T
31.0
22.75
9.55
None
9622
H176HD
63.7
32.3
10.6
Yes
9615
H167T
40.75
23.20
6.25
None
9525
H98J
47.65
31.0
9.25
None
9511
H26AV
49.55
47.90
9.80
Yes
9509
H214KO
23.40
16.45
5.95
Yes
9615
HI67G
45.9
23.2
8.8
Yes
point had earlier been made by Murphy (1971b) who pro-
vided the first workable descriptions for many of the
Whittlesey ceramic types. I have expanded upon and
modified several of the type names proposed by Morgan
and Ellis (1943), Fitting (1964) and Murphy (1971a,
1971b), in an effort to look for significant changes in ce-
ramics through time and space. While not yet demonstra-
ble that the type varieties proposed will reveal everything
we always wanted to ask about the late prehistory of
northeast Ohio, it is unfortunately true that previous ce-
ramic seriations based upon the previous ceramic typol-
ogy have not revealed much at all. For example, the mis-
named type Fairport Plain was first loosely defined by
Fitting ( 1 964: 1 64) who added the statement that most of
these ceramics were cordmarked to the lip, although some
were plain. Fitting also noted that both plain and notched
lips occurred, and, although he reported that there were
variations in the temper of Whittlesey ceramics, he did
not use temper in his type definitions (1964:162).
Based on the ceramics recovered in his 1968 excavation
at Fairport Harbor, Murphy (1971a) suggested that there
did not appear to be any significance to the distributions
of the ca 30% shell-tempered (and mixed shell-and-grit)
ceramics at Fairport Harbor in terms of surface finish, or
lip mode or decoration, a position supported in his more
analytical definition of the type Fairport “Plain” (Murphy
1971 b:299— 300). The term Fairport Plain, with or with-
out quotation marks, is still incorrectly inclusive. It is my
position here (see Brose 1980) that the distinction between
plain or smoothed vessel surfaces and cordmarked vessel
surfaces, is likely to carry some degree of archaeological
information although whether chronological or socio-
geographical is yet unknown. Indeed the use of overly in-
clusive and/or borrowed ceramic terms (e.g., Brose
1985
PREHISTORIC HALE FARM OCCUPATION
51
TABLE 12
Miscellaneous Ground and Polished Stone Artifacts
Catalogue
Location
Description
Max. L.
(mm)
Max. W.
(mm)
Max. T.
(mm)
Inside Diameter of
Drilled Holes
(in mm)
9612
HI 63 AC/
2 portions of 3 hole
37.1
29.6
4.4
3.9 (incomplete)
H163AD
Slate Gorget
29.1
31.9
4.4
4.6and 5.4(broken)
9656
H214KN
Broken 1 hole Gorget of Sandstone
97.65
47.0
12.45
6.2
9647
H-204AA
Broken Slate Gorget
51.2
40.4
9.10
_
9595
H-143A/
Broken clay bead
15.85
7.55
6.45
3.5
H143C
Broken clay bead
16.55
7.55
6.65
3.5
9630
H-184CE
Complete clay bead
22.9
9.0
8.0
4.1
9630
H-184X
Broken limestone bead
12.65
7.9
7.0
3.8
9645
H-202A
Granodiorite Celt
110.65
72.95
32.5
—
1973b) is likely to conceal such potential data. The type
name Fairport Harbor Plain should be reserved for the
plain vessels previously included in the old Fairport Plain
of Fitting and the Fairport “Plain” of Murphy. This
leaves the type name Fairport Harbor Cordmarked to
refer to the cordmarked ceramics previously included by
those authors. I have further used the type-variety system
as propounded by Phillips (1970) to determine possibly
significant varieties within each type, distinguished by the
presence (var. Painesville) or absence (var. Willoughby)
of the horizontal row of punctates below the vessel lip.
The variable lip notching noted by all authors, from
Greenman (1935a, 1935b, 1937) to Murphy (1917b),
appears to be a freely varying vessel mode in the sense
used by Phillips ( 1970( 1):55). As such it deserves descrip-
tion, but until its significant distribution in time and/or
space is demonstrated, need not be given classificatory
status.
A resolution for the issue of the ceramic tempering
material is less obvious in the late prehistoric ceramics of
the Lake Erie region. There seems little justification for
cluttering the literature with a plethora of ceramic type
and variety names redundant in all save the nature of in-
clusive tempering agent when there seems to be a free vari-
ation along a continuum from wholly grit-tempered
through mixed shell-and-grit-tempered, to wholly shell-
tempered ceramics which differ in no other attributes (cf.
Brose 1973b; Brose et al. 1976). Indeed, shell tempering
seems to appear relatively early in the seriation of Whit-
tlesey ceramics (say after about a.d. 1200), at which time
its relative frequency may be as high as 20% of some types
(Brose 1976b, 1978a, 1980, 1984, 1985a, 1985b). There is
not much apparent change in frequencies through time,
for the latest dated Whittlesey components also yield
about 15% to 20% shell and mixed shell-and-grit temper-
ing. Nonetheless, the high relative frequency of shell tem-
pering for some types at the Fairport Harbor site itself
(Murphy 1971b) or at some structures at the South Park
site (Brose 1978b, 1985b) suggests that with closely con-
trolled ceramic lots the frequency of shell tempering may
represent significant social patterning, if not also being
temporally significant (but probably not temporally
diagnostic).
Using the former Ohio State Museum collections of
Greenman (1935a, 1935b, 1937) and Morgan ( 1943), Fit-
ting ( 1 964: 1 65ff) developed a seriation from early to late:
Fairport Plain to Reeve Opposed-Reeve Horizontal to
Tuttle Hill Notched. He concluded that the site sequence
was Fairport Harbor, Reeve, South Park, and Tuttle Hill.
This sequence, and indeed much of the ceramic seriation,
was based on the assumption of single componency for
the sites.
Murphy clearly recognized the fact that multiple occu-
pations had occurred at South Park (1972), at Fairport
Harbor (1971a), and at Lyman (1971c). Recent investiga-
tions (Belovich and Brose 1982; Brose 1973b, 1975, 1976a,
1976b, 1985b; Brose, Wentzel, et al. 1976; Brose, White,
and Ford 1983; Bush 1982; Lee 1982) have confirmed
Murphy’s recognition and added data to support the
multi-component nature of the Reeve and Tuttle Hill sites
also. Lacking the stratigraphic control revealed by subse-
quent extensive excavation, Murphy (1972) seriated his
South Park ceramics into four typological components.
From early to late these were represented by a component
characterized by Mixter series and what Murphy called
Glen Meyer ceramics (1972:33); a component character-
ized by Reeve Horizontal, his newly defined Reeve Fil-
leted, and crude Parker Festooned ceramics; a compo-
52
DAVID S. BROSE
No. 41
nent characterized by Reeve Opposed and Horizontal
and possibly by his recently defined Fairport Filleted ce-
ramics; and a final component dominated by Tuttle Hill
Notched ceramics. Based upon this study, and upon his
earlier typological remarks (1971b), I infer that Murphy
felt the Whittlesey ceramic seriation might be dichoto-
mized into a Cuyahoga Valley and a somewhat different
Fake Shore sequence (but with considerable overlap at
times).
Murphy’s suggested overall sequence, which seems to
have given little significance to the relative frequencies of
shell temper (rightly, I believe), began prior to a.d. 1300
with Mixter and Glen Meyer as external ceramic intro-
ductions into a local assemblage of Fairport “Plain” and
Reeve Horizontal. This was seen as followed by the de-
velopment of Fairport Filleted and Reeve Opposed ce-
ramic types, possible as a response to the introduction
after a.d. 1400 of Parker Festooned from the west and
McFate Incised from the east. Reeve Filleted had devel-
oped from Fairport Filleted after a.d. 1400, and by a.d.
1600 this had in turn developed into Tuttle Hill Notched,
the latest local ceramic type. Murphy recognized that
there were morphological varieties in each of these Whit-
tlesey ceramic types and suggested that these might have
chronological significance. As he stated, “All such defini-
tions of formal ceramic types are inevitably subjective and
are always subject to future refinement. The same is true
of inferred relationships and suggested ages of these
types” ( 1 97 1 b;298).
Although the decade since his study has indeed pro-
vided data for typological refinement, has raised some
questions concerning the cultural relationships, and has
seen further excavation at stratified sites yielding revi-
sions to his suggested ages, the general sequence Murphy
proposed in 1971 has been supported.
Murphy’s typology was artificially limited at its earlier
end, due to the circumstances that no early Whittlesey
sites had been identified. The recent excavations at a
number of small early Fate Woodland sites in Cuyahoga,
Lake, Geauga, and Summit counties ( Belovich and Brose
1982; Brose 1975, 1980; Brose and Pratt 1976; Brose and
Scarry 1976; Brose, White, and Ford 1983; Bush 1982;
Fienga and Lee 1982; Lee 1982) have shown that the
earliest Whittlesey ceramics developed from local Wood-
land antecedents in the period between a.d. 1000 and a.d.
1200, during the Riverview phase, as previously defined
by lithic seriation (Brose 1978c).
The Jonathan Hale Farm site lithic materials were a
critical component of that Riverview phase, and the abo-
riginal ceramics form an assemblage which by seriation
should also fall in the period between a.d. 1000 and a.d.
1 150.
The 1971 and 1980 excavations at the Hale Farm recov-
ered just over five hundred fragments of aboriginal pot-
tery (Tables 13 and 14; Pis. IV and V). The ceramic as-
semblage from the Hale Farm represents 1 1 different
TABLE 13
Distribution of Surface Finish on Prehistoric Ceramics
from the Hale Farm Site
(Rimsherds/Body sherds)
Exterior Surface Treatment
Interior Surface Treatment
Cordmarked
Plain
Brushed or
Fabric
Impressed
Total
Cordmarked
1/6
14/446
0/0
15/452
Plain
0/0
3/0
0/0
3/0
Brushed or
Fabric Impressed
0/0
0/52
0/0
0/52
Total
i/6
17/498
0/0
18/504
vessels. Nine of these, grit-tempered cordmarked jars with
no plastic modifications, are typical of the general early
Late Woodland period in the Great Lakes Region. Two of
the vessels have the notched, thickened lip or folded rim
strip or collar which occurs both in some regional early
Late Woodland ceramic assemblages at about a.d. 1000,
and which, in many varieties, seems to have been consid-
ered a characteristic of Whittlesey focus ceramics, usually
assigned to this period a.d. 1300 to a.d. 1600.
The type Cuyahoga Cordmarked, is represented by five
rimsherds, apparently from three different vessels at the
Hale Farm site. It was previously defined based on ceram-
ics recovered from two rock-shelters located in Boston
Ledges along Richie’s Run, about five miles downstream
and across the Cuyahoga Valley (Brose and Scarry
1976: 134-136). All Cuyahoga Cordmarked ceramics rep-
resent small somewhat outcurved rim, semiconoidal to
subglobular jars with massive grit tempering. Flat,
slightly thickened lips show a slight exterior bevel and are
about 9. 6 mm thick and the rims below the lip are from
8.2 to 12. 6 mm (x = 9.6 mm) in thickness. Vessel exterior
TABLE 14
Summary Statistics for Thickness of
Prehistoric Ceramics from the Hale Farm Site
(in mm)
Sample
Vessel Portion
Size
Mean
S.D.
C. V.
A. Lip
18
6.072
1.044
0.172
Neck
17
7.458
1.054
0.141
Shoulder
16
8.425
1.764
0.209
Body
195
10.704
8.995
0.840
B. Body Sherd with
brushed or Fabric
Impressed Exterior
and/or uniform
curvature 45 12.691 2.482 0.196
Other Body Sherds 150 9.167 3.556 0.388
PI. IV. Aboriginal ceramics from the Hale Farm site; exterior surfaces.
PL V. Aboriginal ceramics from the Hale Farm site; interior surfaces.
1985
PREHISTORIC HALE FARM OCCUPATION
55
surfaces are decorated with vertical to slightly oblique
overlapping impressions of a paddle tightly wrapped with
a two ply z, z, S cord. Vessel interiors show direct or
slightly dragged z, z, S cordwrapped paddle edge or stick
in a zone from the lip to a depth of about 40 mm (Brose
and Scarry 1976:134-136).
From the Hale Farm site one Cuyahoga Cordmarked
vessel with no additional decoration clearly corresponds
to that type description and is here assigned the name
Cuyahoga Cordmarked var. Boston Ledges. While iden-
tical in most attributes to the previous type description its
unmodified lip is slightly thinner in thickness (x = 7.8
mm), lying at the lower limit of the type distribution. Two
additional Cuyahoga Cordmarked vessels from the Hale
Farm site are also represented by unmodified lip modes
but display the single horizontal row of circular punctua-
tions about 25 mm below the lip. These vessels of Cuya-
hoga Cordmarked var. Hale also have a somewhat
thinner (x = 6.8 mm), as well as a somewhat more out-
curved, neck area. In these attributes they differ from the
typical Cuyahoga Cordmarked type in the direction of the
Fairport Harbor Cordmarked type and seem to represent
a typological and temporal intermediate ceramic variety.
Fairport Harbor Cordmarked is represented by one
vessel each of var. Painesville, plain lip mode, and var.
Painesville, notched lip mode; by three vessels of var. Wil-
loughby, plain lip mode; and by one var. Willoughby,
notched lip mode vessel. For all six of these Fairport
Harbor Cordmarked vessels, the ceramic paste is con-
torted and no coil breaks are evident. Temper consists of
crushed granitic grit with some apparent selection for
lighter, acidic angular minerals such as quartz and plagio-
clase. Temper particle size ranges from 0.8 mm to 1 .8 mm
with a mean of about 1.2 mm. Temper density is relatively
high, possibly representing between 20 and 30% by vol-
ume. Thickness at the flattened lip ranges from 5. 1 mm to
6.6 mm with a mean thickness of 6.2 mm. Just below the
slightly extruded lip, the neck ranges from 6. 1 mm to 8.3
mm in thickness with a mean of 6.8 mm for these vessels
with their minimally thinned, minimally outturned rim
profiles (Fig. 7). While rim diameters could not be deter-
mined from these small sherds, similar ceramics from the
Fig. 7. Rim profiles of aboriginal ceramics from the Hale Farm site (interior to right), with recon-
struction of small vessel of the type Fairport Harbor Cordmarked, var. Painesville.
56
DAVID S. BROSE
No. 41
earlier Fairport Harbor site collections of Morgan, Ellis,
and Murphy range from just under 24 cm to over 55 cm.
All exterior vessel surfaces have been malleated from
base to lip with a rather coarse, loosely cordwrapped
paddle. Ceramic cord impressions are predominantly ver-
tical in orientation. They appear almost entirely to be rep-
resentative of s, s, Z twisted two-ply cords, although a
variety of different sized elements are present on the dif-
ferent vessels. All vessel interiors are smoothed, and often
shallow horizontal finger impressions of the potter can be
felt.
The three vessels of Fairport Harbor Cordmarked var.
Painesville are decorated by a single horizontal row of
circular or annular punctates around the upper neck.
These punctates range from 3 to 7 mm in diameter, are
from 8 to 1 5 mm apart, and from 14 to 40 mm below the
lip. Occasionally the punctates were deep enough to
create interior bossing and, rarely, they completely pene-
trated the inner wall (although the lack of regularity for
this attribute on single rimsherds suggests that this was a
production mistake rather than a deliberate style). One of
the Fairport Harbor Cordmarked var. Painesville vessels
has a series of exterior shallow lip edge notches made by
the transverse impression of a rounded tool, and one var.
Painesville vessel has a plain lip mode. The Fairport Har-
bor Cordmarked var. Willoughby vessels, without a row
of punctations, are also represented by one vessel with a
plain lip mode and with two vessels having the notched lip
mode.
One grit-tempered vessel of the type Fairport Filletted
(Murphy 1 97 1 b:299) is represented by two rimsherds
from the Hale Farm site. The folded rim (sometimes
called a thickened lip or collar) is weakly developed, ex-
tending downward from the exterior lip for about 8.5 mm
and being only about 3.8 mm thicker than the flattened lip
of 6.5 mm. The thickest portion of the exterior folded rim
is transversely notched with narrow, fingernail-like im-
pressions. The lip itself has a single shallow longitudinal
grove incised along its center. Apart from this unusual lip
mode the vessel appears similar to ceramics Murphy re-
covered for the Cleveland Museum of Natural History in
1968 which he felt (correctly, I believe) were intermediate
between Fairport “Plain” and Fairport Filleted (Murphy
1 97 1 b:299). I have previously proposed two varieties of
the type Fairport Filleted, distinguished by the presence
(var. Hillside) or absence (var. Fairport) of vertical finger
trailing below the rimfold to the shoulder (Brose 1980,
1984). The Fairport Filleted vessel from the Hale Farm
could thus be considered var. Fairport with an incised rim
mode. I am not comfortable with this pigeonhole, but
only the recovery of further examples of such ceramics in
significant context would justify the establishment of a
new variety.
The final aboriginal vessel from the Hale Farm site is
represented by one grit-tempered, smoothed or plain sur-
face rimsherd with a flat notched lip and a single horizon-
tal row of circular punctates. This is similar to many of the
plain ceramics with the notched lip mode which had been
previously subsumed within the type Fairport Plain (Fit-
ting 1964) or Fairport “Plain”(Murphy 1971a, 1971b). To
follow the terminological treatment for the cordmarked
majority of those ceramics, this plain vessel should be
called Fairport Harbor plain, var. Painesville.
These aboriginal ceramics can all be considered as a
single prehistoric cultural assemblage, assignable to be-
tween a.d. 900 and a.d. 1200. They thus represent the ce-
ramic complex transitional from the still poorly under-
stood early Late Woodland period to the later prehistoric
Whittlesey focus, or tradition (Brose 1973b, 1976a,
1976b, 1978a, 1980; Fitting 1964; Greenman 1937; Griffin
1946, 1967; Murphy 1971b). The earlier post-Hopewell
societies show widespread ceramic similarity throughout
the Midwest. The later Whittlesey tradition represents a
geographically localized complex, similar in many ways
to those protohistoric Iroquoian and Algonkian groups
encountered by the European explorers. The period dur-
ing which the prehistoric occupation of the Hale Farm oc-
curred is thus of considerable importance in understand-
ing the cultural changes associated with the development
of tribal agricultural societies in North America.
The few archaeological sites known prior to a.d. 750, in
northeast Ohio show little evidence for any significant dif-
ferences in the size or composition of the groups which
occupied them, although these sites clearly seem to be sea-
sonally and functionally different. None are clearly agri-
cultural villages although corn and squash are present at
several.
During the past several years a number of early Late
Woodland archaeological sites quite similar to the Hale
Farm site have been excavated or reported upon. These
now permit some understanding of the period between
a.d. 800 and a.d. 1300 in northeastern Ohio.
At the Columbia Road site atop a steep ridge overlook-
ing the Cuyahoga River Valley, test units uncovered four
cultural features, two of which were fire pits (Belovich
and Brose 1982). Most projectile points were notched but
a few were triangular. The plain or cordmarked grit-
tempered ceramics had straight or slightly excurvate rims
with flat lips, some slightly rounded. Over half of the rims
displayed interior cordmarking. Exterior rim decoration
consisted of a single nail-punctate sherd. Two radiocar-
bon dates were a.d. 1 040 ± 80 (DIC #2605), and a.d. 970 +
60 (DIC #2606). Our analyses suggest the site was a family
campsite utilized during the summer.
The Bass Lake site (Fienga and Lee 1982) appears to be
a rather large plant-collecting and fishing camp located
on the north shore of Bass Lake, headwaters of the Cha-
grin River. Excavation revealed 26 oblong, basin-shaped
pits, and single radiocarbon date (DIC #2457) of a.d.
1200 ± 45 was obtained from charcoal.
1985
PREHISTORIC HALE FARM OCCUPATION
57
The projectile points in situ include a few Corner
Notched and Levanna types, as well as numerous points
intermediate between Levanna and Madison. The ceram-
ics are grit-tempered, with cordmarked exteriors and
plain interiors. Rims are everted, or slightly incurving,
and may have a folded incipient collar. Vessel bodies are
rounded. Decoration, restricted to the exterior neck and
rim, to the lower edge of the collar, and to the lips of the
vessels, includes incising, punctates, and cordwrapped
cord impressions, broadly comparable to Wayne (Halsey
1 968) or to Allegan and Spring Creek wares between a.d.
800 and a.d. 1 350 in Michigan (Brashler 1981); to the a.d.
900 to 1200 Clemson’s Island complex of Pennsylvania;
or to the Carpenters Brook phase of New York (Ritchie
1969) .
At the Kernisky village site, on the lower Chagrin
River, recent excavations (Bush 1982) have revealed what
appears to be a floodplain terrace horticultural village
with the remains of several structures. The ceramics are
cordmarked on the exterior only. They are grit-tempered
or grit-and-shell-tempered, with slightly everted rims and
flat, slightly rounded lips, most of which are notched. A
few ceramic vessels from the earlier Cleveland Museum of
Natural History testing in 1971 and 1979 show a folded
incipient collar. All of the recovered projectile points
could be considered examples of the Madison point type.
Although neither the completed analyses nor the absolute
dating are yet available, in my judgement the lithic reduc-
tion and ceramic tradition seen at the Kernisky village
should place the major occupation late in the Fairport
phase, dated somewhere between a.d. 11 50 and a.d. 1350
(Brose 1 978c:90— 9 1 ).
In 1974, as part of an archaeological survey program
for the Cleveland Sewer District, excavations tested and
completely exposed a small prehistoric campsite about 25
km to the north of Hale Farm. A remarkably similar as-
semblage of stone tools and ceramics was encountered in
the refuse-filled hearths and postholes of two circular to
oval single post structures on the second terrace of the
Cuyahoga River (Brose and Pratt 1976:3-8). While no
animal bone was preserved at that site, a charred cob of
eight-row maize was recovered. With no organic remains
appropriate for direct dating, I speculated that the site
was occupied between a.d. 900 and a.d. 1300 based on the
styles of its projectile points and ceramics. It also seemed
related to an early occupation on the high bluffs to the
west where still earlier test excavations had revealed an
overlying large late prehistoric village on Tuttle Hill
(Greenman 1937). When Tuttle Hill was itself destroyed
by the construction of the 1-77 / 1-480 interchange, salvage
excavation recovered a single burial associated with grit-
tempered, exterior and occasionally interior cordmarked,
notched-lip ceramics typical of that early Late Woodland
campsite. These were dated to a.d. 1048 ± 100 (CWRU-
14).
While no direct dates are available, the artifacts, fea-
tures, and possible structures encountered in testing the
Drivers site on the lower Tinker’s Creek floodplain, are
also nearly identical to those from Hale Farm ( Lee 1 982).
Beyond those relatively thoroughly explored sites
within the region, similar ceramic and lithic artifacts have
been encountered in test excavations. In the river valleys,
the Kurtz and the Lee village sites, the Rhodes Farm site,
and the Young sites are relatively large agricultural sites
located on the second terraces of the Cuyahoga, Chagrin,
and Grand rivers. Along the smaller tributaries of these
rivers, well into the uplands, the Mohawk Park Shelter,
the Pineway Trails site, the Gillie (Bernhardt 1973), the
Krill (Prufer, personal communication 1983), the Stow
and Little Mountain Shelters, the Cleveland Zoo, and the
Doan Brook sites are all very small reoccupied fall and
winter hunting camps. It is likely that the sporadic recov-
ery of similar artifacts underlying some early nineteenth-
century activities on the Cuyahoga River floodplain (cf.
Jeff Richner, personal communication 1984; Lee et al.
1983) represent small, even more seasonally limited, sin-
gle family activities. These sites were possibly occupied
for collecting tubers and greens during the early spring,
usually called “the starving moon” by Indians throughout
this area.
There are also a few contemporary sites of differing
sizes such as the Avon Plant site (Brose and Morse 1976),
the Cahoon Creek and the Greenhouse sites (Brose,
White, and Ford 1983), a lower component of the Reeve
site, dated at a.d. 1065 ± 100 (CRWU-13) (Brose 1985a),
and the Ashtabula Gulf site reported by Kraus (1942).
These sites all sit on sand ridges cut by streams as they
enter Lake Erie. They were reoccupied as medium-sized
agricultural villages and as seasonal fishing camps.
By combining evidence from these sites it is possible to
glimpse new economic and geographic patterns which
were adopted by the Indians of this early Hale phase of
northeast Ohio’s Whittlesey cultural tradition between
a.d. 900 and a.d. 1200. This was a period of significant
climatic change, beginning with winters which were both
colder and drier than normal and with mild summers
which were longer and wetter. By the end of this phase,
winters had become rather mild and quite wet while, ex-
cept for a narrow zone along Lake Erie from just west of
Cleveland to Pennsylvania, summers were not only long
and warmer, but also drier (cf. Barreis et al. 1976; Brose
1980; Wendland and Bryson 1974). Such summertime
changes would have made the social commitment to
floodplain corn agriculture easier, and the shorter milder
winters with heavier snow cover would have made late
winter hunting and early spring fishing and plant collect-
ing increasingly reliable.
58
DAVID S. BROSE
No. 41
Conclusion
It must have been some late summer day nearly a thou-
sand years ago that the Hale Farm site, surrounded by
forests, was first cleared and occupied. The prehistoric
group appears to have included a dozen or so American
Indians. Probably there were two or three related fami-
lies. Where these people spent that spring and early
summer is not known. Possibly they had come from one
of those archaeological sites known along the Cuyahoga
River terraces in nearby portions of the valley. Those sites
had been occupied by somewhat larger groups whose pot-
tery and stone tools suggest close relationships with those
found at the Hale Farm site. (Brose 1975; Brose et al.
1981) Those sites also yielded some cultivated maize and
show evidence of animals taken in the late spring or early
summer, including large numbers of fish and migratory
waterfowl.
One or two semicircular to circular prehistoric struc-
tures were built at the Hale Farm site. Both were likely to
have had a covering of bark or reed mats tied over a sim-
ple ring of saplings placed into the ground and tied to-
gether at the top. It is unknown whether any interior brac-
ings were used, although some of the historic Indians of
the Great Fakes who constructed similar houses did use
them (cf. Brose 1970:37-38).
While spending that summer at the Hale Farm site,
these Indians gathered the ripening wild fruits and nuts
and hunted a variety of animals as they waited to harvest
their small crops of corn from the sandy floodplain ter-
races to the east and north. A wide variety of domestic
activities were performed, only some of which have left
archaeologically recovered remains. Chipped stone tools
were made from cobbles picked up in the nearby Cuya-
hoga River and from stone obtained on expeditions to
quarries and outcrops which were several day’s travel
from the Hale Farm site. The stone tools were used for a
range of cutting, scrapping, and piercing work. They were
also broken and resharpened, and some were lost.
Pottery was also made at the site, from the local clays
mixed with fragments of crushed and burned rock. It was
a simple shape, hand built in several different sizes to ac-
cord with the group’s functional needs for both cooking
and for storage. A good deal of it was fragile and several
of the vessels were broken.
With the coming of cooler fall weather after the har-
vest, the natural resources which had sustained the group
for several months were less easily obtained. The fires
would have been put out and the remains of the last meals,
along with the broken and discarded bowls and jars, were
thrown into the pits and hearths. The bark or skin mat-
tings would be taken from the sapling framework of the
houses. Dried nuts and corn and the remaining smoked
meat could be packed along with a few personal posses-
sions, and each family would leave the valley for their
winter hunting camps in the upland. They probably
planned to meet again the following summer, but they
never returned to the Hale Farm site itself. The sapling
framework of the houses decayed in place. Through cen-
turies falling leaves and the spring floods from Hale Run
filled in and eventually covered completely the aban-
doned fire hearths and half-emptied storage pits. The low
rise at the forks of Hale Run was reclaimed by the silence
and the forests, until these, in turn, were disturbed by Ab-
raham Miller’s axe.
Acknowledgments
Beyond the efforts of the various field crew members, 1 wish
to express thanks to the museum volunteers Bettyann Ball, Dan
Best, Nathalie Boswell, Robert Burns, Jr., William Fienga, Judy
McDonald, Jack Moviel, Susan Russell, Lars Sande, Ronald
Skrbin, and Pat Stockwell.
Special consideration is also due to Dean Zimmerman who
curated and resurrected field notes, artifacts, and field invento-
ries from a decade of neglect. Finally I thank Siegfield Buerling
for never having given up all hope that I would eventually com-
plete this report, among others still due.
References
Andreas, Barbara K. 1980. The flora of Portage, Stark, Summit
and Wayne counties, Ohio. Ph.D. diss.. Department of An-
thropology, Kent State University, Kent.
Baerreis, David A., Reid Bryson, and Joan Kutzbach. 1976.
Climate and culture in the western Great Lakes region. Mid-
continental Journal of Archaeology 1(1 ):39— 58.
Banks, P. O., and Rodney M . Feldmann, eds. 1 970. Guide to the
geology of Northeastern Ohio. Northern Ohio Geological
Society, Cleveland.
Belovich, Stephanie, and David S. Brose. 1 982. Survey and eval-
uative testing within the Cuyahoga Valley N.R.A.: Sites
33Su87 and 33Sul02. Archaeological Research Reports
40:1-86. Cleveland Museum of Natural History, Cleveland.
Benson, Donna Lynne. 1978. A reconsideration of the Carolina
pattern. Master’s thesis in Anthropology, Case Western Re-
serve University, Cleveland.
1980. Taxonomy in historical archaeology. Papers of
the Conference on Historical Archaeology, Vol. 1 1. In press.
Bernhardt, Jack E. 1973. Gillie Rockshelter: A Late Woodland
phase in Summit County, Ohio. Master’s thesis in Anthro-
pology, Kent State University, Kent, Oh.
Brashler, Janet G. 1981. Early Late Woodland boundaries and
interaction: Indian ceramics of southern Lower Michigan.
Anthropological Series 3(3): 1-359. Publications of the Mu-
seum, Michigan State University, East Lansing.
Braun, E. Lucy. 1950. Deciduous forests of eastern North Amer-
ica. Philadelphia: Blakeston Co.
1955. The phytogeography of unglaciated eastern
United States and its interpretation. Botanical Review
21:297-375.
Brose, David S. 1970. Middle Woodland cultural ecology and
social organization in the Upper Great Lakes. Case Western
Reserve University Papers in Anthropology No. 1 , Cleveland.
1971. The Southeast ceremonial complex in Northeast-
ern Ohio. The Ohio Archaeologist 2 1(4): 16- 19.
1985
PREHISTORIC HALE FARM OCCUPATION
59
1973a. Report on the analysis of early pioneer settle-
ment in the Connecticut Western Reserve, 1796- 1837: A sta-
tistical evaluation of the archaeological and historic data.
Report on file with the National Endowment forthe Humani-
ties. Washington, D.C.
1973b. A preliminary report on recent excavations at
the South Park site, Cuyahoga County. Ohio. Pennsylvania
Archaeologist 43( I ):25— 43.
1975. Preliminary results of the archaeological recon-
naissance of the Lower Tinkers Creek Valley. The Cuyahoga
Valley National Park Federation Board, pp. 1-58. Peninsula.
1976a. Locational analysis in the prehistory of North-
eastern Ohio. In Cultural change and continuity: Essays in
honor of James B. Griffin, edited by Charles Cleland,
pp. 3-18. New York: Academic Press.
1976b. The Whittlesey occupations of Northeast Ohio:
A second approximation. In The late prehistory of the Lake
Erie drainage basin, edited by David S. Brose, pp. 22-51.
Scientific Papers of the Cleveland Museum of Natural His-
tory, Cleveland.
1978a. The late prehistory of the Upper Great Lakes. In
Handbook of North American Indians. Vol. 15, Northeast,
edited by Bruce Trigger, pp. 569-582. Washington, D.C.:
Smithsonian Institution.
1978b. Some comments on “Temporal models in East-
ern U.S. prehistory” by James B. Stoltman. Current Anthro-
pology 19(4): 729-731.
1978c. A model of changing subsistence technology in
the Late Woodland of Northeast Ohio. In Lithic Technology
and Subsistence, edited by David Davis. Vanderbilt Univer-
sity Publications in Anthropology 20:87-1 16. Nashville.
1980. The Late Prehistoric occupations of the South
Central Lake Erie area. Paper presented at the 1 4th Annual
Meeting of the Ontario Archaeological Society Symposium,
London, Ontario.
1 984. A statistically dendritic analysis of Whittlesey ce-
ramics with an illustrated type-variety classification. Manu-
script on file at the C.M.N.H. Archaeological Archives.
1985a. Recent salvage excavations at two village sites
near Cleveland, Ohio: Redating the occupations of the Tuttle
Hill and Reeves sites. The Ohio Archaeologist. In press.
1985b. South Park: A stratified Whittlesey site in Cuya-
hoga Co., Ohio. Scientific Papers of the Cleveland Museum
of Natural History. In press.
Brose, David S., S. Belovich, M. Brooslin, R. Burns, J. Hall, H.
Haller, C. Pierce, and C. Ubbeohde. 1981. Prehistoric and
historic archaeological investigations of the Cuyahoga Valley
National Recreation Area, Ohio. Archaeological Research
Reports of the Cleveland Museum of Natural History
30:1-524.
Brose, David S., and Donna L. Benson. 1985. The Jonathan
Hale Farm excavations and a socio-ecological analysis of pi-
oneer settlement in the Western Reserve. Manuscript in pre-
paration.
Brose, David S., and David Morse. 1976. An archaeological
survey of the Avon Lake district. Report to the Director of
Utilities, Avon Lake, Ohio. Archaeological Research Reports
of the Cleveland Museum of Natural History 16:1-23.
Brose, David S., and G. Michael Pratt. 1976. Salvage excava-
tions at the Walnut Tree site, 33Cu35: Archaeology of an im-
pacted prehistoric campsite. Cleveland Regional Sewer Dis-
trict Engineer Technical Reports.
Brose, David S., and John F. Scarry. 1 976. Boston Ledges shel-
ter: Comparative analysis of Early Late Woodland occupa-
tions in Summit County, Ohio. Midcontinental Journal of
Archaeology 1(2): 1 79-228.
Brose, David S., Gregory Wentzel, Helga Bluestone, and Pat
Essenpreis. 1976. Conneaut Fort: A prehistoric Whittlesey
focus village in Ashtabula County, Ohio. Pennsylvania Ar-
chaeologist 46(4):29-77.
Brose, David S., Nancy M. White, and Sarah Ford. 1983. Sev-
enteen prehistoric sites in Northern Ohio. Manuscript on file
at the Cleveland Museum of Natural History, Department of
Archaeology, Cleveland.
Bush, David R. 1982. The earlier Late Woodland in the Chagrin
River — The Kernisky site. Paper presented at the 45th An-
nual Meeting of the Midwestern Archaeological Conference.
Cleveland.
Cleland, Charles E. 1966. Prehistoric animal ecology and eth-
nozoology in the Upper Great Lakes. Anthropological Pa-
pers No. 26. Museum of Anthropology, University of Michi-
gan, Ann Arbor.
1976. A re-examination of the focal-diffuse adaptation
model with a note of evolutionary implications. Midcontinen-
tal Journal of Archaeology 1( 1 ):59— 76.
Converse, Robert. 1964. Ohio stone types. The Ohio Archaeol-
ogist 1 3(4):78— 1 20.
Fienga, William, and Alfred Lee. 1982. The Bass Lake site,
Geauga County, Ohio. Paper presented at the 45th Annual
Meeting of the Midwest Archaeological Conference. Cleve-
land.
Fitting, James E. 1964. Ceramic relationships of four Late
Woodland sites in Northern Ohio. Wisconsin Archaeologist
45 (4): 160- 175.
1965. Late Woodland cultures of Southeastern Michi-
gan. Anthropological Papers No. 24. Museum of Anthropol-
ogy, University of Michigan, Ann Arbor.
Fitting, James E., Stephan R. Claggett, and Stephen R. Triech-
ler. n.d. Prehistoric projectile points of Michigan; Great
Lakes informant. Topics in History Ser. 2, No. 5, pp. 1-11.
Michigan Department of State. Michigan History Division,
Lansing.
Ford, Richard I. 1977. Evolutionary ecology and the evolution
of human ecosystems: A case study from the Midwestern
U.S. A. In Explanation of prehistoric change, edited by James
W. Hill, pp. 153-184. Albuquerque: Univ. of New Mexico
Press.
Gordon, Robert B. 1966. Vegetation of Ohio at the time of orig-
inal survey. Map. Columbus, Ohio. Biological Survey.
1 969. The natural vegetation of Ohio in pioneer days.
Bulletin of the Ohio Biological Survey, n.s., 3(2): 1-1 13. Co-
lumbus.
Greenman, Emerson. 1930. Field notes of survey and excava-
tion in Northern Ohio 1929-1930. On file at the Ohio Histori-
cal Society, Division of Archaeology. Columbus.
1935a. Excavation of the Reeve Village site. Lake Co.,
Ohio. The Ohio State Archaeological and Historical Quar-
terly 44( 1 ):2— 64.
1935b. Seven prehistoric sites in Northern Ohio. Tlte
Ohio State Archaeological and Historical Quarterly
44(2):220— 237.
1937. Two prehistoric villages near Cleveland, Ohio.
The Ohio State Archaeological and Historical Quarterly
46(4) :305— 366.
60
DAVID S. BROSE
No. 41
1939. The Wolf and Furton sites, Macomb Co., Michi-
gan. Occasional Contributions of the Museum of Anthropol-
ogy No. 8. University of Michigan, Ann Arbor.
Griffin, James B. 1943. The Fort Ancient aspect. Ann Arbor:
University of Michigan Press.
1 946. Cultural change and continuity in Eastern United
States archaeology. In Man in Northeastern North America,
edited by Frederick Johnson. Papers of the Robert S. Pea-
body Foundation for Archaeology, 3:37-95. Andover.
1967. Eastern American archaeology: A summary.
Science 156(3772): 175- 191. Washington, D.C.
1978. Late prehistory of the Ohio Valley. In Handbook
of North American Indians, Vol. 15, Northeast, edited by
Bruce Trigger, pp. 547-566. Washington, D.C.: Smithsonian
Institution.
Hall, John. 1980. Field notes on the Quaternary geology and
modern geomorphology, soils, and water resources of North-
east Ohio. On file. Department of Geology, CWRU.
Halsey, John R. 1968. The Springwells Mound Group of Wayne
County, Michigan. In Contributions to Michigan Archaeol-
ogy, edited by J. Fitting, pp. 79- 1 59. Anthropological Papers
No. 32. Museum of Anthropology, University of Michigan,
Ann arbor.
Horton, John J. 1961. The Jonathan Hale Farm: Achronicalof
the Cuyahoga Valley. The Western Reserve Historical So-
ciety, Publication No. 116. Cleveland.
Keeley, Lawrence H. 1974. Technique and methodology in mi-
crowear studies: A critical review. World Archaeology
5(3):323— 336. London.
1980. Experimental determination of stone tool uses: A
microwear analysis. Prehistoric Archaeology and Ecology
Series. The University of Chicago Press, Chicago.
Kraus, Bertram S. 1942. Archaeological survey of Ashtabula
County, Ohio, 1941. Mimeographed manuscript on file at
CWRU.
Lee, Alfred M. 1982. The Driver’s site, 33Su24; Evaluative test-
ing and salvage excavations. Cleveland Museum of Natural
History Archaeological Research Reports 41:1 -37.
Lee, Alfred M.,and David S. Brose. 1980. Archaic archaeologi-
cal complexes in Northeastern Ohio. Paper presented at the
14th Annual Meeting of the Ontario Archaeological Society
Symposium. London, Ontario.
Lee, Alfred M., David S. Brose, and Russell Weisman. 1983.
Archaeological reconnaissance of the Akron Water Pollution
Control Facility at Botzum, Northhampton Twp., Summit
County, Ohio. Cleveland Museum of Natural History Ar-
chaeological Research Reports 43:1-54.
Luedtke, Barbara. 1978. Lithic material distributions and inter-
action patterns during the Late Woodland period in Michi-
gan. Ph.D. diss.. Department of Anthropology, University of
Michigan, Ann Arbor.
McKenzie, Douglas, Don Bier, Don Ianone, and Chris Pierce.
1974. The Riverview site: An Early Late Woodland village in
Cuyahoga County, Ohio. Manuscript on file at Cleveland
State University, Department of Anthropology, Cleveland.
Miller, Barry B. 1970. The Quaternary period. In Guide to the
geology of Northeastern Ohio, edited by Phillip Banks and
Rodney Feldmann, pp. 149-164. Northern Ohio Geological
Society, Cleveland.
Miller, George. 1980. Classification and economic scaling of
nineteenth century ceramics. Historical Archaeology
14:1-40.
1983. Ceramic supply in an economically isolated fron-
tier community: Portage County of the Ohio [ric] Western
Reserve 1800-1825. Historical Archaeology 17:19-32.
Morgan, Richard G., and H. Holmes Ellis. 1943. The Fairport
Harbor site. The Ohio State Archaeological and Historical
Quarterly 52(3): 1-62.
Murie, Olhaus J. 1951. The American elk. New York: Stackpole
Press.
Murphy, James L. 1971a. The Fairport Harbor site (33La5),
Lake County, Ohio. Pennsylvania Archaeologist
4 1 (3) :26— 43.
1971b. Whittlesey ceramic types. The Ohio Archaeolo-
gist 21(l):298-303.
1971c. The Lyman site (33La2), Lake County, Ohio.
Pennsylvania Archaeologist 41(3): 12-25.
1972. A note on the South Park ceramics. The Ohio
Archaeologist 22(2):3 1 —35.
Ohio Department of Natural Resources, Division of Recreation
Planning. 1975. Technical Report for the Cuyahoga River
Valley Park Study, 4 vols. Columbus.
Phillips, Philip. 1970. Archaeological survey in the Lower Yaz-
oo Basin, Mississippi, 1949-1955. Peabody Museum of Ar-
chaeology and Ethnology, Harvard University, Papers
No. 60. Cambridge.
Rau, John L. 1968. The evolution of the Cuyahoga River: Its
geomorphology and environmental geology. In Cuyahoga
River wastershed symposium, edited by D. Cooke, pp. 9-41.
Kent State University, Kent.
Read, Matthew C. 1 880. Explorations of a rocky shelter in Bos-
ton, Summit County, Ohio. American Antiquarian
2:203-206. Boston.
Ritchie, A., and J. R. Steiger. 1974. Soil survey of Summit
County, Ohio. USDA Soil Conservation Service in coopera-
tion with ODNR, Washington, D.C.
Ritchie, William A. 1969. The archaeology of New York Slate.
Revised edition. Garden City: Natural History Press.
197 1. A typology and nomenclature for New York pro-
jectile points. New York State Museum and Science Service,
Bulletin No. 384. Second edition, revised. Albany.
Scully, Edward G. 1951. Some Central Mississippi Valley pro-
jectile point types. Museum of Anthropology, University of
Michigan Technical Report 1:1-15. Ann Arbor.
Severinghaus, C. W. 1949. Tooth development and wear as
criteria of age in whitetailed deer. Journal of Wildlife Man-
agement 13(2): 195-2 16.
Shelford, Victor E. 1963. Ecology of North America. Urbana:
University of Illinois Press.
Stout, Wilber, and R. A. Schoenlaub. 1945. The occurrence of
flint in Ohio. Geological Survey of Ohio, Bulletin 46, 4th ser.,
Columbus.
Whallon, Robert C. 1974. Spatial analysis of occupation floors
II: Application of nearest-neighbor analysis. American An-
tiquity 39:16-34.
Wendland, Wayne M., and Reid A. Bryson. 1974. Dating cli-
matic episodes of the Holocene. Quaternary Research
4(l):9-24.
White, George W. 1953a. Pleistocene geology. (Map of Akron
quadrangle) mapping by George White and copying by Jane
Forsyth, 1956. Ohio Geological Survey, Columbus.
1953b. Geology and water-bearing characteristics of
the unconsolidated deposits of Cuyahoga County. In The wa-
ter resources of Cuyahoga County, Ohio, edited by J. D.
1985
PREHISTORIC HALE FARM OCCUPATION
61
Winslow, G. W. White, and E. E. Webber. United States Geo-
logical survey. Water Resource Division, Bulletin No. 26.
1960. Classification of Wisconsin glacial deposits in
Northeastern Ohio. United States Geological Survey, Bul-
letin 1 1 2 1 A: A I —A 1 2.
1 979. Extent of till sheets and ice margins in Northeast-
ern Ohio. Geological Note no. 6, State of Ohio Department of
Natural Resources, Division of Geological Survey. Co-
lumbus.
White, George W., with Stanley M. Totten. 1982. Glacial geol-
ogy of Northeastern Ohio: With a chapter on Pleistocene
beaches and strand lines bordering Lake Erie. State of Ohio,
Department of Natural Resources, Division of Geological
Survey, Bulletin No. 68. Columbus.
Williams, Arthur B. 1936. The composition and dynamics of a
beech-maple climax community. Ecological Monographs
7(3).
1949. The native forests of Cuyahoga County, Ohio.
Cleveland Museum of Natural History, Science Publication
9(1), Cleveland.
Wittine, Arthur H. 1970. A study of Late Pleistocene lake in the
Cuyahoga Valley, Summit County, Ohio. Master’s thesis in
Geology, Kent State University, Kent.
Yarnell, Richard A. 1964. Aboriginal relationship between cul-
ture and plant life in the Upper Lakes region. Anthropologi-
cal Papers No. 23. Museum of Anthropology, University of
Michigan, Ann Arbor.
Ziegler, Allan C. 1973. Inference from prehistoric faunal re-
mains. Addison-Wesley McCaleb Module in Anthropology,
Reading.
' . ,
.
.
STYLE GUIDE FOR AUTHORS:
KIRTLANDIA
Address manuscript correspondence to the editor. Dr.
David S. Brose, Department of Archaeology, Cleveland
Museum of Natural History, Cleveland, Ohio 44106. Re-
prints of this Style Guide for Authors are available on
request from the editor, or the publisher, Kent State Uni-
versity Press, Kent, Ohio 44242.
Manuscripts
General Instructions: The manuscript should have a
uniform style and be submitted exactly as it is to appear in
print. It should consist of the following subdivisions, each
prepared as a unit on separate sheets:
Title page
Abstract Notes
Text Tables
References Figure legends
The manuscript, including references, should be typed
double-spaced on bond or heavy-bodied paper 8 Vi" by
1 1 "(22 cm x 28 cm) with a 1 "(2.5 cm) margin on all sides.
Number the manuscript pages consecutively beginning
with the title page. Submit an original and three copies of
text and illustrations.
Do not divide words at the end of lines; if they are un-
familiar to the printer, they may be incorrectly hyphen-
ated.
Corrections to the manuscript should be typed or
printed legibly in ink.
Do not begin sentences with abbreviations.
The word Figure is not abbreviated in the text, except
when appearing in parentheses: (Fig. 2); (Figs. 4-6).
The spelling of nontechnical terms should be that rec-
ommended in the current Webster’s Third New Interna-
tional Dictionary, Unabridged.
Always spell out numbers when they stand as the first
word in a sentence; do not follow such numbers with ab-
breviations. Numbers indicating time, weight, and mea-
surements are to be in Arabic numerals when followed by
abbreviations (e.g., 2 mm; 1 sec; 3 mil).
Title Page : The title page must contain:
Title
Author’s name (or names)
Insitution from which the paper emanated, with city,
state, and zip code
Number of text pages plus bibliography, figures,
graphs, charts, and tables
Abbreviated title (running headline) not to exceed 48
characters and spaces
Name, address, and telephone number of the person to
whom the proof is to be sent.
Abstract: The abstract should consist of 250 words or
less. When published, it will precede the introductory sec-
tion of the text. The abstract should be written in com-
plete sentences and should succinctly state the objectives,
the experimental design of the paper, and the principal
observations and conclusions; it should be intelligible
without reference to the rest of the paper.
7exT:The general style guide for Kirtlandia is the 13th
revised and expanded edition of The Chicago Manual of
Style (1982). The author-date system of text references
and a full (style B) reference list are required. References
lists must be double-spaced. Titles of works should be
capitalized according to the standard set forth in the Chi-
cago Manual sec. 7. 123, and titles of journals, series, and
papers or proceedings should be spelled out in full. In the
text, references to the literature should be cited by au-
thor’s surname followed by year of publication.
. . . the study of Kirtland ( 1836) reveals . . .
. . . the report of Squier and Davis (1848) describes . . .
. . . studies by Darwin (1859) and Wallace (1868) sug-
gest . . .
. . . the earliest report (Whittlesey 1871) revealed . . .
. . . previous discussions (Stout et al. 1943) indi-
cate . . .
When references are made to more than one paper by the
same author or authors published in the same year they
should be designated in their order of appearance in the
text as, for example. Young (1947a), (Young 1947b), or
(Young 1947a, 1947b) and in the reference list as follows:
Young, C. C. 1947a. Mammal-like Reptiles from Lufeng,
Yunnan, China. Proceedings of the Zoological Society
of London 1 17:537-597.
1947b. On Lufengo saurus magnus Young and
Additional Finds of Lufengo saurus huenei Young. Pale-
ontologica Sinica, n.s., 12:1-53.
The list of references must be arranged first alphabetically
by author’s or authors’ surname(s) and chronologically
for works by any single author or group of authors in the
following style: Author’s surname (or names) and initials;
year of publication; complete publication data, other
than physical facts, as follows:
American Ornithologists’ Union. 1957 Checklist of North
American Birds. Baltimore, 5th ed.
Aronson, J. L., T. J. Schmitt, R. C. Walter, M. Taieb, J.
J. Tierelin, D. C. Johanson, C. W. Naeser, and A. E.
M. Nairn. 1977. New Geochronologic and Paleomag-
netic Data for the Hominid-Bearing Hadar Formation
of Ethiopia. Nature 267:323-327.
Beck, W. M. 1954. Studies in Stream Pollution Biology.
Quarterly Journal of the Florida Academy of Sciences
17:211-227.
Gartley, Richard, Jeff Carskadden, and Tim Gregg. 1973.
The Philo Site, A Fort Ancient Component in the Cen-
tral Muskingum Valley. Ohio Archaeologist 23(4):
15-19.
64
STYLE GUIDE FOR AUTHORS
No. 41
Greber, N’omi. 1976. Within Ohio Hopewell. Ph.D. diss..
Department of Anthropology, Case Western Reserve
University, Cleveland.
Hopwood, A. T. 1926. Fossil mammalia. In The Geology
and Paleontology of the Kaiso Bone Beds. Ed. E. J.
Wayland. Occasional Papers of the Geological Survey
of Uganda, 2: 13-36.
Mayr, E. 1969. Principles of Systematic Zoology.
McGraw-Hill, New York.
Stout, W., and R. A. Schoenlaub. 1945. The Occurrence
of Flint in Ohio. Ohio Geological Survey, 4th ser., Bul-
letin 46.
Uzzell, T. M. 1964. Relations of the Diploid and Triploid
Species of the Ambystoma jeffersonianum Complex
(Amphibia, Caudata). Copeia 1964:257-300.
Notes: Footnotes are not accepted. Endnotes to the
text should be as limited as possible. Endnotes must be
numbered consecutively. The corresponding reference
superscript numbers must be clearly indicated in the text.
Additional references to the identical notes are to be
numbered with the next following consecutive number.
There were subjects for all six experiments.1 These were
obtained through a variety of methods.2
Notes to a table should be typed directly beneath the table
and numbered, 1, 2, 3, etc. They should not be numbered
in sequence with the endnotes to the text.
Tables: All tables must be cited in the text. Tabular
material should be simple and uncomplicated, with as few
vertical and horizontal rules as possible. Table titles
should be complete but brief. Information other than that
defining the data should be presented in the text.
Figures: All figures must be cited in the text. Figure
legends are to be numbered consecutively as follows: Fig.
1 . . . , Fig. 2 ... , and should follow the sequence of
reference in the text.
Whenever possible, figures should be integrated into
the text. Reference to relevant text passages can often re-
duce the length of legends and avoid redundancy.
Metric System: The metric system should be used for
all measurements, weights, etc. Temperatures should be
expressed in degrees Celsius. Metric abbreviations should
be expressed in lowercase without periods.
Symbols: When preceded by a digit, the following sym-
bols are to be used: % for percent; ° for degree; ' for min-
ute; " for second.
Abbreviations: Consult the 13th revised and expanded
edition of The Chicago Manual of Style for permissible
abbreviations.
Illustrations: Photographic illustrations are costly. Au-
thors should limit plates to those which are needed to
adequately present the findings. All illustrations must be
original drawings or high-quality photographic prints,
submitted in complete and finished form with adequate
labeling.
Photographic Prints submitted for reproduction must
be of adequate contrast and if multiple prints are included
in a single figure, they should be of mylar uniform tone.
Prints should be on white, nonmatte paper.
Line Drawings should be drawn with black ink on
medium-weight white paper, mylar or light-weight art-
board. To reduce weight and postal charges, photo-
graphic prints may be submitted in lieu of original draw-
ings. The artwork should be sharp and black to achieve
maximum contrast.
Use stippling and hatching techniques to achieve tonal
quality. Avoid the use of shading (pencil, wash, or air-
brush) for tonal effect unless the drawing is to be repro-
duced as a halftone.
The reverse side of each illustration should indicate:
Author’s name; figure number; top side of illustration;
reduction requested; “Review copy” on those copies in-
tended only for reviewers. Do not fasten illustrations with
paper clips, staples, etc., since they will mar the surface of
the illustrations. Illustrations should be shipped flat and
protected by heavy cardboard.
Illustrations cannot be reduced by less than 20% of
their submitted size. Submitted line drawings cannot ex-
ceed 1 1" x 14" (28 cm x 36 cm). Lettering and labels must
be readable after reduction. When reduced, the minimum
height of a capital letter should not be less than 2.5 mm for
a photomicrograph and 1 mm for a graph or chart.
Printed figures may not exceed 9" x 6" (22.9 cm x 15.3
cm).
General Illustration Instructions: Original illustrations,
and two sets of good-contrast photographic copies should
be submitted with the manuscript for review purposes.
Photocopies of illustrations made on an office duplicat-
ing machine are not acceptable; copies of the manuscript
only will be permitted.
Figures should be trimmed straight on all sides and
“squared.” Figures should be mounted on strong bristol
board of about 1 5 points (0.4 mm) thickness with at least a
1" (2.5 cm) margin surrounding the figure or grouping of
figures.
Illustrations grouped to form a single figure should be
of similar density and tone to prevent loss of detail.
Labels should be large enough to allow for suitable re-
duction and sturdy enough to withstand mailing and
handling in the production process. Labeling should be
done directly on the drawing or photographic print, never
on an overlay. All labeling should be placed at least 1 / 8"
(3 mm) in front of the edges of the illustration. To achieve
adequate contrast between the label or letter and its back-
ground, place white labels over dark backgrounds and
black labels over light backgrounds.
1985
STYLE GUIDE FOR AUTHORS
65
Numbering: Figures, including charts and graphs, must
be numbered consecutively.
Miscellaneous: The editor and publisher reserve the
right to return to the author for revision manuscripts and
illustrations that are not in proper finished form.
Upon acceptance of an article for publication, the au-
thor will be asked to sign a transfer agreement, transfer-
ring copyright to the publisher.
Proofs: A single set of page and illustration proofs will
be sent to the author. All corrections should be marked
clearly, directly on page proofs. Authors are solely re-
sponsible for proofreading their articles.
Reprints: Reprints may be purchased at prices quoted
on the reprint order form. Reprint orders should be
returned with the proofs. It is important to order initially
a sufficient quantity of reprints, since the price is substan-
tially higher if they are ordered after the paper has been
published.
Publication in Kirtlandia requires consideration of edi-
torial costs. These will be assessed at $32 per page for so-
ciety members and $40 per page for nonmembers. Papers
in excess of 32 printed pages may be charged full produc-
tion costs. A page charge invoice will be sent with page
proofs. Unsupported authors may request a waiver of
page charges and will be asked to supply statements from
their supervisor (department chairman, their advisor in
the case of graduate students, and dean) that institutional
and grant funds are not available. To help defray abnor-
mal costs, authors will be requested to pay $75 per page
for photographic illustration pages that exceed 25% of the
total number of printed text pages. These payments will
commonly be levied but are not a prerequisite to the pub-
lication of any article.
CONTENTS
Benthic Colonization in Fresh Water: A Synthesis
Michael J. S. Tevesz 3
An Inventory of the Vertebrates of the Grand River Terraces, Ashtabula County, Ohio
Timothy O. Matson 15
An Echinocarid Crustacean, Echinocaris Auricula, from the Late Devonian of West Virginia
Joseph T. Hannibal and Rodney M. Feldmann 22
Life Habits and Distribution of Riverine Lampsilis Radiata Luteola (Mollusca: Bivalvia)
Michael J. S. Tevesz, David W. Cornelius, and J. Berton Fisher 21
The Prehistoric Occupation of the Hale Farm, Bath Township, Summit County, Ohio
David S. Brose 35
Style Guide for Authors: Kirtlandia 63
1
CLEVELAND, OHIO
NUMBER 42
■
KIRTLANDIA
•NATURAL HISTORY*
PHYSICAL ANTHROPOLOGY
Distal Radius and Proximal Femur Fracture Patterns in the Hamann-Todd Skeletal Collection
Robert P. Mensforth, Susan A. Surovec, and James R. Cunkle 3
PALEONTOLOGY
Echinocaris, A Mid-Paleozoic Crustacean: An Annotated Bibliography
Joseph T. Hannibal and Rodney M. Feldmann 25
BOTANY
A Bluegrass New To Ohio: Poa Saltuensis Fern. & Wieg.
James K. Bissell 53
31 MAY 1987
KIRTLANDIA
The Scientific Publication of the Cleveland Museum of Natural History
David S. Brose, Editor
Brief History and Purpose
Kirtlandia is a publication of the Cleveland Museum of Natural History, an orga-
nization incorporated in 1922 whose predecessor, the Ark , was founded in Cleveland
by Jared Potter Kirtland in 1837. Published since 1926, Kirtlandia is a continuation
of the earlier old series, Scientific Publications volumes 1 to 7 (1928-1937), and new
series volumes 2 to 4 (1962- 1965), and the Survey of Ohio Fluted Points. Supported
by the Kirtlandia Society of the Cleveland Museum of Natural History, Kirtlandia is
devoted to the publication of scientific papers in the various fields of inquiry within
the Museum’s sphere of interest: Cultural and Physical Anthropology; Archaeology;
Botany; Geology; Paleobotany; Invertebrate and Vertebrate Paleontology; System-
atic Ecology; and Invertebrate and Vertebrate Zoology. Issues will vary from single
monographs to collections of short papers, review articles, and brief research notes.
Kirtlandia is distributed by The Kent State University Press, Kent, Ohio 44242.
Authorization to photocopy items for internal or personal use, or the internal or
personal use of specific clients, is granted for libraries and other users registered with
the Copyright Clearance Center (CCC) Transactional Reporting Service, provided
that the base fee of $00.50 per copy, plus .25 per page is paid directly to CCC, 27
Congress Street, Salem MA 09170. The fee code for users of the Transactional Re-
porting Service is 0075-6245/87 $00.50+. 25.
Kirtlandia is abstracted in Biological Abstracts and indexed in Bibliography and
Index of Geology and Zoological Record.
Kirtlandia No. 42
© 1987 by The Cleveland Museum of Natural History
KIRTLANDIA
THE CLEVELAND MUSEUM OF NATURAL HISTORY
Cleveland, Ohio
Spring 1987
Number 42
PHYSICAL ANTHROPOLOGY
Distal Radius and Proximal Femur Fracture Patterns in the Hamann-Todd Skeletal Collection
Robert P. Mensforth, Susan A. Surovec, and James R. Cunkle 3
PALEONTOLOGY
Echinocaris, A Mid-Paleozoic Crustacean: An Annotated Bibliography
Joseph T. Hannibal and Rodney M. Feldmann 25
BOTANY
A Bluegrass New To Ohio: Poa Saltuensis Fern. & Wieg.
James K. Bissell 53
ISSN: 0075-6245
KIRTLANDIA
EDITOR
DAVID S. BROSE
Cleveland Museum of Natural History
ASSOC I A TE EDITORS
MARY BAUM
Research Librarian
Cleveland Museum of Natural History
JAMES K. BISSELL
Curator of Botany
Cleveland Museum of Natural History
BRUCE LATIMER
Curator of Physical Anthropology
Cleveland Museum of Natural History
SONJA TERAGUCHI
Manager of Collections
Cleveland Museum of Natural History
JOSEPH T. HANNIBAL
Curator of Invertebrate Paleontology
Cleveland Museum of Natural History
SIBLEY HOOBLER
Editorial Assistant
Cleveland Museum of Natural History
EDITORIAL A D VISOR Y BOA RD
RODNEY FELDMANN
Professor of Geology
Kent State University
MICHAEL C. HANSEN
Geologist
Ohio Geological Survey
RICHARD MEINDL
Associate Professor of Anthropology
Kent State University
G. MICHAEL PRATT
Associate Professor of Anthropology
Heidelberg University
DAVID H. STANSBERY
Director, Museum of Zoology
Ohio State University
FREDERICK H. UTECH
Curator of Botany
Carnegie Museum of Natural History
ED VOSS
Curator of the Herbarium
University of Michigan
ANDREW M. WHITE
Professor of Biology
John Carroll University
DISTAL RADIUS AND PROXIMAL FEMUR FRACTURE
PATTERNS IN THE HAMANN-TODD
SKELETAL COLLECTION
ROBERT P. MENSFORTH
Department of Anthropology
University of Oklahoma
Norman, Oklahoma 73019
and
SUSAN A. SUROVEC
JAMES R. CUNKLE
Department of Anthropology
Cleveland State University
Cleveland, Ohio 44115
Abstract
The study presents a retrospective analysis of distal radius and
proximal femur fractures that occurred in 938 Hamann-Todd Collec-
tion skeletons. Individuals included in the investigation were retrieved
from dissecting room cadavers in Cleveland, Ohio, between the years
1910 and 1938. Demographic analysis showed that mean age at death
was 41.9 years for Blacks and 53.8 years for Whites examined in the
study. Observations of remodeling status and side of involvement were
recorded for all fractures identified. It was found that the age, sex, and
race specific fracture patterns, which characterize the early twentieth-
century Hamann-Todd sample, strongly correspond to those seen in
modern American and European communities. The most striking
difference between Hamann-Todd fracture patterns and those seen in
modern groups concerns that much greater total frequency of traumatic
injuries that occurred in the former group. Although hip fractures
appear to be a primary result of age progressive skeletal fragility, it is
suggested that the early onset, and high incidence, of distal radius
fractures that occur in climacteric Caucasian women may be more
directly due to accidents initiated by a greater frequency, intensity, and
duration of vasomotor disturbances, which are known to accompany
estrogen withdrawal in perimenopausal White females.
Introduction
Over the past several decades American and European
populations have experienced a demographic shift
whereby increasing numbers of people survive to old age
(Gordon 1984; Johnell et al. 1984). Although continuing
advances in medical technology and improvements in
health care delivery systems are primarily responsible for
such trends, we also have witnessed changes in the major
causes of morbidity and mortality in these groups
(Kilbourne and Smillie 1969). Thus, it is not surprisingto
find that contemporary sociological, epidemiological,
and clinical researchers have been devoting greater
attention to the health care problems of the aged.
One of the major problems that characterizes the
geriatric segment of our society is a high incidence of
traumatic injuries. For example, Gordon (1984) reports
that 225,000 hip fractures occur annually in the United
States. Over two-thirds of these occur in elderly women.
The financial costs associated with hip fractures in the
United States now approximates 3.8 billion dollars a
year. However, old-age fractures are costly not only in
terms of clinical management, but also in constituting a
major risk of mortality. Compared to younger individu-
als, it is well recognized that old people who have frac-
tures are much more vulnerable to complications as a re-
sult of surgery and/or immobilization (Robbins 1974).
These include thromboembolism, pneumonia, and death.
Thus, it is not remarkable that hip fractures are now the
12th leading cause of death among the elderly in the
United States (Gordon 1984).
These problems motivated Buhr and Cooke (1959) to
conduct one of the first comprehensive surveys of
fracture epidemiology in a modern urban industrial
community. They examined fractures in individuals from
England and Wales whose ages ranged from birth to 80+
years. It was found that traumatic injuries were charac-
terized by marked age and sex specific patterns, which
could be classified as one of four types (Buhr and Cooke
1959). These are L-Type, J-Type, A-Type, and Compos-
ite fracture patterns.
For example, an L-Type fracture pattern (e.g., supra-
condylar fractures of the distal humerus; Fig. 1, upper
left) typically shows a peak incidence at an early age,
followed by a subsequent decrease in frequency to levels
that are relatively insignificant. In contrast, the J-Type
pattern identifies old-age fractures (Fig. 1, upper right).
Here, the frequency of traumatic injury is low throughout
childhood, and early and middle adulthood. Beyond 60
years of age J-Type fractures (e.g., hip fractures) show an
Kirtlandia, No. 42
® by The Cleveland Museum of Natural History
4
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
A-TYPE COMPOSITE (A,J)
Fig. 1. Typical age and sex specific fracture patterns observed in modern human groups. Fracture
frequencies are variably reported as number of cases per million, annual incidence per 10,000,
incidence per 10,000 per age class, or percent of individuals affected per age class. The patterns
shown here were adapted from Buhr and Cooke (1959).
1987
HAMANN-TODD SKELETAL COLLECTION
5
age progressive increase, and usually reach peak inci-
dence in the oldest decade of life. The A-Type fracture
pattern (e.g., phalangeal fractures; Lig. 1, lower left)
describes those traumatic injuries that have a low initial
frequency, then rise to reach peak incidence, and subse-
quently decline to a low frequency once again. Linally, a
Composite fracture pattern (e.g., distal radius fractures;
Lig. 1, lower right) refers to those traumatic injuries that
display two, or more, of the age related patterns
described thus far. The different types of fracture pat-
terns observed may also be characterized by sex differ-
ences in age at onset and peak frequency of occurrence.
The two J-Type traumatic injuries of old age that have
been examined most thoroughly in contemporary epi-
demiological studies are distal radius and proximal
femur fractures. The former primarily are represented by
fractures of the Colies, Smith, and Barton types (Bacorn
and Kurtzke 1953; Older et al. 1965; Koefed 1983).
Proximal femur fractures include compression fractures
of the femoral head, subcapitular and basilar fractures of
the femoral neck, and intertrochanteric fractures (Buhr
and Cooke 1959; Bauer 1960; Alffram and Bauer 1962;
Knowelden et al. 1964; Little 1973; Lreeman et al. 1974;
Johnell et al. 1984; Bengner and Johnell 1985; Solgaard
and Petersen 1985).
The age and sex specific patterns that characterize
distal radius and proximal femur fractures are illustrated
in Fig. 2. The shown fracture patterns represent trau-
matic injuries that occurred in White males and females
from England and Wales (Buhr and Cooks 1959), and
Age in Years
Fig. 2. Incidence of distal forearm and proximal femur fractures in a
contemporary population from England and Wales. Annual fracture
rates are represented as number of cases per million. The left vertical
margin shows the frequency of distal forearm fractures. The right
vertical margin shows the frequency of proximal femur fractures.
Adapted from Buhr and Cooke (1959).
they are by no means unique. Although annual incidence
per 10,000 may vary from one geographic region to
another, identical patterns for these two types of trau-
matic injury have been observed repeatedly (Bauer 1960;
Johnell et al. 1984; Gordon 1984; Solgaard and Petersen
1985).
The fundamental age and sex related differences,
which discriminate between distal radius and proximal
femur fracture patterns, can be described as follows. Dis-
tal radius fractures in females show an early age at onset
where a high incidence is established in the menopausal
years. Thereafter, distal radius fractures in females attain
what might be described as a plateau phase where a high
frequency is maintained, but does not appear to increase.
In contrast, males over 30 years of age have a low
frequency of distal radius fractures throughout life and
do not exhibit an old age J-Type pattern (Buhr and
Cooks 1959; Bauer 1960).
Proximal femur fracture rates illustrated in Pig. 2 show
that these traumatic injuries are uncommon in males and
females under 60 years of age. Thereafter, both sexes
exhibit a typical J-Type age progressive increase in the
frequency of hip fractures. However, females display
fracture rates that are at least two times greater than
those of males. Moreover, accelerated fracture rates in
males begin to rise approximately ten years later than in
females.
With regard to etiology, a number of exogenous and
endogenous factors are known to promote fracture risk
in modern industrial communities (Buhr and Cooke
1959; Alffram and Bauer 1962). The former represent a
group of nonbiological environmental circumstances,
such as occupational hazards and automobile accidents.
Geographic location and seasonality also have been
implicated in differential fracture incidence in human
groups (Gordon 1984; Solgaard and Petersen 1985; Beng-
ner and Johnell 1985).
However, among old people it is well recognized that
endogenous factors play a dominant role in elevated risk,
and increased severity, of traumatic injury (Bauer 1960).
These include disease conditions that often result in
pathological fractures (e.g., bone cysts, bone cancers,
Paget’s disease, etc.) and degenerative consequences of
biological aging (Buhr and Cooke 1959). The latter are
far more important in fracture risk and include degenera-
tive sequelae, such as failing eyesight, impaired motor
coordination, muscular weakness and atrophy, and pro-
nounced bone loss with age. Thus, as biological age
progresses in men and women, the frequency of acciden-
tal injuries continues to rise (Alffram and Bauer, 1962;
Knowelden et al 1964).
Indeed, Solgaard and Petersen (1985) report that the
majority of distal radius fractures that occur in women
(87%) and men (64 %) are the result of falls from level
ground. Studies also have shown that hip fractures in
individuals over 70 years of age are most often due to falls.
6
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
or movements we would ordinarily regard as trivial (e.g.,
arising from a seated position; Jaffe 1972; Johnell 1984).
Thus, a greater frequency of relatively minor accidents in
old people results in a greater incidence of moderate to
severe traumatic injuries.
A significant body of clinical and epidemiological
research has established a strong positive association
between age and sex specific fracture patterns and bone
loss with age. The results of such studies are summarized
elsewhere, and they will not be reviewed in detail here (see
Alffram and Bauer 1962; Chalmers 1973; Goldsmith et al.
1973; Wasnich et al. 1985; and references therein). It will
suffice to state that a slow normal rate of bone loss begins
in males and females during the 4th decade of life
(Goldsmith et al. 1973). Elowever, during the menopausal
years women experience an accelerated rate of bone loss
due to the cessation of ovarian function and estrogen
withdrawal (Meema and Meema 1976; Horsman et al.
1977). Therefore, by age 60 adult females often show a
significant reduction in skeletal mass, marked osteoporo-
sis, and subsequently display an earlier onset age for
J-Type fractures compared to males (Chalmers and
Weaver 1966; Chalmers 1973).
Racial differences in skeletal metabolism have also
been demonstrated. Studies show that Black men and
women attain higher peak bone mineral density values
compared to Whites (Trotter et al. 1960; Goldsmith et
al. 1973). Although Blacks lose bone at rates that are
comparable to Whites, they enter the older decades of life
with a relatively, and absolutely, greater bone mineral
content (Goldsmith et al. 1973). Thus, it is not remark-
able to find that the incidence of osteoporosis, and
old-age fractures, is significantly lower in older Black
men and women compared to Whites, and other racial
groups as well (Gordon 1984).
Given the epidemiological relationships described thus
far, the purpose of the present investigation is to examine
the frequency with which distal radius and proximal
femur fractures occurred in a sample of skeletons that
were assembled in Cleveland, Ohio, during the earlier
part of the twentieth century. The study is therefore
retrospective and replicative in nature. Here, it is specif-
ically hypothesized that:
(1) White females will display a greater frequency of
distal radius fractures compared to all other
sex/ race subgroups,
(2) age at onset for increased incidence of distal
radius fractures in White females will occur
during the perimenopausal and menopausal
years (i.e. , the period between 40 and 60
years),
(3) White females will display a greater frequency of
hip fractures compared to all other sex/ race
subgroups,
(4) age at onset for increased incidence of hip
fractures in White females will be 60 years of
age,
(5) age at onset for increased incidence of hip
fractures in White males will be 70 years of age,
or approximately 10 years later than in
females,
(6) a low frequency of hip fractures should charac-
terize all sex/ race subgroups under 60 years of
age,
(7) Black males and females will show a lower
frequency of distal radius and proximal femur
fractures at all ages compared to Whites,
(8) Black males and females will show no significant
increase in the frequency of distal radius and
proximal femur fractures with age, and
(9) Black males will show the lowest frequency of
distal radius and proximal femur fractures com-
pared to all other sex/ race subgroups.
Although age at death, sex, and race are known for the
individuals used in this study, a potential problem
concerns the fact that adequate medical histories are
unavailable for the skeletons examined. This means that
the age at which fractures occurred is only known for
those traumatic injuries appearing unremodeled at time
of death. Therefore, the fracture frequency data to be dis-
cussed are best regarded as cumulative in nature. None-
theless, sex and race differences in age at onset for in-
creased incidence of distal radius and proximal femur
fractures should remain unaffected and easy to detect.
Therefore, the J-Type fracture patterns that occurred in
an early twentieth-century urban industrial American
community were investigated in the following manner.
Materials
The human skeletal remains employed in this study are
from the Hamann-Todd Collection. The bulk of this
historic collection was assembled between the years 1910
and 1938, by Western Reserve University anatomists.
Under the direction of T. Wingate Todd, the skeletons of
over 3,400 dissecting room cadavers were retrieved for
future scientific research. These skeletal materials are
now permanently curated by the Cleveland Museum of
Natural History. It is important to realize that many of
the individuals included in the Hamann-Todd Collection
were transients, indigents, and persons of low socioeco-
nomic status. Thus, the sample used in this study is best
regarded as a biased cross section of early twentieth-
century urban industrial America.
At present, the Hamann-Todd Collection contains
records that document age and sex for the skeletal
remains of 3,157 Black and White Americans. These
specimens are partitioned by sex and race in Table 1. It
can be seen that sex ratios are markedly skewed in favor
of males. The male:female sex ratios for Blacks and
1987
HAMANN-TODD SKELETAL COLLECTION
7
TABLE 1
Black and White Adults in the
Hamann-Todd Cadaver Records
Race
Male
n\
Female
m
Sex
Ratio
Blacks
901
279
3.23
Whites
1725
252
6.85
Total3
2626
531
4.95b
“Among the 3,157 Hamann-Todd Black and White adults (+18
years) listed in cadaver records, a total of 1, 180 (37.4%) Blacks
and a total of 1,977 (62.6%) Whites are represented.
bLor both races combined males outnumber females by a factor
of five.
Whites are 3.23 and 6.85, respectively. In addition, 88
adults and an unspecified number of subadults that are
listed in cadaver records were returned to relatives for
burial, or were cremated, following autopsy. Given these
circumstances, the following sampling strategy was
employed.
At the time of observation a total of 262 Black and 207
White females were available for study. An equivalent
number of Black and White males were then age-matched
to their respective female samples. This was done to
establish balanced adult age distributions and sex ratios.
Thus, a total of 938 adult skeletons (Black n = 524; White
n= 414) were examined for distal radius and proximal
femur fractures.
Methods
All Hamann-Todd specimens included in the study
were macroscopically examined for the presence/
absence of distal radius and proximal femur fractures.
Distal radius fractures are defined as those fractures that
occurred within 4 cm. of the distal articular surface of the
radius. Proximal femur fractures were divided into two
classes. The first is intracapsular fractures of the proxi-
mal femur (PIC), and the second is extracapsular frac-
tures of the proximal femur (PEC). The former are
represented by compression fractures of the femoral head
and subcapitular fractures of the femoral neck. The latter
include basilar fractures of the femoral neck and intertro-
chanteric fractures of the proximal femur. No fractures
that were situated below the intertrochanteric line were
included in the study.
Four skeletal sites were examined for each specimen.
These were the left and right proximal femora and distal
radii, respectively. Adult males who were missing one or
more of the sampling sites were replaced by complete
age-matched specimens. Only three adult females were
missing one observation per individual (i.e., .08% of all
observations). For these females the missing skeletal site
was assumed not to have been fractured.
The identified fractures were then scored with re-
spect to symmetry. Unilateral left, unilateral right, and
bilateral involvements were recorded. Fractures were
also qualitatively assessed as remodeled, or unremod-
eled, at time of death. This was done in order to estimate
the risk of mortality associated with each fracture type.
All fractures that were identified in our initial survey
were then re-examined on a second occasion. This proce-
dure ensured that bony changes due to degenerative joint
disease, and other pathological conditions, would not
mistakenly be diagnosed as traumatic injuries (i.e., patho-
logical fractures were not included in the analysis). Like-
wise, all bones that showed evidence of post-mortem
damage were carefully gleaned so that all fracture ‘mim-
ics’ were excluded from consideration. Thus, the fracture
data to be summarized in the study are reported with a
high degree of confidence.
The investigation results are presented in graphic and
tabular form. Where appropriate, nonparametric proba-
bility statistics were used to evaluate differences in
fracture symmetry, remodeling status, and age, sex, and
race differences in the fracture patterns observed.
Results
Demographic Composition of the Hamann- Todd Sam-
ples
The age distributions for Blacks and Whites examined
in the study are listed by five-year age intervals in Table 2.
For comparison, more recent aggregate census data for
the United States in 1 960, is given in Table 3 (Thomlinson
1965). Survivorship curves for these data are illustrated
in Fig. 3. It can be seen that the Hamann-Todd Black and
White age distributions differ markedly from the modern
age profile. The greater survivorship, which characterizes
Fig. 3. Survivorship curves for the Hamann-Todd Black and White
samples, compared to survivorship reported for more recent aggregate
census data for the United States in 1960 (see Thomlinson 1965).
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
TABLE 2
Age Distributions for the Black and White
Male and Female Adult Samples Used in
the Fracture Study
Age
n\
White
Sample
%
h
ni
Black
Sample
%
h
18-24
5
2.4
100.0
30
11.5
100.0
25-29
14
6.8
97.6
34
13.0
88.5
30-34
10
4.8
90.8
29
11.1
75.5
35-39
24
1 1.6
86.0
50
19.1
64.4
40-44
19
9.2
74.4
29
11.1
45.3
45-49
22
10.6
65.2
21
8.0
34.2
50-54
24
1 1.6
54.6
23
8.8
26.2
55-59
9
4.3
43.0
8
3.1
17.4
60-64
19
9.2
38.7
10
3.8
14.3
65-69
13
6.3
29.5
10
3.8
10.5
70-74
13
6.3
23.2
10
3.8
6.7
75-79
20
9.7
16.9
2
0.8
2.9
+80
15
7.2
7.2
6
2.3
2.1
Total2
207b
262°
aThe total number of Hamann-Todd specimens included in the study
equals 938.
bThe 207 adult white females available for analysis were age-matched
with 207 adult white males. The sample of adult whites examined thus
equals 414 individuals.
"The 262 adult black females available for analysis were age-matched
with 262 adult black males. The sample of adult blacks examined thus
equals 524 individuals.
TABLE 3
Age Distribution for Modern Adults
Age3
N
%
lx
20
558
0.58
100.0
25
611
0.64
99.4
30
732
0.76
98.8
35
1 124
1.17
98.0
40
1719
1.79
96.9
45
2595
2.70
95.1
50
3987
4.15
92.4
55
5616
5.84
88.2
60
7911
8.23
82.4
65
11132
11.58
74.1
70
13851
14.40
62.6
75
14173
14.74
48.2
80
14383
14.96
33.4
+85
17767
18.48
18.5
Total
96159
aThese are aggregate United States census data for 1960 as reported in
Thomlinson 1965.
the latter group, is undeniably a consequence of techno-
logical advances in medicine and improvements in health
care delivery systems that have occurred over the past 60
years.
Marked differences also characterize the Hamann-
Todd Black and White age distributions. Mean ages for
the Black and White samples used in the study are 41.9
and 53.8 years, respectively. A Kolmogorov-Smirnov test
indicates that the two age distributions are significantly
different (x2 = 38.90; p<.001). The lower mean age at
death among Blacks may reflect socioeconomic and racial
factors that compromised access to health care during the
late nineteenth and early twentieth centuries. Here, the
circumstance that only a small proportion of the Blacks
sampled are over 60 years of age is thus regarded as a
potential source of error in our attempt to accurately es-
timate the incidence of old age fractures in this group.
General Observations
Several of the distal radius and proximal femur
fractures, which were identified in the Hamann-Todd
series, are illustrated in Fig. 4. In general, degree of
severity varied widely for the distal radius fractures that
were observed. Traumatic involvement ranged from
relatively benign hairline fractures of the subchondral
bony joint surfaces, to those where severe trauma
resulted in marked displacements and angular distortion
of the distal radius. Proximal femur fractures were much
more obvious. Intracapsular fractures of the proximal
femur (PIC) typically displayed a marked shortening of
functional femur neck length, combined with distortion
of normal spherical contours of the femoral head.
Extracapsular proximal femur fractures (PEC) were
frequently situated at the base of the femoral neck, and
they were often accompanied by the bony fragments of
associated intertrochanteric fractures.
Among the 938 skeletons that were examined, a total of
77 (8.2%) individuals displayed one or more distal radius
fracture. Only 34 (3.6%) specimens exhibited proximal
femur fractures. Thus, distal radius fractures were 2.3
times more frequent than hip fractures. A chi square
comparison indicates that this difference is highly signifi-
cant (x2 = 17.71; p<.001). For proximal femur fractures
alone, a total of 22 (2.4%) individuals had extracapsular
hip fractures, while only 12 (1.3%) specimens displayed
intracapsular hip trauma. The difference in frequency for
these two classes of hip fracture is not significant (x2 =
3.00; p>.05). All fractured individuals are partitioned by
sex and race in Appendix A. Information summarized
there identifies each specimen’s catalogue number, age at
death, as well as the specific type, symmetry, and remodel-
ing status for those fractures that were observed (Appen-
dix A Table 21).
Fig. 4. Examples of proximal femur and distal radius fractures which were observed in the
Hamann-Todd Collection. A, left proximal femur which manifests a basilar neck fracture combined
with an intertrochanteric fracture (CMNH-HT-3032, 89-year-old White female); B, Subcapital
fracture of the right femoral neck which subsequently led to ischemic necrosis of the femoral head
(CMNH-HT-1024, 84-year-old White female); C, anterior view of a Colie’s fracture affecting the
right distal radius (CMNH-HT-14Q2, 81-year-old White female); D , posterior view of Colie’s
fracture affecting the left radius; and E posterior view of Colie’s fracture affecting the right radius of
the same individual (CMNH-HT-1581, 83-year-old White female).
10
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
Fracture Symmetry
The frequency with which distal radius and proximal
femur fractures were seen on the left, right, or both sides
of affected individuals is summarized in Table 4. These
data are partitioned by sex and race in Appendix B
(Table 22). In general, 8-9% of all fractures were seen to
TABLE 4
Distal Radius and
Proximal Femur Fracture Symmetry
Total
No. of
Individuals
with Left Right Bilateral
Fracture Type Fractures m % m_ % m %
Distal Radius 77 36 46.8 32 41.6 7 9.1
Proximal Femur 34 16 47.1 15 44.1 3 8.8
occur bilaterally. Nonetheless, it was not possible to
determine if bilateral involvements were the result of the
same traumatic episode. For several individuals this did
not seem to be probable. Patterns for unilateral involve-
ment indicate no side preference for distal radius or
proximal femur fractures. A chi square comparison
confirms that no significant difference in unilateral
expression characterizes the fractures which were identi-
fied in the study (x2 = 0.02; p>.80).
Fracture Remodeling Status
The frequencies with which distal radius and proximal
femur fractures appeared remodeled, or unremodeled, at
time of death is given in Table 5. These data are parti-
tioned by sex and race in Appendix C (Table 23). It was
found that 88% of distal radius fractures, and 87% of in-
tracapsular hip fractures, were healed at time of death.
Though varying degrees of functional immobility un-
doubtedly accompanied many of these injuries, it is con-
cluded that distal radius and intracapsular proximal
TABLE 5
Distal Radius and Proximal Femur
Fracture Remodeling Status
Total
No. of
Fractures Remodeled Unremodeled
Fracture Type
Observed
m
%
rii
%
Distal Radius
84
74
88.1
10
1 1.9
Femur (PIC)
15
13
86.7
2
13.3
Femur (PEC)
22
5
22.7
17
77.3
femur fractures were well tolerated from a clinical
perspective.
In contrast, only 23% of extracapsular hip fractures
were remodeled at time of death. Chi square values listed
in Table 6 show that the marked remodeling differential
between extracapsular hip fractures, and those described
above, is statistically significant. It is concluded that indi-
viduals who acquired extracapsular fractures of the prox-
imal femur had a very high risk of mortality. That is,
approximately eight out of every ten Hamann-Todd spec-
imens with femur neck fractures probably died as a direct,
or indirect, consequence of the traumatic injury.
TABLE 6
Chi Square Values for Comparisons of Distal Radius and
Proximal Femur Fracture Remodeling Status
Comparison
Chi Square
Distal Radius
vs.
39.24a
Femur (PEC)
Distal Radius
vs.
0.26
Femur (PIC)
Femur (PIC)
vs
13.14“
Femur (PEC)
“Significant at .001 level of probability.
Old Age Fracture Patterns in the Hamann- Todd
Collection
The frequency with which distal radius and proximal
femur fractures were observed for Hamann-Todd Black
and White males and females is given by decade in Tables
7, 8, 9, and 10. These data are listed by five-year age inter-
vals in Appendix D; Tables 24, 25, 26, and 27. The Black
and White age and sex specific fracture patterns are illu-
strated in Figure 5. The total frequency with which frac-
tures occurred is partitioned by sex and race for each frac-
ture type in Table 11. Data summarized there are
expressed as the total number of individuals affected.
Results show that White females experienced a greater
frequency of distal radius and proximal femur fractures
compared to all other sex/ race subgroups. In descending
frequency, the rank order for distal radius fractures is
White female (19.3%), White male (7.2%), Black female
(6.5%), and Black male (1.9%). The rank order for
proximal femur fractures is White female (9.7%), White
male (3.4%), Black male ( 1.5%), and Black female ( 1 .2%).
Chi square values (Table 1 2) indicate that both Black and
White women had a significantly greater frequency of
distal radius fractures compared to those of the proximal
femur. Similar comparisons for Black and White men
were insignificant.
1987
HAMANN-TODD SKELETAL COLLECTION
TABLE 7
Age Specific Lracture Lrequencies for White
Lemales Listed by Decade
TABLE 10
Age Specific Lracture Lrequencies for Black
Males Listed by Decade
Total
Distal Femur Femur Proximal
Radius (PIC) (PEC) Femur
Age
N
ri\
%
fl2
%
m
%
m
%
18-29
19
0
0.0
0
0.0
0
0.0
0
0.0
30-39
34
1
2.9
0
0.0
0
0.0
0
0.0
40-49
41
6
14.6
0
0.0
1
2.4
1
2.4
50-59
33
10
30.0
1
3.0
1
3.0
2
6.1
60-69
32
6
18.8
0
0.0
5
15.6
5
15.6
70-79
33
9
27.3
0
0.0
4
12.1
4
12.1
+80
15
8
53.3
3
20.0
5
33.3
8
53.3
Total
207
40
19.3
4
1.9
16
7.7
20
9.7
TABLE 8
Age Specific Lracture Lrequencies for White
Males Listed by Decade
Age
N
Distal
Radius
rt\ %
Femur
(PIC)
ni %
Femur
(PEC)
W %
Total
Proximal
Femur
tlA %
18-29
19
0
0.0
0
0.0
0
0.0
0
0.0
30-39
34
3
0°
oo
0
0.0
0
0.0
0
0.0
40-49
41
1
2.4
0
0.0
0
0.0
0
0.0
50-59
33
1
3.0
0
0.0
0
0.0
0
0.0
60-69
32
3
9.4
1
3.1
1
3.1
2
6.3
70-79
33
5
15.2
0
0.0
1
3.0
1
3.0
+80
15
2
13.3
1
6.7
3
20.0
4
26.7
Total
207
15
7.2
2
1.0
5
2.4
7
3.4
TABLE 9
Age Specific Lracture Lrequencies for Black
Lemales Listed by Decade
Age
N
Distal
Radius
rt\ %
Femur
(PIC)
ni %
Femur
(PEC)
m %
Total
Proximal
Femur
m %
18-29
64
0
0.0
0
0.0
0
0.0
0
0.0
30-39
79
4
5.1
0
0.0
0
0.0
0
0.0
40-49
50
4
8.0
0
0.0
0
0.0
0
0.0
50-59
31
2
6.5
0
0.0
0
0.0
0
0.0
60-69
20
6
30.0
1
5.0
1
5.0
2
10.0
70-79
12
0
8.3
1
8.3
0
0.0
1
8.3
+80
6
1
16.7
0
0.0
0
0.0
0
0.0
Total
262
17
6.5
2
0.8
1
0.4
3
1.1
Total
Distal Femur Femur Proximal
Radius (PIC) (PEC) Femur
Age
N
n\
%
rt2
%
%
ri\
%
18-29
64
\
1.6
1
1.6
0
0.0
1
1.6
30-39
79
0
0.0
1
1.3
0
0.0
1
1.3
40-49
50
1
2.0
1
2.0
0
0.0
1
2.0
50-59
31
1
3.2
0
0.0
0
0.0
0
0.0
60-69
20
2
10.0
1
5.0
0
0.0
1
5.0
70-79
12
0
0.0
0
0.0
0
0.0
0
0.0
+80
6
0
0.0
0
0.0
0
0.0
0
0.0
Total
262
5
1.9
4
1.5
0
0.0
4
1.5
TABLE 1 I
Summary of Number and Lrequency of Fractures
Observed in the Hamann-Todd Sample
White
White
Black
Black
Female
Male
Female
Male
(N-207)
(N-207)
(N-262)
(N-262)
Fracture Type
n %
n %
n %
n %
Distal Radius
40
19.3
15
7.2
17
6.5
5
1.9
Femur (PIC)
4
1.9
2
1.0
2
0.8
4
1.5
Femur (PEC)
16
7.7
5
2.4
1
0.4
0
0.0
Total Proximal
Femur3
20
9.7
7
3.4
3
1.2
4
1.5
Total
53h
25.6
22
10.6
20
7.6
9
3.4
■“Figures listed for total proximal femur are simply the summed observa-
tions for the catergories of femoral ( PIC) and femoral (PEC) fractures.
hA total of 7 adult White female individuals displayed combined frac-
tures of the distal radius and proximal femur at time of death. Thus a
total of 53 White females (60-7) had one or more old-age fractures at
time of death. No White males and no Blacks displayed combined distal
radius and proximal femur fractures at time of death.
TABLE 12
Chi Square Values Comparing the Difference in Total
Frequency for Distal Radius versus Proximal Femur Fractures
Distal
Proximal
Radius
vs.
Femur
White Female
7.803
White Male
3.07
Black Female
9.253
Black Male
0.11
“Significant at the .01 level of probability.
Proximal Femur Fractures Distal Radius Fractures
H-T White
H-T Black
H-T White H-T Black
Fig. 5. Age, sex, and race specific frequency of distal radius and proximal femur fractures that were
observed in the Hamann-Todd sample.
1987
HAMANN-TODD SKELETAL COLLECTION
13
Chi square values given in Table 13 show that male/
female differences in the frequency of distal radius
fractures are statistically significant for comparisons
within each race. These tests also indicate that White
males and females had a significantly greater frequency
of distal radius fractures compared to Black males and
females, respectively. Therefore marked sex and race
differences characterize the total frequency of distal
radius fractures that were encountered in the study. With
TABLE 13
Chi Square Values Comparing Sex and Race
Differences in the Total Frequency of Distal Radius Fractures
White
Female
White
Male
Black
Female
Black
Male
White Female
—
1 3. !0a
17.84a
40.43“
White Male
—
0.01
8.07
Black Female
—
6.83b
Black Male
—
“Significant at the .01 level of probability.
bSignificant at the .001 level of probability.
regard to proximal femur fractures, the chi square values
listed in Table 14 show only that White women had a
significantly higher frequency of hip fractures compared
to all other sex/race subgroups. Given the results pre-
sented thus far, the age, sex, and race specific fracture
patterns that were found can be described as follows.
Distal Radius Fractures in Hamann- Todd Whites
White males and females displayed dramatic differ-
ences in distal radius fracture patterns (see Fig. 5; Tables
7 and 8). In general, fractures were uncommon in White
women under 40 years of age. Thereafter, females showed
a marked rise in the frequency of fractures. An early peak
occurred in the sixth decade. Slightly lower frequencies
were seen in the seventh and eighth decades. Nonetheless,
TABLE 14
Chi Square Values Comparing Sex and Race
Differences in the Total Frequency of Proximal Femur
Fractures
White
Female
White
Male
Black
Female
Black
Mate
White Female
—
6.70“
17.99b
1 5. 76b
White Male
—
2.77
1.74
Black Female
—
0.14
Black Male
—
“Significant at the .01 level of probability.
bSignificant at the .00! level of probability.
the frequency of distal radius fractures remained high in
females over 40 years of age, and reached peak incidence
in the ninth decade.
In contrast. White males had a low frequency of distal
radius fractures compared to females overall. Men
exhibited a minor A-Type fracture pattern during the 20
to 50 year period, and a diminutive J-Type pattern over
60 years of age. The age at onset for increased incidence
of distal radius fractures in older men was during the
seventh decade. The peak frequency for distal radius
fractures in White males occurred in the eighth and ninth
decades.
Thus, White men and women showed dramatic differ-
ences in age at onset, and frequency of occurrence, for
distal radius fractures. These findings can be summarized
as follows. Females displayed an age at onset for
increased incidence of distal radius fractures that was 20
years earlier than age at onset for males. Also, women
achieved a fracture incidence in the fifth decade that was
not attained in males until the eight decade. Finally,
females displayed a ninth decade distal radius fracture
incidence that was four times greater than that seen in
comparable age-matched males.
Distal Radius Fractures in Hamann- Todd Blacks
In contrast to Whites, the Blacks examined in the study
showed lower frequencies of distal radius fractures, and
did not display a typical J-Type pattern (see Fig. 5; Tables
9 and 10). The latter finding may be due to the fact that
significantly fewer Blacks survived to the later decades of
life. However, age and sex specific fracture patterns were
seen among the Blacks who were examined. Although
Black women under 60 years of age exhibited a slightly
higher frequency of distal radius fractures compared to
men, the overall incidence of fractures was low for both
sexes during the early and middle years of adulthood.
Age at onset for increased frequency of distal radius
fractures, as well as age at peak incidence, occurred in the
seventh decade for both sexes. The only noteworthy
difference between males and females was confined to
individuals over 60 years of age. At this time Black
women displayed a frequency of distal radius fractures
that was 3.5 times greater than that seen in males.
Proximal Femur Fractures in Hamann- Todd Whites
White males and females displayed a typical J-Type
pattern for proximal femur fractures (see Fig. 5; Tables 7
and 8). In general, hip fractures were uncommon in men
and women that were under 50 years of age. Age at onset
for increased incidence of hip fractures was during the
sixth decade in females, and the seventh decade in males.
Thereafter, the frequency of hip fractures showed a
general trend of age progressive increase in both sexes.
White men and women in the ninth decade of life showed
the highest frequency of hip fractures. Sex differences in
14
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
hip fracture incidence were also observed. For all decades
over 50 years, White females exhibited frequencies of
proximal femur fractures that were at least two times
greater than those seen in males. Thus, hip fractures in the
Hamann-Todd White sample showed a 10-year sex dif-
ference in age at onset, and sex differences in overall fre-
quency of occurrence.
Proximal Femur Fractures in Hamann-Todd Blacks
In contrast to Whites, the Blacks showed a low overall
frequency of hip fractures, and did not display a charac-
teristic J-Type pattern (see Fig. 5; Tables 9 and 10). The
latter finding is presumed to be, in part, a consequence of
poor sampling in the later decades of life, as was the case
with distal radius fracture patterns. The age and sex
specific patterns for hip fractures that were observed
among Blacks are as follows. Hip fractures were uncom-
mon in males and females under 60 years of age. Age at
onset for increased incidence of proximal femur fractures
was in the seventh decade for men and women. With
regard to sex differences. Black females over 60 years of
age displayed a frequency of hip fracture that was three
times greater than that seen in males. Although age at
onset appears to be similar, a sex difference in the
frequency of hip fractures among older Blacks is thus
suggested.
Evaluation of Age Specific Fracture Patterns
In order to assess the statistical significance of age
related differences in the frequency of old-age fractures
that were observed, the following procedure was used.
Fracture frequency data were compressed into three age
groups, which identify early adulthood (18-39 years),
middle adulthood (40-59 years), and late adulthood (+60
years). These age intervals also correspond to the pre-
menopausal, climacteric, and postmenopausal years in
adult women, respectively. Fracture data that were
summarized in this manner are listed in Tables 1 5 and 1 6,
and illustrated in Fig. 6.
Chi square values for sex specific comparisons of age
related differences in the frequency of distal radius
fractures are given in Table 17. Results show that (a)
White women over 40 years of age had a significantly
greater frequency of fractures compared to those women
under 40 years of age, (b) White men over 60 years of age
had a significantly greater frequency of fractures com-
pared to men that were 40 to 59 years of age, and (c) both
Black men and women over 60 years of age had a
significantly greater frequency of fractures compared to
those individuals that were 18 to 39 years of age. Thus,
significant age-related differences in the frequency of
distal radius fractures characterize all sex/race sub-
groups that were examined in the study.
Chi square values listed in Table 18 compare male/
TABLE 15
White Female and Male Fracture Data Compressed
into Age Intervals
Age
N
Distal
Radius
n\ %
Femur
(PIC)
n2 %
Femur
(PEC)
m %
Total
Proximal
Femur
m %
White
Female
18-39
53
1
1.9
0
0.0
0
0.0
0
0.0
40-59
74
16
21.6
1
1.4
2
2.7
3
4.1
+60
80
23
28.8
3
3.8
14
17.5
17
21.3
Total
207
40
19.3
4
1.9
16
7.7
20
9.7
White
Male
18-39
53
3
5.7
0
0.0
0
0.0
0
0.0
40-59
74
2
2.7
0
0.0
0
0.0
0
0.0
+60
80
10
12.5
2
2.5
5
6.3
7
OO
oo
Total
207
15
7.2
2
1.0
5
2.4
7
3.4
TABLE 16
Black Female and Male Fracture Data Compressed
into Age Intervals
Age
N
Distal
Radius
m %
Femur
(PIC)
m %
Femur
(PEC)
m %
Total
Proximal
Femur
«4 %
Black
Female
18-39
143
4
2.8
0
0.0
0
0.0
0
0.0
40-59
81
6
7.4
0
0.0
0
0.0
0
0.0
+60
38
7
18.4
2
5.3
1
2.6
3
7.9
Total
262
17
6.5
2
0.8
1
0.4
3
1.1
Black
Male
18-39
143
1
0.7
2
1.4
0
0.0
2
1.4
40-59
81
2
2.5
1
1.2
0
0.0
1
1.2
+60
38
2
5.3
1
2.6
0
0.0
1
2.6
Total
262
5
1.9
4
1.5
0
0.0
4
1.5
female age related differences in the frequency of distal
radius fractures that were observed within each race.
Results indicate that (a) White females over 40 years of
age had a significantly greater frequency of fractures
compared to White males, and (b) age related differences
in the frequency of distal radius fractures among Black
males and females were statistically insignificant.
1987
HAMANN-TODD SKELETAL COLLECTION
15
A
30-
25'
ABC ABC ABC ABC
White White Black Black
Female Male Female Male
B
White White Black Black
Female Male Female Male
TABLE 17
Chi Square Values for Distal Radius Fracture
Age Group Comparisons Partitioned by Sex and Race
Age Group
White White
Black Black
Comparison
Female Male
Female Male
18-39 vs. 40-59
9. 1 8a 0.71
2.58 1.22
18-39 vs. +60
14.25b 1.69
1 2.84h 4.25“
40-59 vs. +60
1.03 5.14“
3.22 0.78
'“Significant at the
.01 level of probability.
hSignificant at the
.001 level of probability.
“Significant at the
.05 level of probability.
TABLE 18
Chi Square Values for Distal Radius Fracture
Age Group Comparisons between the Sexes
and Partitioned by Race
White
Black
Age
Male vs. Female
Mate vs. Female
18-39
0.84
1.53
40-59
1 l.26a
1.52
+60
6.45h
2.44
“Significant at the
.001 level of probability.
hSignificant at the
.05 level of probability.
TABLE 19
Chi Square Values for Distal Radius Fracture
Age Group Comparisons between the Races
and Partitioned by Sex
Male
Female
Age
Black vs. White
Black vs. White
18-39
4.80a
0.01
40-59
0.01
6.42a
+60
0.95
1.45
“Significant at the .05 level of probability.
Fig. 6. Frequency of distal radius and proximal femur fractures
partitioned by sex and race for the age groups A ( 1 8-39), B (40-59), and
C (+60 years).
Linally, chi square values given in Table 19 compare
male/ male and female/female age related differences in
the frequency of distal radius fractures between each
race. Results show that (a) White females over 40 years of
age had a significantly higher frequency of fractures
compared to Black females, and (b) young adult White
males had a significantly higher frequency of fractures
compared to young adult Black males. No other race
differences in the age specific frequency of distal radius
fractures were statistically significant.
Statistical analyses similar to those presented above
were not conducted for proximal femur fractures. This is
because hip fractures were infrequent overall, and they
were particularly rare in those individuals that were under
60 years of age at time of death. Nonetheless, the age re-
lated patterns in Lig. 5 clearly demonstrate that hip frac-
tures were of no epidemiological significance in either
Blacks or Whites prior to the onset of old age.
Discussion
With regard to the replicative goals of this study, a
number of statements can be made concerning the age.
16
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
sex, and race specific fracture patterns that were pre-
dicted earlier. Hypotheses 1 , 2, 3, 6, and 7 were confirmed
as stated. Hypotheses 4, 5, and 9 were confirmed with
minor qualifications. Only hypothesis number 8 was
found to be incorrect. For each hypothesis given, the
fracture patterns that were observed can be summarized
as follows.
(1) White females displayed a significantly greater
frequency of distal radius fractures compared to
all other sex/ race subgroups.
(2) Age at onset for increased incidence of distal
radius fractures in White women occurred in the
40 to 60 age group.
(3) White females displayed a significantly greater
frequency of hip fractures compared to all other
sex/ race subgroups.
(4) Age at onset for increased incidence of hip
fractures in White women occurred during the
sixth decade of life (i.e., 10 years earlier than
predicted.
(5) Age at onset for increased incidence of hip
fractures in White men occurred during the
seventh decade (i.e., 10 years earlier than pre-
dicted. Age at onset in males did occur approxi-
mately 10 years later than in females.
(6) A low frequency of hip fractures characterized
all sex/ race subgroups that were under 60 years
of age at time of death.
(7) Black men and women showed a lower fre-
quency of distal radius and proximal femur
fractures at all ages compared to White men and
women, respectively. The only exception to this
pattern was that Black females had a slightly
higher incidence of distal radius fractures in the
fourth and seventh decades compared to White
females. However, these minor differences were
insignificant, and not apparent in smoothed
comparisons of age grouped data.
(8) Contrary to the pattern predicted earlier. Black
men and women that were over 60 years of age
did display a significantly greater frequency of
distal radius fractures compared to Black men
and women in the 18 to 39 age group.
(9) Black males displayed the lowest frequency of
distal radius and proximal femur fractures com-
pared to all other sex/ race subgroups. The only
exception to this was that Black men showed a
slightly greater, but insignificant, total fre-
quency of hip fractures compared to Black
women. However, data summarized earlier indi-
cate that several hip fractures in Black males
occurred in individuals that were under 60 years
of age at time of death. As predicted, Black
women over 60 years showed a higher age
specific frequency of hip fractures compared to
Black men.
Therefore, it is quite clear that the age, sex, and race
specific patterns that characterize distal radius and
proximal femur fractures among Hamann-Todd Blacks
and Whites strongly conform to fracture patterns that
have been reported for more recent American and
European urban industrial populations. However, one
interesting observation concerns the finding that
Hamann-Todd Black and White men both displayed an
increased frequency of distal radius fractures in individu-
als over 60 years of age. Fracture epidemiology studies,
which were conducted during the mid 1950’s, report that
older males showed no increased incidence of distal
radius fractures with age (Buhr and Cooke 1959; Bauer
1960). Nonetheless, more recent surveys document a
secular trend whereby men, and women, display marked
increases in the incidence of both distal radius and hip
fractures in individuals over 60 years of age (Nilsson and
Obrant 1978; Zetterberg and Andersson 1982; Frandsen
and Kruse 1983; Swanson and Murdoch 1983; Wallace
1983; Zain et al. 1984; Bengner and Johnell 1985;
Solgaard and Petersen 1985). Thus, the increased fre-
quency of distal radius fractures that was seen in older
Hamann-Todd Collection Black and White men should
not be regarded as an unexpected, or unusual, finding.
The most striking difference found in this investigation
concerns the remarkable frequency with which distal
radius and proximal femur fractures occurred in
Hamann-Todd individuals. In order to emphasize this
point, the Hamann-Todd fracture data were converted to
total incidence per 10,000. This information is summa-
rized in Table 20. Also given is the total incidence per
10,000 for distal radius and hip fractures as computed for
data reported for the population in Malmo, Sweden in
1955 (Bauer 1960).
Fracture ratios for the distal radius indicate that
TABLE 20
Total Incidence of Old Age Fractures
Distal Radius Proximal Femur
Fractures Fractures
Subgroup
H-T M-Sb
Ratio
H-T
M-S
Ratio
White Female
1932
116
16.7
966
93
10.4
White Male
725
22
33.0
338
30
11.3
Black Female
649
116
5.6
115
93
1.2
Black Male
191
22
8.7
153
30
5.1
“Total incidence per 10,000 for the Hamann-Todd study sample of early
twentieth century American Blacks and Whites.
bTotal incidence per 10,000 for distal radius and proximal femur frac-
tures reported for European White males and females in Malmo,
Sweden in 1955 (Bauer 1960).
1987
HAMANN-TODD SKELETAL COLLECTION
17
Hamann-Todd total fracture incidence is (a) 17 times
greater for White women, (b) 33 times greater for White
men, (c) 6 times greater for Black women, and (d) 9 times
greater for Black men compared to the modern group.
Although less marked, fracture ratios for hip trauma
show that Hamann-Todd total fracture incidence is (a) 10
times greater for White females, (b) 1 1 times greater for
White males, (c) nearly identical for Black females and
(d) 5 times greater for Black males. Given the known
biased composition of the Hamann-Todd Collection,
and lacking any other fracture data from the earlier part
of the 20th century in urban America, it is suggested here
that Hamann-Todd individuals represented a cross sec-
tion of society at very high risk of traumatic injury.
The final issue to be addressed here concerns the
etiology of distal radius fractures in Caucasian women.
All fracture epidemiology studies that have been con-
ducted thusfar (i.e. , those cited herein) are in agreement
with the following perspectives. First, the majority of
distal radius fractures that occur in all age/sex/race
groups are the simple consequence of accidental falls
from level ground. Second, the age at onset for a marked
increase in the incidence of distal radius fractures in
White females is strongly associated with the climateric
years of life. Third, studies have shown that bone loss
with age is most pronounced in Caucasian women
compared to all other age-matched sex and race groups.
Fourth, skeletal fragility is regarded as the primary risk
factor responsible for the early onset, and high incidence,
of distal radius fractures in Caucasian females.
Although senile and postmenopausal osteoporosis are
clearly implicated in the pathogenesis of hip fractures in
men and women that are over 60 years of age, we consider
it improbable that bone loss alone could account for the
early onset, high frequency, and marked differences that
have been observed for distal radius fracture patterns in
Caucasian women compared to men. Indeed, do peri-
menopausal and menopausal females ‘fall’ more fre-
quently than their premenopausal peers? If so, it would
be of interest to know whether or not such falls were
merely routine accidents, or whether such episodes were
initiated by physiological disturbances common to this
particular age and sex group.
Thus, it is suggested here that physiological conse-
quences of estrogen withdrawal other than bone loss may
be important factors, which play a role in promoting
accidents and elevating fracture risk during the climac-
teric years of life in women. The latter is defined as the
period encompassing the onset of menstrual irregularity,
overt menopause, and several years following the com-
plete cessation of menses. That is, the period from 45 to
60 years, with the median age of menopause occurring at
approximately 50 years (Frommer 1964; Dennerstein
and Burrows 1978).
It is important for fracture epidemiologists to realize
that over the last decade clinical researchers have devoted
greater attention to the psychological, behavioral, bio-
chemical, and physiological changes that occur in women
undergoing the menopausal transition (McKinley and
Jefferys 1974; Dennerstein and Burrows 1978; Casper et
al. 1979; Tataryn et al. 1980). The problems encountered
by these individuals include an increased incidence of
psychosomatic illness, vaginal atrophy osteoporosis and
vasomotor disturbances (Voda 1981). However, the
latter are by far the most frequent symptoms (e.g., hot
flushes, night sweats, etc), which generate discomfort and
anxiety. Indeed, studies show that 75% of women
undergoing natural menopause experience a variable
frequency, intensity, and duration of vasomotor distur-
bances during the climacteric period. (McKinley and
Jefferys 1974; Voda 1981; Felman et al 1985).
One of the most informative studies of the menopausal
syndrome in Caucasian women was conducted by Bun-
gay and colleagues ( 1 980). They found that (a) symptoms
of vasomotor disturbance were strongly associated with
the climacteric period, and (b) peak frequency of symp-
toms closely corresponded to the median age of meno-
pause. The vasomotor disturbances displaying this
marked age and sex specific pattern were classified as
Type 3a responses (see Bungay et al. 1980). The latter
include hot flushes, night sweats, day sweats, dizzy spells,
tiredness, palpitations, difficulty in concentration, diffi-
culty in making decisions, loss of confidence, forgetful-
ness, and feelings of unworthiness. Unfortunately, simi-
lar information about the menopausal experience in
Black women is unavailable.
The age specific pattern for the frequency of vasomo-
tor disturbances as seen in modern White climacteric
women (Bungay et al. 1980) is illustrated in Fig. 7. Also
shown is the age specific frequency with which distal
radius fractures were observed in Hamann-Todd White
females. The vasomotor disturbance data exhibits a
typical A-Type pattern. It can be seen that the age related
pattern for distal radius fractures in Hamann-Todd
Caucasian women parallels the frequency of vasomotor
disturbances for those individuals that ranged in age
from 40 to 65 years.
Although falls from level ground are recognized as the
major circumstance in which distal radius fractures
occur, it is perhaps worthy of comment that no studies
have ever reported the results of inquires as to why
patients fell in the first place. That is, we have no direct
knowledge of the extent to which dizzy spells or other
vasomotor disturbances may have preceeded, or ini-
tiated, accidental falls. Thus, the data illustrated in Fig. 7
prompt us to suggest that vasomotor disturbances may
play a significant role in initiating a greater frequency of
accidents and, subsequently, an early onset and elevated
incidence of distal radius fractures in climacteric White
females.
18
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
White Female
E
o
a
H E
3
O
a
o
c
a
o
CO
-
Fig. 7. Frequency of distal radius fractures in Hamann-Todd White
females, compared to the age specific frequency of Type 3a menopausal
symptoms reported for modern White females (see Bungay et al., 1980).
Type 3a responses are a class of symptoms due to vasomotor
disturbances that accompany estrogen withdrawal in women during the
climacteric years of life.
Based on the age and sex specific physiological rela-
tionships described thus far, it is further suggested that
distal radius fractures in adult White women may be
characterized by an overlapping A/ J-Composite fracture
profile. The A and J-Type components that correspond
to the hypothetical fracture pattern are illustrated in Fig.
8A. The additive distal radius fracture incidence which
would result from such a pattern is shown in Fig. 8B. It is
hypothesized that the A-Type distal radius fracture
component should bracket the climacteric period. Dur-
ing this time, vasomotor disturbances would be regarded
as the dominant, but not sole, factor in fracture etiology.
Therefore, the age at onset, peak incidence, and age at
decline for distal radius fractures should parallel chang-
ing frequencies of vasomotor disturbances.
The second, J-Type distal radius fracture component
would be regarded as a primary consequence of age
progressive bone loss and other degenerative sequelae
discussed earlier. Thus, as vasomotor disturbances de-
crease in frequency at the end of the climacteric period, it
would be expected that skeletal fragility will play a
dominant role in the increased incidence of distal radius
fractures, as is the circumstance with hip trauma among
older men and women alike.
The striking similarity between the hypothetical com-
posite White female distal radius fracture pattern, and
the pattern that was observed for Hamann-Todd White
females, has not escaped our attention. However, several
problems which limit our basis of inference deserve
comment here. The hypothetical composite fracture
pattern would receive support from our findings if the
A
White Female
B
White Female
Fig. 8. Hypothetical age specific distal radius fracture pattern for
Caucasian women. The proposed A-Type pattern during the climac-
teric years, and the J-Type pattern for postmenopausal females are
illustrated separately in Fig. 8A. The cumulative frequency of distal
radius fractures which would be observed for the A/ J-Composite
pattern is shown in Fig. 8B.
majority of distal radius fractures that were seen in
Hamann-Todd individuals had occurred with five years
of age at death. This circumstance is considered to be
unlikely. Most of these traumatic injuries were, to some
1987
HAMANN-TODD SKELETAL COLLECTION
19
degree, remodeled at time of death and true age at
occurrence remains unknown. Therefore, the age specific
frequencies that were reported for Hamann-Todd distal
radius fractures are best regarded as cumulative data.
Furthermore, the high 6th decade incidence of distal
radius fractures that was seen in Hamann-Todd White
females, relative to 7th and 8th decade values, can just as
readily be interpreted as a consequence of sampling
errors.
More recent epidemiological studies do not permit
evaluation of the hypothetical White female composite
distal radius fracture pattern for other reasons. The most
important among these is the effective clinical manage-
ment of bone loss. Over the past two decades, prophylac-
tic and therapeutic use of hormonal and nutritional
supplements has become increasingly widespread as a
means to slow bone loss in aging Caucasian women
(Gordon 1961; Nicholas and Wilson 1963; Davis et al.
1966; Lafferty et al. 1969; Riggs etal. 1976; Gallagher and
Riggs 1978). Numerous clinical studies document that
accelerated rates of bone loss can be retarded if estrogen
replacement therapy is begun within the first few years of
natural, or surgical, menopause (Meema et al. 1975;
Meema and Meema 1976; Aitken et al. 1976; Nordinet al.
1976; Horsman et al. 1977).
With regard to fracture patterns, it is important to
realize that widespread use of estrogen replacement
therapy in climacteric women also has the effect of
reducing the frequency, intensity, and duration of vaso-
motor disturbances that would otherwise occur in these
individuals (Bungay et al. 1980; Hammar et al. 1984).
Thus, it is reasonable to expect that the hypothesized
A-Type climacteric distal radius fracture component in
modern White females will be reduced in magnitude and
shifted upward in age. The A- and J-Type fracture
patterns would then show significant overlap. The result-
ant modern fracture profile might then give the appear-
ance of an early onset J-Type pattern that attains a
plateau phase, or simply an age progressive J-Type
pattern. The most recent distal radius fracture patterns,
which have been reported for Caucasian women, are in
accord with the expectations posited above (see Solgaard
and Petersen 1985; and Bengner and Johnell 1985).
The retrospective analysis presented here, and more
contemporary research described above, do not allow us
to assess the validity of the hypothetical A/ J-Composite
distal radius fracture pattern as posited for Caucasian
women. Nonetheless, it is interesting that Hamann-Todd
White females display an age specific distal radius frac-
ture pattern similar to the one we would predict to occur
in a population that did not experience the benefits of
estrogen replacement therapy and recommended nutri-
tional supplementation. Therefore, resolution of the vaso-
motor disturbance hypothesis, as well as information
about fracture patterns and other aspects of aging in
modem Black men and women, deserve attention in fu-
ture gerontological research.
Summary and Conclusion
The study presents a retrospective analysis of distal
radius and proximal femur fractures that occurred in 938
Hamann-Todd Collection skeletons. The Black and
White individuals included in the investigation were
retrieved from dissecting room cadavers in Cleveland,
Ohio, between the years 1910 and 1938. Thus, the sample
represents a biased cross section of an early twentieth cen-
tury American urban industrial society.
Demographic analysis demonstrated that the mean age
at death for Blacks, 41.9 years, was significantly lower
than the mean age at death for Whites, 53.8 years, for
those individuals examined. Therefore, the small propor-
tion of Blacks over 60 years of age that were represented
was considered a potential source of sampling error, with
respect to our ability to accurately estimate the frequency
of old age fractures in this group.
Fracture symmetry data indicate that 8-9% of all distal
radius and hip fractures occurred bilaterally. However, it
was not possible to determine if bilateral involvements
were the result of one or more traumatic injuries. It is
probable that several of these occurred at different times
during the life of affected individuals. Unilateral involve-
ments showed no side preference for any of the fracture
types that were examined.
With regard to fracture repair status, it was found that
88% of distal radius fractures, and 87% of mtracapsular
hip fractures, displayed moderate to marked bone re-
modeling at time of death. It was concluded that these
fractures were well tolerated from a clinical perspective.
In contrast, only 23% of extracapsular hip fractures were
remodeled at time of death. It was therefore concluded
that Hamann-Todd individuals with these traumatic
injuries were at a significantly high risk of mortality.
The age, sex, and race specific fracture patterns, which
characterize the Hamann-Todd sample, were found to
correspond strongly to those seen in modern European
and American communities. Therefore, the replicative
goals of the study were confirmed with few exceptions.
Results are summarized as follows.
(1) White women displayed a significantly greater
frequency of distal radius fractures compared to
all other sex/race subgroups.
(2) Age at onset for increased incidence of distal
radius fractures in White females occurred dur-
ing the 40 to 60 year period.
(3) Caucasian women displayed a significantly
greater frequency of hip fractures compared to
all other sex/ race subgroups.
(4) A low frequency of hip fractures characterized
all sex/ race subgroups that were under 60 years
of age at time of death.
(5) Age at onset for increased incidence of hip
fractures in White females occurred during the
20
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
6th decade, and peak frequency of hip trauma
was seen in the ninth decade.
(6) Age at onset for increased incidence of hip
fractures in White males occurred during the
seventh decade, 10 years later than females, and
peak frequency of hip trauma was observed in
the ninth decade.
(7) Black men and women showed a low frequency
of distal radius and proximal femur fractures at
all ages compared to White men and women,
respectively.
(8) Black men and women that were over 60 years of
age showed a significantly greater frequency of
distal radius fractures compared to those Blacks
in the 18 to 39 year age group.
(9) Black males exhibited the lowest frequency of
distal radius and proximal femur fractures com-
pared to all other sex/ race subgroups. The only
exception to this was that Black females had a
slightly lower total number of hip fractures
compared to males.
The most dramatic difference between fracture pat-
terns reported in modern groups and those which were
observed for the Hamann-Todd sample concerns the
much greater total frequency with which distal radius and
proximal femur fractures occurred in the latter group.
Distal radius fracture ratios indicate that the total
frequency of traumatic injury for Hamann-Todd was (a)
17 times greater in White women, (b) 33 times greater in
White men, (c) 6 times greater in Black women, and (d) 9
times greater in Black men, compared to a modern
group.
Similarly, hip fracture ratios show that the total
frequency of these traumatic injuries for Hamann-Todd
was (a) 10 times greater in White females, (b) 1 1 times
greater in White males, (c) nearly identical in Black
females, and (d) 5 times greater in Black males, compared
to a modern group. These findings support the conclu-
sion that individuals in the Hamann-Todd Collection
represent a cross section of early twentieth century urban
industrial society that was at very high risk of traumatic
injury.
Although hip fracture patterns appear to be a primary
consequence of age progressive skeletal fragility, it is
suggested that bone loss alone may not explain the early
onset and high incidence of distal radius fractures that are
known to characterize middle-aged Caucasian women.
Alternatively, it is posited that vasomotor disturbances,
which accompany estrogen withdrawal in climacteric
women, may play a more important role in initiating a
greater frequency of accidental falls. The latter would
result in elevated distal radius fracture rates in climacteric
White females, prior to significant reduction of skeletal
mass and biomechanical strength.
Finally, it is further suggested that adult Caucasian
women in pre-estrogen replacement therapy societies
may be characterized by a A/J-Composite distal radius
fracture pattern. Here, vasomotor disturbances would
dominate fracture risk in the climacteric years, and
skeletal fragility would dominate fracture risk in individ-
uals over 60 years of age. With respect to Black men and
women, we regard the lack of comparative information
about fracture patterns, aging effects in general, and the
menopausal syndrome in Black women in particular, to
constitute gerontological issues that deserve attention in
future research.
Acknowledgments
The CSU-OAFS (Cleveland State University— Old Age
Fracture Study) research group would like to express our
gratitude to the faculty and staff of the Cleveland Museum of
Natural History and Cleveland State University department of
anthropology for their assistance in and support of this
investigation. The authors would specifically like to thank
Bruce Latimer, Dr. Byron Hoffman, Lyman Jellema, Dave
Brose, Stephanie Belovich, Lori Linden, Jari Cardinal, Toni
Hutton, Luba Gudz, Doris Webster, and the wonderful
volunteer staff at the CMNH. We would also like to thank Leon
Soule, John Lallo, Laura Martin, John Blank, and Paul
Aspelin at CSU for their kindness and encouragement. Finally,
we thank Jim Ohman, Ted Coombs, C. Owen Lovejoy, and
Richard Meindl who tolerated endless bad puns about break-
throughs in our research. The senior author would also like to
thank the junior authors whose interest, motivation, and pro-
ductivity were enormously appreciated. Also, I hope that the
junior authors did not find this project to have been too much of
a traumatic experience. This research was funded in full by Pass-
the-Hat, Inc.
References
Aitken, J. M., D. M. Hart, R. Lindsay, and J. B. Anderson.
1976. Prevention of bone loss following oophorectomy in
premonopausal women. Israel Journal of Medical Sciences
12:607-614.
Alffram, P. A., and G. C. H. Bauer. 1962. Epidemiology of
fractures of the forearm. A biomechanical investigation of
bone strength. Journal of Bone and Joint Surgery 44A:
105-1 14.
Bacorn, R. W., and J. F. Kurtzke. 1953. Colles’fracture: a study
of two thousand cases from the New York State workmen’s
compensation board. Journal of Bone and Joint Surgery
35A: 643-658.
Bauer, G. C. H. 1960. Epidemiology of fractures in aged
persons. Clinical Orthopedics 17:219-225.
Bengner, U., and O. Johnell. 1985. Increasing incidence of
forearm fractures. A comparison of epidemiologic patterns
25 years apart. Acta Orthopedica Scandinavica 56:158-160.
Buhr, A. J., and A. M. Cooke. 1959. Fracture Patterns. Lancet
(1):53 1-536.
Bungay, G. T., M. P. Vessey, and C. K. McPherson. 1980.
Study of symptoms in middle life with special reference to the
menopause. British Medical Journal 28 1 : 1 8 1 - 1 83.
1987
HAMANN-TODD SKELETAL COLLECTION
21
Casper, R. F., S. S. C. Yen, and M. M. Wilkes. 1979.
Menopausal flushes: a neuroendocrine link with pulsatile
luteinizing hormone secretion. Science 205:823-825.
Chalmers, J. 1973. Distribution of osteoporotic changes in the
aging skeleton. Clinics in Endocrinology and Metabolism
2:203-219.
Chalmers, J., and J. K. Weaver. 1966. Cancellous bone: its
strength and changes with aging and an evaluation of some
methods for measuring its mineral content. Journal of Bone
and Joint Surgery 48A:299-308.
Davis, M. E., N. M. Strandjord, and L. H. Lanzl. 1966.
Estrogens and the aging process. The direction, prevention,
and retardation of osteoporosis. Journal of the American
Medical Association 196:219-224.
Dennerstein, L., and G. D. Burrows. 1978. A review of studies
of the psychological symptoms found at the menopause.
Maturitas 1:55-64.
Feldman, B. M., A. Voda, and E. Gromseth. 1985. The
prevalence of hot flash and associated variables among
perimenopausal women. Research in Nursing and Health
8:261-268.
Frandsen, P. A.,andT. Kruse. 1983. Flip fractures in the county
of Funen, Denmark. Acta Orthopedica Scandinavica 54:68 1 —
686.
Freeman, M. A. R., R. C. Todd, and C. J. Pirie. 1974. The role
of fatigue in the pathogenesis of senile femoral neck fractures.
Journal of Bone and Joint Surgery 56B:698~702.
Frommer, D. J. 1964. Changing age of the menopause. British
Medical Journal 2:349-35 1 .
Gallagher, J. C., and B. L. Riggs. 1978. Current concepts in
nutrition. (Nutrition and bone disease). New England Jour-
nal of Medicine 298: 193-195.
Goldsmith, N. F., J. O. Picetti, and C. Garcia. 1973. Bone
mineral in the radius and vertebral osteoporosis in an insured
population. Journal of Bone and Joint Surgery 55A:1276-
1293.
Gordon, G. S. 1961. Osteoporosis: diagnosis and treatment.
Texas State Journal of Medicine 57:740-747.
Gordon, G. S. 1984. Prevention of bone loss and fractures in
women. Maturitas 6:225-242.
Hammar, M., G. Berg, L. Fahraeus, and U. Larsson-Cohn.
1984. Climacteric symptoms in an unselected sample of
Swedish women. Maturitas 6:345-350.
Horsman, A., B. E. C. Nordin, J. C. Gallagher, P. A. Kirby, R.
M. Milner, and M. Simpson. 1977. Observations on sequen-
tial changes in bone mass in postmenopausal women: a
controlled trial of estrogen and calcium therapy. Calcified
Tissue Research 22:217-224.
Jaffe, H. L. 1972. Metabolic, Degenerative, and Inflammatory
Diseases of Bones and Joints. Philadelphia: Lea & Febiger.
Johnell, O., B. Nilsson, K. Obrant, and I. Sernbo. 1984. Age
and sex patterns of hip fracture — changes in 30 years. Acta
Orthopedica Scandinavica 55:290-292.
Kilbourne, E. D.,and W. G. Smillie. 1969. Human Ecology and
Public Health. Fourth edition. London: MacMillan Com-
pany.
Knowelden, J., A. J. Buhr, and O. Dunbar. 1964. Incidence of
fractures in persons over 35 years of age. British Journal of
Preventive Social Medicine 1 8: 1 30- 141 .
Koefed, H. 1983. Comminuted displaced Colles’fractures. Acta
Orthopedica Scandinavica 54:307-31 1.
Lafferty, F. W., G. E. Spencer, and O. H. Pearson. 1969. Effects
of androgens, estrogens, and high calcium intake on bone
formation and resorption in osteoporosis. American Journal
of Medicine 36:514-528.
Little, K. 1973. Bone Behavior. New York and London:
Academic Press.
McKinlay, S. M., and M. Jefferys. 1974. The menopausal
syndrome. British Journal of Preventive Social Medicine
28:108-1 15.
Meema, S., M. L. Bunker, and H. E. Meema. 1975. Preventa-
tive effect of estrogen on postmenopausal bone loss. Archives
of Internal Medicine 135:1436-1440.
Meema, S., and H. E. Meema. 1976. Menopausal bone loss and
estrogen replacement. Israel Journal of Medical Sciences
12:601-606.
Nicholas, .1. A., and P. D. Wilson. 1963. Osteoporosis of the
aged spine. Clinical Orthopedics 26:19-33.
Nilsson, B. E., and K. J. Obrant. 1 978. Secular tendencies of the
incidence of fracture of the upper end of the femur. Acta
Orthopedica Scandinavica 49:389-391.
Nordin, B. E. C., A. Horsman, R. Brook, and D. A. Williams.
1976. The relationship between estrogen status and bone loss
in postmenopausal women. Clinical Endocrinology 5, sup-
plement, 353-361.
Older, T. M., E. V. Stabler, and W. H. Cassebaum. 1965. Colles
fracture: evaluation and selection of therapy. Journal of
Trauma 5:469-476.
Riggs, B. L., J. Jowsey, P. J. Kelley, D. L. Hoffman, and C. D.
Arnaud. 1976. Effects of oral therapy with calcium and
vitamin D in primary oseoporosis. Journal of Clinical
Endocrinology & Metabolism 42: 1 139-1 144.
Robbins, S. L. 1974. Pathologic Basis of Disease. Philadelphia:
W. B. Saunders Company.
Solgaard, S., and V. S. Petersen. 1985. Epidemiology of distal
radius fractures. Acta Orthopedica Scandinavica 56:391-
393.
Swanson, A. J. G., and G. Murdoch. 1983. Fractured neck of
the femur. Pattern of incidence and implications. Acta
Orthopedica Scandinavica 54:348-355.
Tataryn, 1. V., P. Lomax, .1. G. Bajorek, W. Chesarek, D. R.
Meldrum, and H. L. Judd. 1980. Postmenopausal hot
flushes: a disorder of thermoregulation. Maturitas 2:101-
107.
Thomlinson, R. 1965. Population Dynamics. Causes and
Consequences of World Demographic Change. New York:
Random House.
Trotter, M., G. E. Broman, and R. R. Peterson. 1960. Densities
of bones of White and Negro skeletons. Journal of Bone and
Joint Surgery 42A:50~58.
Voda, A. M. 198! Climacteric hot flash. Maturitas 3:73-90.
Wallace, W. A. 1983. The increasing incidence of fractures of
the proximal femur: an orthopdaedic epidemic. Lancet (i):
1413-1414.
Wasnich, R. D., P. D. Ross, L. K. Heilbrun, and J. M. Vogel.
1985. Prediction of postmenopausal fracture risk with use of
bone mineral measurements. American Journal of Obstetrics
and Gynecology 153:745-751.
Zain Elabdien, B. S., S. Olerud, G. Karlstrom, and B. Smedby.
1984. Incidence of hip fracture in Uppsala County 1965-1980.
Acta Orthopedica Scandinavica 55:284-289.
Zetterberg, C., and G. B. J. Andersson. 1982. Fractures of the
22
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
proximal end of the femur in Goteborg, Sweden, 1940-1979.
Acta Orthopedica Scandinavica 53:419-426.
Appendix A
CSU-OAFS (Cleveland State University — Old Age Fracture
Study) List of All Fractured Individuals
Code:
a) H-T No., Hamann-Todd Catalogue Number.
b) Age, Cadaver Record Stated Age at Death.
c) Distal Radius, all fractures that were located within 4
cm. of the distal articular surface of the radius.
d) Femur (PIC), Intracapsular Proximal Femur Frac-
tures.
e) Femur (PEC), Extracapsular Proximal Femur Frac-
tures.
f) R, Right Side Affected.
g) L, Left Side Affected.
h) r, all fractures which were remodeled, or evinced only
minimal and initial remodeling, at time of death.
CSU-OAF STUDY
Adult White Female Fracture Specimens
H-T
No.
Age
Distal
Radius
Femur
Head
Femur
Neck
2478
35
R-r/L-r
0228
40
R-r
0415
40
R-r
0783
42
R-u
1119
45
R-r
2437
45
R-r
2414
47
R-r/L-r
0355
49
L-u
0229
50
L-r
0411
50
L-u
0742
50
R-r
1149
50
R-r
3123
51
L-r
1846
53
R-r
3140
53
L-r
0785
54
L-r
3118
54
L-r
3352
54
L-u
2949
58
L-u/ R-u
3327
58
R-r
0022
60
R-r/L-r
L-u
0541
60
R-u
1426
60
L-r
3132
61
R-r
R-r
3164
62
L-r
0234
65
R-u
1200
68
L-U
2307
68
R-r
R-r
1413
72
R-r
H-T
No.
Age
Distal Femur
Radius Head
Femur
Neck
1811
73
L-r
3151
74
R-u
1451
75
L-r
1750
75
R-r
2188
75
R-r
2021
76
R-u
3183
77
L-r
3278
77
L-u
1191
78
L-r
1779
78
L-r
3359
78
L-r
L-u
0927
80
L-r
1680
80
R-r
1753
80
R-r
1754
80
L-u
1402
81
L-r/ R-r R-r
1505
81
L-r
L-r
2007
81
R-r
1581
83
L-r/ R-r
1024
84
R-r/L-r
1433
84
R-r
2121
85
L-r
3032
89
R-r
L-u
1708
93
R-u
Adult White Male Fracture Specimens
2689
38
L-r
2580
40
R-r
1765
40
L-r
2618
48
L-u
2198
58
R-r
1064
60
L-r
1464
60
L-u
0963
65
R-u
2626
65
R-r
3147
69
L-r
0759
73
L-r
3341
75
L-u
1989
75
L-u
1 166
76
L-r
2801
78
R-r
3081
78
R-r
1021
81
L-r
1307
83
R-u
2635
84
R-r
1663
85
R-r
3214
89
L-u
3025
93
R-u
Adult Black Female Fracture Specimens
0918
35
R-u
2311
35
R-r
2612
39
R-r
2942
39
L-r
1702
40
R-r
3131
43
L-r
1987
HAMANN-TODD SKELETAL COLLECTION
23
H-T
Distal
Femur
Femur
Appendix B
No. Age
Radius
Head
Neck
TABLE 22
1022 46
R-r
Distal Radius and Proximal Femur Fracture
3182 48
L-r
Symmetry Data
2660 5 1
R-r
Individuals
2269 58
R-u/ L-u
with
0839 60
R-r
Fractures Left
Right Bilateral
0773 60
R-u
N n\
%
n2 % W
%
1551 60
L-r
1912 60
L-r
DISTAL RADIUS
3174 60
L-r
White Female
40 16
40.0
19 47.5 5
12.5
2773 64
L-r
White Male
15 10
66.7
5 33.3 0
0.0
0751 65
R-r
Black Female
17 9
52.9
7 41.2 1
5.9
2039 65
L-r
Black Male
5 1
20.0
3 60.0 1
20.0
1367 72
L-r
Total
77 36
46.8
32 41.6 7
9.1
0967 87
R-r
FEMUR (PIC)
Adult Black Male Lracture Specimens
White Female
4 0
0.0
2 50.0 2
50.0
White Male
2 0
0.0
2 100.0 0
0.0
0598 25
R-r
Black Female
2 2
100.0
0 0.0 0
0.0
2368 26
L-r/R-
r
Black Male
4 2
50.0
1 25.0 1
25.0
1245 31
L-r
129! 42
L-r
Total
12 4
33.3
5 41.7 3
25.0
1338 42
R-r
FEMUR (PEC)
2339 52
L-r
White Female
16 10
62.5
6 37.5 0
0.0
1452 60
R-r
White Male
5 2
40.0
3 60.0 0
0.0
1 /33 60
R-r
Black Female
1 0
0.0
I 100.0 0
0.0
1528 67
R-r/L-r
Black Male
0 0
0.0
0 0.0 0
0.0
Total
22 12
54.5
10 45.5 0
0.0
TABLE 21
Summary of Lracture Observations
Appendix C
White White
Black
Black
TABLE 23
Female Male
Female
Male
Total
Distal Radius and Proximal Femur Fracture
Remodeling Status
Distal Radius Lx
45 15
18
6
84
Total
Remodeled
41 12
15
6
74
Fractures
Remodeled Unremodeled
Unremodeled
4 3
3
0
10
Observed
n\
% n2
%
Unilateral Left
16 10
9
1
36
Unilateral Right
19 5
7
3
34
DISTAL RADIUS
Bilateral
5 0
1
1
7
White Female
45
41
91.1 4
8.9
White Male
15
12
80.0 3
20.0
Lemur (PIC) Lx
6 2
2
5
15
Black Female
18
15
83.3 3
16.7
Remodeled
4 2
2
5
13
Black Male
6
6
100.0 0
0.0
Unremodeled
2 0
0
0
2
Total
84
74
88.1 10
1 1.9
Unilateral Left
0 0
2
2
4
Unilateral Right
2 2
0
1
5
FEMUR (PIC)
Bilateral
2 0
0
1
3
White Female
6
4
66.7 2
33.3
Lemur (PEC) Lx
16 5
1
0
22
White Male
2
2
100.0 0
0.0
Remodeled
5 0
0
0
5
Black Female
2
2
100.0 0
0.0
Unremodeled
11 5
1
0
17
Black Male
5
5
100.0 2
13.3
Unilateral Left
10 2
0
0
12
Total
15
13
86.7 2
13.3
Unilateral Right
6 3
1
0
10
Bilateral
0 0
0
0
0
FEMUR (PEC)
White Female
16
5
313 11
68.8
TNILxa
53 22
20
9
104
White Male
5
0
0.0 5
100.0
TNLxb
67 22
21
1 1
121
Black Female
1
0
0.0 1
100.0
Black Male
0
0
22.7 17
77.3
“Total no. of individuals that displayed one
or more OAF fractures.
Total no. of OAF fractures observed in each subsample.
Total
22
5
22.7 17
77.3
24
MENSFORTH, SUROVEC, AND CUNKLE
No. 42
Appendix D
TABLE 24 TABLE 26
White Female Age Specific Fracture Frequencies Black Female Age Specific Fracture Frequencies
Age
Group
N
Distal
Radius
rt\ %
Femur
(PIC)
n2 %
Femur
(PEC)
m %
Total
Proximal
Femur
ru %
Age
Group
N
Distal
Radius
n\ %
Femur
(PIC)
12 %
Femur
(PEC)
rt 3 %
Total
Proximal
Femur
ru %
18-24
5
0
0.0
0
0.0
0
0.0
0
0.0
18-24
30
0
0.0
0
0.0
0
0.0
0
0.0
25-29
14
0
0.0
0
0.0
0
0.0
0
0.0
25-29
34
0
0.0
0
0.0
0
0.0
0
0.0
30-34
10
0
0.0
0
0.0
0
0.0
0
0.0
30-34
29
0
0.0
0
0.0
0
0.0
0
0.0
35-39
24
1
4.2
0
0.0
0
0.0
0
0.0
35-39
50
4
8.0
0
0.0
0
0.0
0
0.0
40-44
19
3
15.8
0
0.0
0
0.0
0
0.0
40-44
29
2
6.9
0
0.0
0
0.0
0
0.0
45-49
22
3
13.6
0
0.0
1
4.5
1
4.5
45-49
21
2
9.5
0
0.0
0
0.0
0
0.0
50-54
24
9
37.5
0
0.0
1
4.2
1
4.2
50-54
23
1
4.4
0
0.0
0
0.0
0
0.0
55-59
9
1
11.1
1
11.1
0
0.0
1
11.1
55-59
8
1
12.5
0
0.0
0
0.0
0
0.0
60—64
19
5
26.3
0
0.0
2
10.5
2
10.5
60-64
10
4
40.0
1
10.0
1
10.0
2
20.0
65-69
13
1
7.7
0
0.0
3
23.1
3
23.1
65-69
10
2
20.0
0
0.0
0
0.0
0
0.0
70-74
13
3
23.1
0
0.0
0
0.0
0
0.0
70-74
10
0
0.0
1
10.0
0
0.0
1
10.0
75-79
20
6
30.0
0
0.0
4
20.0
4
20.0
75-79
2
0
0.0
0
0.0
0
0.0
0
0.0
+80
15
8
53.3
3
20.0
5
33.3
8
53.3
+80
6
1
16.7
0
0.0
0
0.0
0
0.0
Total
207
40
19.3
4
1.9
16
7.7
20
9.7
Total
262
17
6.5
2
0.8
1
0.4
3
LI
TABLE 25
White Male Age Specific Fracture Frequencies
TABLE 27
Black Male Age Specific Fracture Frequencies
Age
Group
N
Distal
Radius
ru %
Femur
(PIC)
ni %
Femur
(PEC)
rt) %
Total
Proximal
Femur
ru %
Age
Group
N
Distal
Radius
rt\ %
Femur
(PIC)
rt2 %
Femur
(PEC)
rt 3 %
Total
Proximal
Femur
ru %
18-24
5
0
0.0
0
0.0
0
0.0
0
0.0
18-24
30
0
0.0
0
0.0
0
0.0
0
0.0
25-29
14
0
0.0
0
0.0
0
0.0
0
0.0
25-29
34
1
2.9
1
2.9
0
0.0
1
2.9
30-34
10
0
0.0
0
0.0
0
0.0
0
0.0
30-34
29
0
0.0
1
3.4
0
0.0
1
3.4
35-39
24
1
4.2
0
0.0
0
0.0
0
0.0
35-39
50
0
0.0
0
0.0
0
0.0
0
0.0
40-44
19
2
10.5
0
0.0
0
0.0
0
0.0
40-44
29
1
3.4
1
3.4
0
0.0
1
3.4
45-49
22
1
4.5
0
0.0
0
0.0
0
0.0
45-49
21
0
0.0
0
0.0
0
0.0
0
0.0
50-54
24
0
0.0
0
0.0
0
0.0
0
0.0
50-54
23
1
4.3
0
0.0
0
0.0
0
0.0
55-59
9
1
11.1
0
0.0
0
0.0
0
0.0
55-59
8
0
0.0
0
0.0
0
0.0
0
0.0
60-64
19
2
10.5
0
0.0
0
0.0
0
0.0
60-64
10
1
10.0
1
10.0
0
0.0
1
10.0
65-69
13
1
7.7
1
7.7
1
7.7
2
15.4
65-69
10
1
10.0
0
0.0
0
0.0
0
0.0
70-74
13
1
7.7
0
0.0
0
0.0
0
0.0
70-74
10
0
0.0
0
0.0
0
0.0
0
0.0
75-79
20
4
20.0
0
0.0
1
5.0
1
5.0
75-79
2
0
0.0
0
0.0
0
0.0
0
0.0
+80
15
2
13.3
1
6.7
3
20.0
4
26.7
+80
6
0
0.0
0
0.0
0
0.0
0
0.0
Total
207
15
7.2
2
1.0
5
2.4
7
3.4
Total
262
5
1.9
4
1.5
0
0.0
4
1.5
ECHINOCARIS , A MID-PALEOZOIC CRUSTACEAN:
AN ANNOTATED BIBLIOGRAPHY
JOSEPH T. HANNIBAL
Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
and
RODNEY M. FELDMANN
Department of Geology
Kent State University
Kent, Ohio 44242
Abstract
Echinocaris is a genus of malacostracan crustacean that has been re-
ported in rocks ranging in age from Early Devonian through Early M is-
sissippian in North America, Europe, Asia, and New Zealand. More
than 200 citations in primary and secondary literature have focused on
one or more aspects of the taxon including its morphology, paleoecol-
ogy, and phylogenetic position. This annotated bibliography represents
a compilation of citations to works on the taxon and summarizes impor-
tant observations in each reference.
Introduction
Purpose
The genus Echinocaris embraces a small group of mal-
acostracan arthropods known exclusively from the fossil
record of the mid-Paleozoic. Twenty-seven species have
been formally assigned to the genus and more than 20
citations have associated specimens with Echinocaris,
which were so poorly preserved as to preclude precise
identification. Of those identified and described, proba-
bly more than 1 5 are valid species. The remainder are
either synonymous with valid taxa or are assignable to
different organisms. The described species range in age
from Early Devonian through Early Mississippian. Al-
though the preponderance of taxa have been described
from marine rocks in the central and eastern United
States, some have been identified in Canada, Great Brit-
ain, the Soviet Union, Burma, and New Zealand.
Thus, the echinocaridids are a relatively small group of
organisms that one might assume would occupy a rela-
tively insignificant position in the study of earth history.
On the contrary, Echinocaris — or taxa that at one time
were referred to the genus — and trace fossils presumed to
be due to the activity of Echinocaris have been cited in
more than 200 technical and nontechnical publications.
Kirtlandia, No. 42
© by The Cleveland Museum of Natural History
Echinocaris has often been illustrated (Fig. 1), and has
been selected by authors as either the sole example, or one
of a few examples, of Paleozoic crustaceans. Reference to
the genus is made, not only in the primary literature, but
also in a number of textbooks of paleontology and histor-
ical geology published in the last 100 years.
The prominence of Echinocaris in the literature proba-
bly stems, in part, from the influential role played by clas-
sical, early American paleontologists, such as James Hall,
Robert Parr Whitfield, and John Clarke. Their work not
only formed the basis for description of the rich and var-
ied Devonian faunas of northeastern United States but
also helped to set the standard of excellence for subse-
quent paleontological work in America. These authors
described in detail, and considered the implications of,
Echinocaris in the record of Devonian rocks.
With the exception of the trilobites and the eurypterids,
Echinocaris was probably the most common large ar-
thropod to be collected from Paleozoic rocks. As such,
the taxon was recognized early as a potential ancestor to
subsequent malacostracans and thus, served as a splendid
illustration of the rootstock of this group. It was also rec-
ognized, in the nineteenth century, that Echinocaris bore
resemblance to the living leptostracans, including Nebalia
and its allies. The presumed central position of Echino-
caris in the phylogeny of the Malacostraca and its mor-
phological similarity to living phyllocarids, coupled with
its large size and distinctive morphology, rendered the
genus a useful and lasting example of early arthropod.
Because of the extraordinary prominence of Echino-
caris in the literature and because the genus has been cited
as the possible precursor to subsequent crustacean groups
it is appropriate to focus attention on the knowledge of
the group accumulated since the first specimens of Cera-
tiocaris\-Echinocaris]punctata were described by Hal! in
1 863. The purposes of this work, therefore, are to present
Fig. 1. Echinocaris punctata (Hall, 1863). A. Reproduction of the original illustration by Beecher
(1884, Plate I, fig. 1 3), which is the most widely reproduced, and probably the best known, illustration
of a specimen of Echinocaris. Beecher described the specimen as, “a nearly entire individual showing
the form and relation of the parts, and the number of naked abdominal segments.” B. Photograph of
the specimen of E. punctata , New York State Museum 13342/4 (old number, NYSM4401), from
which the drawing was made. Bar scales = I cm.
as complete a bibliography of the genus Echinocaris as
possible and to provide annotations summarizing the sig-
nificant systematic, morphologic, stratigraphic, and eco-
logic observations in these works. Additionally, trace fos-
sils ascribed to the work of Echinocaris spp. and taxa
previously thought to be included within Echinocaris
(particularly I Dunsopterus wrightianus and Eleuthero-
caris wrightiana ) are treated. Finally, because many of the
references cited treat a broad spectrum of Paleozoic ar-
thropods, the bibliography will serve as a general entry
into the literature of early malacostracans.
Annotation procedure
Every attempt has been made to secure as complete a
bibliography on Echinocaris as possible. Standard biblio-
graphic sources, such as the Bibliography and Index of
Geology and its precursors, including the Bibliography of
North American Geology and the Bibliography and In-
dex of Geology Exclusive of North America , as well as the
Zoological Record and Van Straelen and Schmitz’s Fos-
silium Catalogus (1934) on phyllocarids, served as the
basic resources. Additionally, references within articles
on Paleozoic arthropods were searched for citations of
the genus. Examination of stratigraphical and general pa-
leontological articles relevant to the units known to con-
tain echinocaridids provided additional references. A
number of references were also drawn from W. D. E
Rolfe’s manuscript of an updated manuscript Fossilium
Catalogus on phyllocarids.
Each of the citations has been examined to provide de-
tailed page references to significant information relative
to Echinocaris. Although these page citations refer to
nearly all mentions of the genus, some passing notices
have not been cited in situations where more inclusive or
specific record is made in the same work. We have at-
tempted to identify citations in textbooks and nontechni-
cal literature, but we have not attempted to confirm ci-
1987
ECHINOCA RIS: AN ANNOTATED BIBLIOGRAPHY
27
tations in every edition of every textbook. Unpublished
theses, with the exception of two which contain impor-
tant, relevant remarks, are not included. In all cases, the
original materials, or photocopies of the original mate-
rials, have been examined so that errors of citation in sec-
ondary literature have been identified and eliminated.
Original spellings have been preserved and, where incor-
rect, corrected spellings are indicated in square brackets.
When the trivial portion of species names have been de-
rived from the names of individual people, and their ini-
tial letter was capitalized in the original article or book,
they are not capitalized here (nor is this considered a
misspelling herein). Authors of species, when not sup-
plied by authors cited, are added.
Two kinds of documenting information are given
within the annotations. References to previous works and
acknowledgments of illustrations cited by the author are
presented in the text or are set off by parentheses. Those
that were not cited by the original author, but are inserted
by us, are set off by square brackets. Additionally, com-
ments intended to clarify points of misinterpretation in
the original text, to correct misspellings, or to guide the
user to other, relevant references are placed in square
brackets. Specimen numbers and depositories, when not
provided previously in the literature, if known to us, are
also provided in square brackets.
To avoid misunderstanding, we have attempted to uti-
lize standard terminology with reference to citations.
Thus, when the term, “after” is used, it means that il-
lustrative material has been taken directly from a pre-
vious author, without modification. In the event that il-
lustrations have been altered, but an original source can
be identified, we have used the terminology, “modified
from.”
Cross references are included where a paper could be
construed as having been written by an author other than
the first author listed. Also, cross references are provided
where an author’s name may have varied spellings in
English.
Acknowledgments
W. D. 1. Rolfe, The Royal Museum of Scotland, Edin-
burgh, generously allowed Hannibal access to his exten-
sive files on Echinocaris , including his manuscript copy of
an updated Fossilium Catalogus on phyllocarids, at an
early stage of this work. Natalie Sidel, Cleveland Public
Library, provided translations of materials in Russian.
Rolfe, Loren Babcock, then at Kent State University,
D. E. Butler, then with the British Geological Survey, and
Andrew K. Rindsberg, Golden, Colorado, provided us
with copies of important articles. Roy E. Plotnick, the
University of Illinois, Chicago, provided information on
? Dunsopterus wrightianus. Kathleen M. Farago, Lake-
wood Public Library, provided invaluable assistance with
proofreading. Rolfe and Murray J. Copeland, Geological
Survey of Canada, provided helpful reviews of this paper.
This work was supported by NSF Grant EAR 83 1 2798 to
Feldmann. A portion of this study was supported by a
Cleveland Museum of Natural History Staff Enhance-
ment Award, funded by Mr. and Mrs. Willard Hirsh, to
Hannibal. Contribution 312, Department of Geology,
Kent State University, Kent, Ohio 44242.
Bibliography
Allan, R. S. 1935. The fauna of the Reefton beds (Devo-
nian), New Zealand; with notes on Lower Devonian
animal communities in relation to the base of the
Devonian System. New Zealand Department of Scien-
tific and Industrial Research, Geological Survey
Branch, Palaeontological Bulletin 14:1-72.
Reported, based on an elongate, tapering, cercopod
(p. 30; PI . Ill, fig. 8), Echinocaris sp. indet. from the
Fig. 2. Illustrations of an arthropod spine (New Zealand Geological
Survey AR1 131), originally identified by Allan (1935, p. 30) as Echino-
caris sp. indet. The specimen does not, however, seem to be referable to
Echinocaris. A. Counterpart of the sole specimen. B. Specimen illus-
trated by Allan ( 1935, Plate 3, fig. 8). C. Enlargement of a portion of the
specimen illustrated in B showing articulated spinelets arranged along
the convex margin of the spine. Bar scales = 1 cm.
28
HANNIBAL AND FELDMANN
No. 42
Reefton beds of New Zealand. [The specimen. No.
AR 1131 in the collection of the New Zealand Geo-
logical Survey, is from the Bolitho Mudstone For-
mation, Middle Siegenian to Lower Emsian (Fig. 2).
It is five or more sided and bears spinelets that are
less than one mm long on at least two of these sides.
It may be a phyllocarid telson-spine, but it is unlike
that of any Echinocaris species.]
Baird, G. C. 1978. Pebbly phosphorites in shale: a key to
recognition of a widespread submarine discontinuity in
the Middle Devonian of New York. Journal of Sedi-
mentary Petrology 48:545-555.
Reported (p. 548) the uppermost portion of the
Kashong Shale Member of the Moscow Formation
in western New York state to be, “characterized by
the brachiopods Tropidoleptus carinatus and Lin-
gula sp., ramose and fenestrate bryozoans, the trilo-
bites Greenops boothi and Dipleura dekayi , and the
phyllocarid Echinocaris sp.”
1979. Sedimentary relationships of Portland
Point and associated Middle Devonian rocks in central
and western New York. New York State Museum Bul-
letin 433:1-24.
Noted (p. 12) Tropidoleptus, Pleurodictyum [sic],
Modiomorpha, Pseudoaviculopecten, Orthonata
[.sic], Grammysia and Echinocaris as common taxa
in the Middle Devonian Kashong Member of the
Moscow Formation in central and western New
York.
Barrois, C. 1891. Memoire sur la faune du gres armori-
cain. Annales de la Societe Geologique du Nord
19:134-237.
Compared Trigonocarys lebescontei n. gen. and sp.
with some other phyllocarids, including (p. 224)
Equisitides [-1 Duns opt erus wrightianus ] and (p.
225) Echinocarvs [= Echinocaris-, Rolfe, 1969, noted
that Echinocarys was a nomen vanum],
Barron, L. S., and F. R. Ettensohn. 1980. A bibliography
of the paleontology and paleoecology of the Devonian-
Mississippian black-shale sequence in North America.
U.S. Department of Energy, Morgantown Energy
Technology Center, DOE/METC/5202-13, 86 p.
Listed several publications dealing with, or mention-
ing, Echinocaris .
1981. Paleoecology of the Devonian- Mississip-
pian black-shale sequence in eastern Kentucky with an
atlas of some common fossils. U.S. Department of
Energy, Morgantown Energy Technology Center,
DOE/ ET/ 12040-151, 75 p.
Noted (p. 21) that, “fossils such as Echinocaris and
Spathiocaris are commonly reported from the black
shales” and that, “if they were indeed crustaceans,
they probably led a nektonic life in the upper part of
the water column of the black-shale sea and were as-
sociated with flotage as are modern arthropods in
the Sargasso Sea. ” [ Echinocaris is certainly a crusta-
cean. However, unlike Spathiocaris, it is not com-
monly found in classic black shales, such as the
Cleveland Shale; see also Hlavin, 1976.]
Bate, R. H., J. S. H. Collins, J. E. Robinson, and W. D. I.
Rolfe. 1967. Arthropoda: Crustacea. In Die Fossil
Record , p. 535-563. Geological Society of London.
Noted (p. 555) the range of the order Archaeostraca
Claus to be Tremadocian through Carnian.
Bather. See British Museum (Natural History).
Beecher, C. E. 1884. Ceratiocaridae from the Chemung
and Waverly groups of Pennsylvania. Second Geologi-
cal Survey of Pennsylvania, Report of Progress,
PPPT-22.
Reviewed (p. 1-3) literature on Paleozoic “phyllo-
pods,” discussed (p. 3-4) “the optic spot” of these
crustaceans, and listed (p. 5) geologic ranges for
Echinocaris punctata (Hall), E. [ -"'Ceratiocaris ”]
longicauda Hall [see Hall, 1863], E. sublevis Whit-
field, E. pustulosa Whitfield, E. multinodosa Whit-
field, and E. socialis Beecher. Redescribed E. punc-
tata (p. 6-10; PI. I, figs. 13-16) [Fig. 1 herein] from
the Hamilton Group of New York and described E.
socialis n. sp. (p. 10-13; PI. I figs. 1-12) from the
shales at the base of the Chemung Group at Warren,
Pennsylvania. Also described mandibles (p. 9-10;
PI. 2, figs. 9-11) associated with specimens of E.
punctata , but belonging to “a species otherwise un-
known.” Presented (p. 4) a labeled, diagrammatic il-
lustration of Echinocaris. [Museum numbers for E.
socialis, PI. I, figs. 5 and 6, E. punctata , PI. I, figs.
13-16, and for the mandibles, PI. II, figs. 9-11, are
given in Clarke and Ruedemann, 1903. These au-
thors listed the mandibles as E. punctata .]
1900. Restoration of Stylonurus lacoanus, a
giant arthropod from the Upper Devonian of the
United States. American Journal of Science, Fourth
Series, 10:145-150. Also, Geological Magazine, New
Series, 7(XI):48 1-485.
Discussed (p. 148) Stylonurus (?) ( Echinocaris ?)
Dunsopterus] wrightianus (Dawson), suggesting
that the type specimen represents two proximal
joints of one of the large crawling feet of a form re-
lated to Stylonurus.
1902. Revision of the Phyllocarida from the
Chemung and Waverly groups of Pennsylvania. Quar-
terly Journal of the Geological Society of London
58:441-449.
Described Echinocaris randallii n. sp. (p. 443; PI.
XVIII, fig. 8) and E. clarkii n. sp. (p. 443-444; PI.
XVIII, fig- 9) from the Waverly Group, Lower Car-
boniferous, near Warren, Pennsylvania. Both forms
were compared to E. socialis Beecher. E. clarkii was
also compared to Pephricaris horripilata Clarke.
1987
ECHINOCARIS: AN ANNOTATED BIBLIOGRAPHY
29
Also presented additional information on the mor-
phology of E. socialis (p. 441-442; PI. XVII; PI.
XVIII, figs. 1 -7) based on specimens from the “phyl-
locarid-beds” in the Upper Devonian Chemung
Group at Warren, Pennsylvania. [Beecher’s primary
type, and figured, specimens are in the Peabody Mu-
seum of Natural History, Yale University.]
Bernard, F. 1895. Elements de Paleontologie. Librairie
J.-B. Bailliere et Fils, Paris, 1 168 p.
Diagnosed (p. 328) Echinocaris and provided (fig.
159A) a labelled, diagrammatic illustration [modi-
fied from Beecher, not Beecker, 1884] of E. punctata
(Hall).
Bigsby, .1. J. 1878. Thesaurus Devonico-Carboniferus:
The Flora and Fauna of the Devonian and Carbon-
iferous Periods. John van Voorst, London, 447 p.
Listed (p. 26) Ceratiocaris armata Hall [= Echino-
caris punctata (Hall)] and Ceratiocaris punctata
Hall [= Echinocaris punctata (Hall)] from New York
State, after Hall, 1863.
Bolton, T. E. 1966. Catalogue of Type Invertebrate Fos-
sils of the Geological Survey of Canada. Vol. III. Geo-
logical Survey of Canada, Department of Mines and
Technical Surveys, Ottawa, 203 p.
Listed (p. 115) type and figured specimens of Echi-
nocaris beecheri Copeland, E. castorensis Copeland,
E. consanguina Eller, and E. sp. figured and de-
scribed by Copeland in Copeland and Bolton, 1960.
[See Copeland, 1960a.]
Boule, M., and J. Piveteau. 1935. Fossiles: Elements de
Paleontologie . Masson et cie, Paris, 899 p.
Mentioned (p. 175) Echinocaris as a Devonian
malacostracan.
British Museum (Natural History). 1907. A Guide to the
Fossil Invertebrate Animals in the Department of
Geology and Palaeontology in the British Museum
( Natural History ) . . . The Trustees of the Museum,
London, 182 p.
Noted (p. 95) that the shield [carapace] of
Echinocaris is bivalved. “This book has been written
by Dr. Francis Arthur Bather . . .” [from the
Preface],
1911 .A Guide to the Fossil Invertebrate Animals
in the Department of Geology and Palaeontology in
the British Museum ( Natural History) . . . Second
Edition. The Trustees of the Museum, London, 183 p.
Made the same comments (p. 95) as in the entry
above. “The First Edition of this Guide . . . was
written by Dr. Francis Arthur Bather . . . who has
also . . . revised the book for this, the second, Edi-
tion” [from the Preface],
Brooks, H. K. 1957. Chelicerata, Trilobitomorphia, Crus-
tacea (exclusive of Ostracoda) and Myriapoda. In
Treatise on Marine Ecology and Paleoecology , Vol-
ume 2, Paleoecology, edited by H. S. Ladd, p. 895-929.
Geological Society of America Memoir 67.
Noted (p. 898) the burrows on the carapace of a spec-
imen of Echinocaris punctata (Hall) shown in Clarke,
1919 [= Clarke, 1921],
Brown, R. W. 1956. Composition of Scientific Words.
Revised edition. Smithsonian Institution Press, Wash-
ington, D.C., 882 p.
Included (p. 235) Echinocaris punctata (Hall) as an
example of an animal named using the Latin caris.
Buehler, E. J., and I. H. Tesmer. 1963. Geology of Erie
County, New York. Buffalo Society of Natural
Sciences Bulletin 21 (3): 1-1 18.
Listed (p. 51) Echinocaris punctata (Hall) and E. sp.
as occurring in the Ledyard Shale Member and (p.
56) Echinocaris punctata (Hall) and “mandibles of
Phvllocaris" [.vzc] as occurring in the Middle Devo-
nian Wanakah Shale Member of the Ludlowville
Formation in Erie County, New York. Other taxa,
including additional phyllocarids, occurring in these
units are also listed.
Butler, D. E. 1980. North Devon Athenaeum — Barnsta-
ple, B. Figured Devonian fossils in the collections. The
Geological Curator 2(9- 1 0): 588— 592.
Listed (p. 59 1 ) specimens of Echinocaris whidbornei
Jones and Woodward and E. sloliensis Partridge,
figured by Partridge (1902), as being in the collec-
tions of the North Devon Athenaeum. Supplied
specimen numbers.
1981. Marine faunas from concealed Devonian
rocks of southern England and their reflection of the
Frasnian transgression. Geological Magazine 118:
679-697.
Reported Echinocaris from Frasnian or early Fa-
mennian rocks from the Little Chishill borehole (p.
689), Essex, England and the Steeple Aston borehole
(p. 691), Oxfordshire, England.
Carll, J. F. 1883. Geological Report on Warren County
and the neighboring oil regions with additional oil well
records. Second Pennsylvania Geological Survey Re-
port: 1-439.
Included (p. 304-307) a description of a stratigraph-
ic section of the rocks in the Warren area by F. A.
Randall. Reported (p. 306) crustaceans ( Ceratio -
caris!) from division R at Tanner’s Hill near the
brewery in western Warren. Also reported (p. 304)
that fossils given to the state museum by Randall
were keyed to this section. [The brewery mentioned
was probably that of Adolph Saltsman, once located
in King’s Hollow. The rocks noted as containing
crustaceans are in the Conewango Formation
(=Chadakoin & Venango formations). The crusta-
ceans probably included one or more of the species
of Echinocaris described in Beecher, 1884 and/or
1902, although it is not possible to reconcile Ran-
30
HANNIBAL AND FELDMANN
No. 42
dall’s section with the exact horizons given by
Beecher.]
Case, G. R. 1982. A Pictorial Guide to Fossils. Van Nos-
trand Reinhold, New York, 515 p.
Suggested (p. 130-132) Pseudodontichthys whitei
Skeels to be a junior synonym of Echinocaris. [This
is probably incorrect; see Rolfe and Denison, 1966.]
Illustrated (fig. 15-3) a specimen of P. whitei from
the Silica Shale from Milan, Michigan and (fig.
15-6) Echinocaris socialis Beecher from the Upper
Moravian beds of the Tully Limestone, Ithaca
Group, Moravia, New York. [The latter specimen is
not E. socialis. It is, however, comparable to E.
punctata (Hall).]
Caster, K. E. 1930. Higher fossil faunas of the Upper Al-
legheny. Bulletins of American Paleontology 15(58):
1-332.
Listed and illustrated Echinocaris socialis Beecher
(p. 97; PI. 55, figs. 4, 6-8 after [modified from]
Beecher, 1902) from the “Chemung Group” at
Warren, Pennsylvania and E. clarkii Beecher (p. 97;
PI. 54, fig. 5 after [modified from] Beecher, 1902) and
E. randalli Beecher [=E. randallii] (p. 98; PI. 55, fig.
5, after [modified from] Beecher, 1902) from the
Waverly Group at Warren.
1934. The stratigraphy and paleontology of
northwestern Pennsylvania. Part I: Stratigraphy. Bul-
letins of American Paleontology 2 1 (7 1 ): 1 - 1 85.
Listed (p. 75) the following occurrences of Echino-
caris in the “Bradfordian fauna” of Ohio, Pennsylva-
nia, and New York; E. socialis Beecher in the Upper
Chadakoin (Ellicott) Member, E. socialist and E.
clarkiC. Beecher in the Amity Shale Member, and E.
randalli Beecher [-E. randallii ] in the Oswayo —
lower Riceville Member and the Cussewago series.
Chadwick, G. H. 1935. Faunal differentiation in the
Upper Devonian. Bulletin of the Geological Society of
America 46:305-342.
Listed Eleutherocaris whitfieldi (Clarke) and Cera-
tiocarisf!) \=“ Ceratiocaris"] heecheri as a diagnostic
species (p. 3 15) and Echinocaris punctata (Hall) as a
nondiagnostic and last-appearing species of the Na-
ples Group (p. 317). Listed Echinocaris condylepis
Hall and Clarke as a diagnostic species of the Can-
adaway Group (p. 325), Echinocaris socialis Whit-
field as a diagnostic species of the Conewango Group
(p. 330), Echinocaris clarkii Beecher and E. randalli
Beecher [=iT. randallii ] as diagnostic of the Cus-
sewago-Knapp (p. 366), and Stylonurus (?)[-r!Dun-
sopterus ] wrightianus (Dawson) diagnostic of the
Chemung Group (p. 320).
Chamberlin, T. C., and R. D. Salisbury. 1905 (a publica-
tion date of 1907 is indicated on the title page). Geol-
ogy. Vol. II, Earth History. Genesis — Paleozoic. Sec-
ond Edition, Revised. Henry Holt and Company, New
York, 692 p.
Noted that, in contrast to trilobites, phyllocarids and
cirripeds were amply represented in the Hamilton
fauna (p. 473), illustrating (fig. 210a) Echinocaris
punctata (Hall) [modified from Beecher, 1884],
1909. A College Text-book of Geology. Henry
Holt and Company, New York, 978 p.
Illustrated (fig. 419d) [modified from Beecher, 1884]
Echinocaris punctata (Hall) on a figure showing
“representative Hamilton fossils,” and referred to
the species as, “a crustacean more highly organized
than the trilobites.”
Chernyshev. See Tschernyshev
Chhibber, H. L. 1934. The Geology of Burma. McMillan,
London, 538 p.
Listed (p. 168) Echinocaris asiatica Reed as one of
the most important fossils found in the Devonian
Wetwin shales of Burma.
Chlup&c, I. 1963. Phyllocarid crustaceans from the Silu-
rian and Devonian of Czechoslovakia. Palaeontology
6( 1 ):97— 1 18.
Corrected (p. 98, 1 13) the family name Echinocari-
dae Clarke (in Zittel, 1900) to Echinocarididae.
Clarke, J. M. 1885a. A brief outline of the geological suc-
cession in Ontario Co., N. Y., to accompany a map.
Report of the [New York] State Geologist for the year
1884-. 9-22.
Listed (p. 20) Echinocaris [= Eleutherocaris ] whit-
fieldi Clarke as occurring in the “Portage” group in
Ontario county.
1 885b. On the higher Devonian faunas of Ontar-
io County, New York. U. S. Geological Survey Bulletin
16:1-86.
Commented (p. 44-45) on Echinocaris and described
Echinocaris [= Eleutherocaris] whitfieldi n. sp. (p. 45,
PI. II, figs. 3, 4) from the Devonian Naples beds at
Hatch Hill, Naples, New York. [Clarke (1902) later
erected the genus Eleutherocaris for this species.]
Also described Ceratiocaris [-“Ceratiocaris"] bee-
cheri n. sp. (p. 44, PI. II, fig. 1) [see Clarke, 1892],
Noted (p. 66) that he had not found any specimens of
Equisetides wrightiana Dawson [=? Dunsopterus
wrightianus (Dawson)].
1891. The fauna with Goniatites intumescens ,
Beyrich, in western New York. The American Geolo-
gist 8:86-105.
Listed (p. 93) Echinocaris [= Eleutherocaris ] whit-
fieldi Clarke and Ed beecheri Clarke [nomen nu-
dum] as occurring in the G. intumescens fauna of the
Naples beds. Noted, in introducing the list, that,
“many interesting forms are undescribed and their
affinities can only be indicated.” [Echinocarisl
beecheri Clarke, 1891, not Copeland, 1960, has
never been illustrated or described. Although it is
probable that Clarke was referring to an undescribed
form, it is possible that he could have been referring
to “ Ceratiocaris ” beecheri Clarke, 1885.]
1987
ECHINOCARIS. AN ANNOTATED BIBLIOGRAPHY
31
1892. List of the original and illustrated speci-
mens in the palaeontological collections, Part I
Crustacea. 45th Annual Report of the New York State
Museum : 373-437 and 1 1th Annual Report of the New
York State Geologist'. 57-121.
Listed type and figured specimens (including repro-
ductions) of crustaceans in the collections of the New
York State Museum. Included, under the main head-
ing Echinocaris (p. 427-430), specimens of E. punc-
tata ( Hall), mandibles of Phyllocarida ( E. punctata ),
(?) Ceratiocaris [ Ceratiocaris ”] longicauda Hall,
Echinocaris [-Eleutherocaris] whitfieldi Clarke, (?)
Ceratiocaris [-” Ceratiocaris”] beecheri Clarke, E.
condylepis Hall and Clarke, E. socialis Beecher, E.
multinodosa Whitfield, E. pustulosa Whitfield, and
E. sub/aevis [ -sublevis ] Whitfield. [It is unclear why
C. beecheri was listed under Echinocaris.] Also in-
cluded, (p. 426) Stylonurusl ( Echinocaris ?) [=? Dun-
sopterus] wrightianus (Dawson).
1898a. The stratigraphic and faunal relations of
the Oneonta sandstones and shales, the Ithaca and Por-
tage groups in central New York. New York State Mu-
seum, Forty-ninth Annual Report of the Regents ,
1895, 2:27-81.
Listed (p. 53) Echinocaris [-Eleutherocaris] whit-
fieldi Clarke, and E. (?) beecheri Clarke [nomen nu-
dum; see Clarke, 1891] in a list of species of the nor-
mal Portage (Naples) fauna in Ontario and Living-
ston counties. New York.
1898b. Notes on some crustaceans from the
Chemung Group of New York. I. A singularly orna-
mented phyllocarid genus, Pephricaris. New York
State Museum, Forty-ninth Annual Report of the Re-
gents, \S95,2:73\-733.
Compared (p. 731-733) Pephricaris horripilata n.
sp., from the Chemung Sandstone at Alfred, New
York, with Echinocaris , and concluded that the new
species was closely related to Echinocaris.
1902. Notes on Paleozoic crustaceans, 3, some
Devonic Phyllocarida from New York. 54th Report of
the New York State Museum , 1 (Appendix 3):97— 103.
Noted (p. 98) that, although no phyllocarids were
common in the Hamilton fauna, Echinocaris punc-
tata (Hall) is more common than other species and
found (p. 99) some Ithaca rocks at Laurens and near
Noblesville crowded with E. punctata and Rhino-
caris columbina Clarke. Also noted that these forms
“in every recorded instance” are found in true marine
faunas. Discussed (p. 103) the history of the name
Eleutherocaris , noting that he had introduced the
name in “Eastman-Zittel’s” Text-book of Paleontol-
ogy for Ceratiocaris [= Eleutherocaris ] whitfieldi
Clarke [undoubtedly Clarke was referring to the
taxon then known as Echinocaris whitfieldi Clarke;
seeZittel, 1900] and briefly compared Eleutherocaris
to Echinocaris (p. 103).
1904. Naples fauna in western New York. New
York State Museum Memoir 6:199-454.
Listed (p. 352) Ceratiocaris ( Echinocaris ?) as occur-
ring in the Kellwasser Limestone of the Eifel area
[Germany], Included Eleutherocaris whitfieldi
Clarke and Stylonurusl [=? Dunsopterus] wrightia-
nus (Dawson) in a list of taxa (p. 358) found in the
Cashaqua shale in the Naples section. Listed (p. 360)
S.l wrightianus and E. whitfieldi as occurring,
rarely, in the Naples subprovince, and Echinocaris ?
longicauda Hall [-E. punctata (Hall)] as occurring in
the Styliola or prenuncial fauna of the Naples sub-
province. Noted (p. 373) Whitfield’s description of
Echinocaris multinodosa, E. pustulosa , and E. sub-
laevis [-E. sublevis] from the Huron [actually the
Chagrin] Shale and that Echinocaris could be found
in the Intumescens zone of New York.
1905. Ithaca fauna of central New York. New
York State Museum Bulletin , 82:53-70.
Listed (p. 61 ) Echinocaris punctata (Hall) as having
been collected from the Ithaca beds at Norwich (p.
55) and Burdick (p. 58) and from the Sherburne
Sandstone at Noblesville and Laurens (p. 59).
1921. Organic dependence and disease; their
origin and significance. New York State Museum Bul-
letin, 221-222:1-113. Also, Yale University Press,
New Haven.
Illustrated (fig. 86) with a specimen of Echinocaris
punctata (Hall) from the Hamilton Group, consist-
ing of a single left valve, with marks of Clionolithes
borings. [The cover gives a date of publication of
1919 in addition to 1921, thus this article is some-
times cited as being published in 1919. The specimen
of E. punctata is Princeton University 89238, from
the Moscow Shale, Smyrna, New York.]
Clarke, J. M. See also Zittel, K. A.
Clarke, J. M.,and D. D. Luther. 1904. Stratigraphic and
paleontologic map of Canandaigua and Naples quad-
rangles. New York State Museum Bulletin , 63, Paleon-
tology, 7: 1-76.
Listed (p. 59) Echinocaris ? [=“ Ceratiocaris”] longi-
cauda Hall as occurring in the Genundewa Limestone
of the Genesee [Group] and (p. 61) Eleutherocaris
whitfieldi Clarke and Stylonurusl [=? Dunsopterus]
wrightianus (Dawson) as occurring in the Naples
fauna of the Cashaqua Shale of the Portage [Sonyea
Group],
Clarke, J. M., and R. Ruedemann. 1903. Catalogue of
type specimens of Paleozoic fossils in New York State
Museum. New York State Museum Bulletin , 65, Pa-
leontology, 8:1-847.
Listed museum numbers, rock units, localities, and
other data for type specimens (including hypotypes
and plastotypes) of Echinocaris condylepis Hall and
Clarke (p. 700-701), E. multinodosa Whitfield (p.
701), E. punctata (Hall) (p. 701-703), E. pustulosa
32
HANNIBAL AND FELDMANN
No. 42
Whitfield (p. 704), E. socialis Beecher (p. 704), and E.
sublaevis Whitfield [ =E . sublev is] (p. 704), as well as
Eleutherocaris whitfieldi Clarke (p. 705), Ceratio-
caris [=“ Ceratiocaris”] longicauda Hall (p. 668) and
Stylonurus (?) [=? Dunsopterus] wrightianus (Daw-
son) (p. 761).
1912. The Eurypterida of New York. New York
State Museum Memoir 14:1-439.
Discussed (p. 311-312), provided a synonymy for (p.
311) and illustrated (text fig. 68) the type and sole
specimen of Stylonurus (?) [=? Dunsopterus] wright-
ianus (Dawson) (-Echinocaris wrightianus), and
concurred with Beecher, 1900, that the specimen rep-
resented the crawling legs of a form related to Stylo-
nurus. [See also Dawson, 1881b, and Waterston,
1968.]
Claypole, E. W. 1903. The Devonian Era in the Ohio
basin. Part II, Devonian palaeontology of the Appa-
lachian Gulf. American Geologist 32:240-250.
Listed (p. 249), after Newberry, 1873, Ceratiocaris ?
and Ceratiocaris sp. [not Ceratiocaris , but at least in
part Echinocaris ; see Newberry, 1873] as occurring
in the Erie [^Chagrin] Shale of Ohio.
Cleevely, R. J. 1983. World Palaeontological Collections.
British Museum (Natural History) and Mansell Pub-
lishing, London, 365 p.
Noted (p. 52) that C. E. Beecher donated specimens
of Echinocaris [including a metatype of E. socialis],
and Tropidocaris to the British Museum (Natural
History) and (p. 225) that the North Devon Athe-
naeum has material of E. M. Partridge, including
Echinocaris whidbornei. [See also Butler, 1980.]
Cleland, H. F. 1903. A study of the fauna of the Hamilton
Formation of the Cayuga Lake section in central New
York. U.S. Geological Survey Bulletin 206:1-112.
Reported (p. 81) specimens of Echinocaris punctata
(Hall) in the Upper and Lower Hamilton in the
Cayuga Lake section. Listed (p. 104) E. punctata as
occurring, very rarely, in the Michelinia zone, first
Cvpricar della- Athyris zone, and second Cypricar-
della-Athyris zone of the Lower Hamilton, and in
the Cystodictya zone of the Upper Hamilton of the
section.
1911. The fossils and stratigraphy of the Middle
Devonic of Wisconsin. Wisconsin Geological and Nat-
ural History Survey Bulletin, Scientific Series No. 6,
21:1-222.
Described (p. 145-146), after Hall [and Clarke], and
illustrated (PI. XLIV, fig. 3) a specimen of Echino-
caris punctata (Hall) from Middle Devonian rocks
of the Milwaukee cement quarry, Berthelet, Wiscon-
sin. Also described (p. 146) and illustrated (PI.
XLIV, fig. 4) a specimen consisting of an abdomen
and telson that, “differs from those of New York E.
punctata . . . and may be a new . . . species.”
[Both specimens are now in the United States Na-
tional Museum of Natural History, USNM 78216.
See also Monroe and Teller, 1899.]
Coogan, A. H., L. E. Babcock, J. T. Hannibal, D. W.
Martin, K. S. Taylor, and D. C. Wehn. 1986. Late
Devonian and Early Mississippian strata at Stebbins
Gulch, Geauga County, and Quarry Rock, Cuyahoga
County, Ohio. Field Trip Guidebook, Field Trip
Number 1, Geological Society of America Northcen-
tral Section Meeting, Kent, Ohio, 16 p.
Noted (p. 4) that the Chagrin Shale, at some locali-
ties particularly to the east of Stebbins Gulch, con-
tains a fauna with Echinocaris, brachiopods, gas-
tropods, bivalves, nautiloids, and other taxa.
Copeland, M. J. 1960a. The occurrence of Echinocaris
and Spathiocaris (Phyllocarida) in western Canada. In
Canadian fossil Arthropoda, Eurypterida, Phyllocar-
ida and Decapoda, edited by M. J. Copeland and T. E.
Bolton. Geological Survey of Canada Bulletin 60: 1 - 1 1 .
Described Echinocaris castorensis n. sp. (p. 4, fig. 1,
PI. I, nos. 1-5) from the Upper Devonian Alexo
Formation, “Beaver Ridge,” Alberta, comparing it
to E. auricula Eller; E. consanguina Eller (p. 5-6, PI.
I, nos. 6, 6a) from the same unit and locality, com-
paring it to E. condylepis Hall and Clarke; E. sp. (tel-
son) (p. 6, PI. 1, no. 7) from the Upper Devonian
Duvernay Formation, Alberta; and E. beecheri n. sp.
[not E. beecheri of Clarke (nomen nudum); see
Clarke, 1891] (p. 6-7, fig. 2, PI. I, no. 8) from the
Mississippian Banff? Formation, Alberta, compar-
ing it to E. whidbornei Jones and Woodward and E.
randalli Beecher [ -E . randallii]. Also discussed the
stratigraphic position of these species (p. 2) and
listed (p. 3) previously described species of Echino-
caris, including “E.” [=? Dunsopterus] wrightiana
(Dawson) (= Stylonurus ? wrightianus ) from North
America.
1960b. New occurrences of Ceratiocaris and Pty-
chocaris (Phyllocarida) from the Canadian Arctic. In
Canadian fossil Arthropoda, Eurypterida, Phyllocar-
ida and Decapoda, edited by M. J. Copeland and T. E.
Bolton. Geological Survey of Canada Bulletin 60:
49-54.
Diagnosed (p. 51) the Echinocaridae (sic) and pre-
sented a key to the genera in the family: Echinocaris,
Silesicaris, Eleutherocaris and Ptychocaris. Listed
(p. 52) Echinocaris sp. McLaren (then a manuscript
name) in the synonymy for Ptychocaris novaki n. sp.
[See McLaren 1963.]
Copeland, M. J., and T. E. Bolton. 1985. Fossils of Ontar-
io. Part 3: The eurypterids and phyllocarids. Royal On-
tario Museum Life Sciences Miscellaneous Publica-
tions: 1-48.
Noted (p. 39) and illustrated (fig. 25) the occurrence
of Echinocaris sp. [£. cf. E. punctata], collected from
the Middle Devonian, Arkona Formation, Hamilton
Group, at Hungry Hollow, Ontario. Also mentioned
1987
ECHINOCARIS'. AN ANNOTATED BIBLIOGRAPHY
33
(p. 39) that Echinocaris pustulosus ( Hall) is “the typ-
ical Middle Devonian representative of the genus in
New York State.” [E. pustulosa Whitfield is a Late
Devonian form; E. punctata (Hall) is the typical
Middle Devonian form.]
Dacque, E. 1921. Vergleichende biologische Formen-
kunde der fossilen niederen Tiere. Gebruder Born-
traeger, Berlin, 111 p. (Reprinted, 1980, by Arno Press,
New York)
Discussed (p. 289) Paleozoic phyllocarids, illustrat-
ing (fig. 113) Echinocaris [ punctata (Hall)], from
[after] Zittel, 1915 [which is, in turn, modified from
Beecher, 1884], [The figure is reversed.]
Dana, J. D. 1895. Manual of Geology , Lourth Edition.
American Book Company, New York, 1087 p.
Mentioned (p. 614-615) that there are a number of
species of Echinocaris in Portage rocks and noted
the occurrence of Echinocaris [= Eleutherocaris ]
whitfieldi Clarke and E.l beecheri Clarke [nomen
nudum; see Clarke, 1891] in the Portage Naples beds
(p. 620-621, after Clarke 1891, 1892) and the occur-
rence of E. socialis Beecher in the Chemung beds of
New York and Pennsylvania (p. 621). Illustrated E.
punctata (Hall) (fig. 920), [modified from] Beecher
[1884],
Dawson, J. W. 1881a. On new Erian (Devonian) plants
(abstract). Canadian Naturalist , New Series, 9:475-476.
Briefly described (p. 476) a new fern, Equisetites
wrightianus n. sp., from New York. [See also Daw-
son, 1881b.]
1881b. Notes on new Erian (Devonian) plants.
Geological Society of London Quarterly Journal 37:
299-308.
Described (p. 301, PI. XII, fig. 10; PI. Xll I, fig. 20) a
new species of fern, Equisetides wrightiana [ =? Dun-
sopterus wrightianus (Dawson)], from the Portage
Group at Italy, New York. [Several names have been
applied to this enigmatic species. Dawson himself re-
ferred the species to both Equisetites and Equisitides
in 1881. This taxon was subsequently considered re-
ferable to Echinocaris , later, to the Eurypterida,
and, most recently, questionably to the genus Dun-
sopterus : see Waterston, 1968. We have not been
able to locate the holotype of this enigmatic fossil,
but casts of the original, and sole, specimen of the
taxon are in at least two museums, the British Mu-
seum (Natural History) and the New York State Mu-
seum. A copy of correspondence between J. Hall and
H. Woodward on the specimen is also deposited in
the British Museum. One eurypterid specialist, R.
Plotnick (personal communication) doubts that this
specimen really is a portion of an eurypterid. Based
upon the asymmetry of the specimen, as seen in the
illustrations, and our examination of the material at
the British Museum, it could even be a trace fossil.]
1882. The fossil plants of the Erian (Devonian)
and upper Silurian formations of Canada. Part 2. Geo-
logical Survey of Canada Report ; 93-142. [Report 429
in Ferrier’s catalogue.]
Noted (p. 126) Wright’s discovery of Equisetites
wrightiana Dawson [ -1 Dunsopterus wrightianus
(Dawson); see Dawson, 1881b.]
Derby, A. G. 1906. A bibliography of Ohio geology. Part
1. A subject index of the publication of the Geological
Survey of Ohio. . . . Geological Survey of Ohio,
Fourth Series, Bulletin 6:1-233.
Indexed (p. 179) descriptions of Echinocaris in Whit-
field, 1893.
Duluk, C. E. 1964. Fossil fauna of the Silica Formation.
Earth Science 17(6):250— 255. Reprinted in Fossils of
the Mid-continent of North America , 1965, p. 37-42.
Earth Science Publishing Co., Downers Grove, Illi-
nois.
Reported Echinocaris (p. 40) and “jaws” variously
identified as fish and mandibles of crustaceans, “par-
ticularly Echinocaris" (p. 40-41, fig. 15b), from
Middle Devonian rocks exposed in the Medusa Port-
land Cement Quarries in Sylvania, Ohio [see Dun-
kle, 1965; Rolfe, 1966; and Stumm and Chilman,
1969, for further discussion of the identification of
the mandibles],
Dunkle, D. H. 1965. The presumed holocephalan fish
Pseudodontichthys whitei Skeels. Scientific Publica-
tions of the Cleveland Museum of Natural History,
New Series, 4(2): 1-10.
Discussed (p. 3-5) and illustrated (figs. 1 and 2; PI. 1 )
mouth parts, previously identified as representing a
holocephalan fish, from the Silica Shale of Ohio.
Observed that these mouth parts closely resembled
those occurring with Echinocaris punctata as noted
by Beecher ( 1884) and Hall and Clarke (1888). [See
also Rolfe and Denison, 1966.]
Dzik, J. 1980. Isolated mandibles of Early Palaeozoic
phyllocarid Crustacea. Neues Jahrbuch fur Geologie
und Palaontologie, Monatshefte 1 980( 2):87— 1 06.
Observed (p. 97) that the mandibles of Montecarisl
sp. from the Devonian of Poland had a gnathal lobe
like that of Echinocaris , but with different tooth
morphology. Also concluded (p. 101-102) that, “ar-
chaeostracans were feeding on relatively firm food,
which did not need cutting into pieces, but tritura-
tion [sic],” consistent with a postulated habitat dose
to the sea bottom.
Eastman, C. R. 1913. See Zittel 1900.
Easton, W. H. 1960. Invertebrate Paleontology. Harper
and Brothers, New York, 701 p.
Noted (p. 563) Echinocaris to be a characteristic Pa-
leozoic genus of phyllocarid. Illustrated (fig. 1 3. 1 2.8)
E. socialis Beecher, after [modified from] Hall and
Clarke, 1888.
34
HANNIBAL AND FELDMANN
No. 42
Edmonds, E. A., B. J. Williams, and R. T. Taylor. 1979.
Geology of Bideford and Lundy Island. Geological
Survey of Great Britain. Memoir for 1 :50 000 geologi-
cal sheet 292, New Series, with sheets 275, 276, 291 and
part of sheet 308 , 143 p.
Reported (p. 20) Echinocaris sloliensis Partridge and
E. whidbornei Jones and Woodward from the shaly
sequence of the coastal sections of the Upper Devo-
nian Baggy Sandstones (after Goldring, 1971) and
(p. 21) E. sp. and E. whidbornei (PI. 2, fig. 10) from
the inland sections of the Baggy Sandstones in a
quarry near Croyde Hoe, Georgeham, England.
Eller, E. R. 1935. New species of Echinocaris from the
Upper Devonian, of Alfred Station, New York. Annals
of the Carnegie Museum 24:263-274.
Described Echinocaris consanguina n. sp., (p. 268-
269, PI. Ill, figs. 1-4), E. turgidan. sp., (p. 269-270,
PI. Ill, figs. 5, 6) and E. auricula n. sp., (p. 271, PI.
Ill, fig. 7) from the Upper Devonian [Famennian?]
Alfred Shale [Canadaway Group] exposed at Alfred
Station, New York. The species are compared with
each other and with E. punctata (HaU), E. condylepis
Hall and Clarke, E. socialis Beecher (also incorrectly
spelled E. sociales) and E. sublaevis Whitfield [-E.
sublevis], Eller also discussed (p. 264-267) several
aspects of the morphology of Echinocaris finding no
evidence of an eye on the “eye node” and doubting
the existence of a “nuchal furrow” separating the ce-
phalic and thoracic regions.
1937. Echinocaris crosbyensis , a new species
from the Upper Devonian of New York. Annals of the
Carnegie Museum 25:257-259.
Described (p. 257-258) and illustrated (fig. 1) Echi-
nocaris crosbyensis n. sp. from near Crosby Creek,
New York, “possibly from the same horizon” as spe-
cies previously described (Eller, 1935) from the
Alfred Shale, and compared it with E. whidbornei
Jones and Woodward, E. condylepis Hall and
Clarke, E. socialis Beecher, and E. turgida Eller.
Etheridge, R., H. Woodward, and T. R. Jones. 1884. Re-
port of the committee . . . on the fossil Phyllopoda of
the Palaeozoic rocks. Report of the fifty-third meeting
of the British Association for the Advancement of
Science held at Southport in September, 7553:215-223.
Listed (p. 217) Echinocaris as a leperditioid phyl-
lopod [sic] with four spiny exposed abdominal seg-
ments and three telson styles or caudal spines.
1886. Third Report of the committee ... on
the fossil Phyllopoda of the Palaeozoic rocks. Report
of the fifty-fifth meeting of the British Association for
the Advancement of Science held at Aberdeen in Sep-
tember, 7555:326-361.
Abstracted and commented on (p. 358-361) refer-
ences to species of Echinocaris , and taxa at one time
referred to that genus, in Hall, 1863; Whitfield, 1 880;
Dawson, 1881; Beecher, 1884; and Jones and Wood-
ward, 1884.
1889. Sixth report of the committee . . . on the
fossil Phyllopoda of the Palaeozoic rocks. Report of
the fifty-eighth meeting of the British Association for
the Advancement of Science held at Bath in Sep-
tember, 7555:173-181.
Abstracted and commented on references to Echino-
caris, and taxa at one time referred to that genus, in
Clarke, 1885 (p. 175) and Hall and Clarke, 1888 (p.
180). Noted (p. 180) that Hall preferred the name E.
punctata (Hall) over E. armata (Hall).
1891a. Seventh report of the committee ... on
the fossil Phyllopoda of the Palaeozoic rocks. Report
of the sixtieth meeting of the British Association for
the Advancement of Science held at Leeds in Sep-
tember, 7590:63-68. [A reprint of this article indicates
that this paper was read at the Newcastle-upon-Tyne
meeting.]
Briefly commented (p. 63-64) on a newly discovered
specimen of Echinocaris found in a quarry near
Sloly, England. [This specimen was the basis for the
description of E. whidbornei Jones and Woodward,
see Jones and Woodward, 1889.]
1891b. Eighth report of the committee ... on
the fossil Phyllopoda of the Palaeozoic rocks. Report
of the sixtieth meeting of the British Association for the
Advancement of Science held at Leeds in September,
7590:424-428.
Noted (p. 426) the presence of Echinocaris in Devon
and in the Devonian strata of North America.
Feldmann, R. M., R. M. Boswell, and T. W. Kammer.
1986. Tropidocaris salsiusculus, a new rhinocaridid
(Crustacea: Phyllocarida) from the Upper Devonian
Hampshire Formation of West Virginia. Journal of Pa-
leontology 60(2):379— 383.
Noted (p. 379) references to Echinocaris sp. (Wil-
liams and Kindle, 1905) and Echinocaris auricula
Eller (Hannibal and Feldmann, 1985) as theonly rec-
ords of the genus in West Virginia.
Feldmann, R. M., and J. T. Hannibal. 1985a. Finger-
printing fossils. Earth Science 38(1): 14-15.
Commented on the biological relationships of Echi-
nocaris, as presented by Rolfe, 1981. Illustrated a
specimen of E. punctata (HaU) (p. 15), after Hall and
Clarke, 1888, and included a map (p. 15) showing
some locations in North America and England
where echinocaridids [specifically Echinocaris ] have
been found.
1985b. Paleobiogeography of Echinocaris. Geo-
logical Society of America Abstracts with Programs
17(5):287.
Reviewed the paleobiogeography of Echinocaris and
reported a new West Virginia locality [see Hannibal
1987
ECHINOCARIS: AN ANNOTATED BIBLIOGRAPHY
35
and Feldmann, 1985] for Echinocaris. Questioned
certain finds of echinocaridid “cercopods”[see Feld-
mann, Hannibal and Babcock, 1986and Allan, 1935,
herein] and found the genus “to be a good index in
Middle and Late Devonian fine clastic rocks depos-
ited in tropical and subtropical marine habitats lack-
ing in shelly faunal elements.”
Feldmann, R. M., 3. T. Hannibal, and L. E. Babcock.
1986. Fossil worms from the Devonian of North Amer-
ica (. Sphenothallus ) and Burma (“Vermes”) previously
identified as phyllocarid arthropods. Journal of Pa-
leontology 60(2):341-346.
Illustrated (fig. 3), and redescribed as “Vermes” (p.
343-345), fossils previously identified as “cerco-
pods”of Echinocaris asiatica Reed [see Reed, 1908]
from the Wetwin shales of Burma, [if. asiatica is in-
correctly spelled E. asiaticus in the abstract of this
paper.]
Feldmann, R. M., and S. McKenzie. 1981. Echinocaris
multispinosis, a new echinocarid (Phyllocarida) from
the Chagrin Formation (Late Devonian) of Ohio.
Journal of Paleontology 55(2):383— 388.
Described Echinocaris multispinosis n. sp. (p.
385-386; text fig. 2) based on a single specimen from
the Chagrin Shale (Famennian) at Indian Point,
Lake County, Ohio. Noted (p. 386) that the type lo-
cality of the new species, “may well be the type local-
ity of the species described by Whitfield ( 1880).”[See
Whitfield, 1880.] Found that arthropods, predomi-
nantly Echinocaris , comprised 50% of the fauna
found in a sample of concretions at this locality.
Provided a key to the identification of arthropods,
including E. pulcra Sturgeon, Hlavin and Kesling
[=£. pulchra ], E. ohioensis Sturgeon, Hlavin and
Kesling, E. sublaevis Whitfield [-E. sublevis ], and
E. multinodosa Whitfield from the Chagrin and sug-
gested that E. sublaevis and E. ohioensis are quite
similar. Also discussed the paleoecology of Echino-
caris and provided (text fig. 1 , modified from Stur-
geon, Hlavin and Kesling, 1964) an illustration of a
generalized Echinocaris.
Feldmann, R. M., R. G. Osgood, Jr., E. J. Szmuc, and
D. W. Meinke. 1978. Chagrinichnites brooksi , a new
trace fossil of arthropod origin. Journal of Paleontol-
ogy 52(2):287— 294.
Noted (p. 292) that Echinocaris multinodosa Whit-
field, E. sublevis Whitfield, E. pulchra Sturgeon,
Hlavin and Kesling, and E. ohioensis Sturgeon,
Hlavin, and Kesling have been described from the
Famennian age Chagrin Shale in northeastern Ohio
but concluded (p. 292-293) that these, and other, ar-
chaeostracans had morphology that would make it
difficult to ascribe the traces from the Chagrin de-
scribed as Chagrinichnites brooksi n. sp. to them.
Fenton, C. L., and M. A. Fenton. 1958. The Fossil Book.
Doubleday, Garden City, New York, 482 p.
Illustrated (p. 235) Echinocaris punctata (Hall)
[modified from Beecher, 1884],
Fisher, D. W. 1951. Marcasite fauna in the Ludlowville
Formation of western New York. Journal of Paleon-
tology 25:365-37 1 .
Listed (Table 1) Echinocaris (?) sp. as occurring,
rarely, in the marcasite horizon of the Ledyard Shale
in western New York.
Frey, R. W., H. A. Curran, and S. G. Pemberton. 1984.
Tracemaking activities of crabs and their environmen-
tal significance: the ichnogenus Psilonichnus. Journal
of Paleontology 58(2):333— 350.
Listed (p. 333) Chagrinichnites as one of several ich-
nogenera of trace fossils of crustacean origin.
Gekker, R. F. 1941. Deposits, fauna and flora of the main
Devonian field. In Fauna of the Main Devonian Field,
/, edited by R. F. Gekker, p. 1 7-84. USSR Academy of
Sciences, Palaeontological Institute, Moscow.
Noted (p. 51) the occurrence of Echinocaris in the
main Devonian field. Listed, in a faunal distribution
list (p. 65), Echinocaris tudrensis Tschernyshev as
occurring in the Bilova Series (Famennian) and E.
sp. as occurring in the Pskov beds (Frasnian). [In
Russian, with English summary. See also Tscherny-
shev, 1941.]
1983. Tafonomicheskiye i ekologicheskiye oso-
bennosti fauny i flory glavnogo Devonskogo polya
[Taphonomic and ecological characteristics of the
fauna and flora of the main Devonian Field], Aka-
demia Nauk SSSR. Trudy Paleontologicheskoga Insti-
tuta 190:1-144.
Summarized (p. 75, after Tschernyshev, 1941) in-
formation on Echinocaris from the main Devonian
field of the Russian Platform, contrasting the pau-
city of specimens with the greater number of speci-
mens of the genus found in the Devonian sediments
of the Urals. Mentioned E. tudrensis Tschernyshev.
[In Russian.]
Goldring, R. 1971. Shallow-water sedimentation as illus-
trated in the Upper Devonian Baggy Beds. Memoirs of
the Geological Society of London , 5:1-80.
Reported (p. 9) that Echinocaris sp. has been found
in the Diplocraterion yoyo facies near Croyde Hoe,
England and (p. 32) that E. sloliensis Coomaraswany
[sic] [the author of the species is actually Partridge]
and E. whidbornei Jones and Woodward have been
associated with lingulids (p. 31) in the Lingula facies
of the Baggy beds (Famennian). Speculated on the
environment in which Echinocaris lived, noting that
Echinocaris “seems to be very much a facies fossil”
(p. 32), and mentioned Partridge’s description of a
species of Echinocaris (p. 2).
1978. Devonian. In The Ecology of Fossils .edited
36
HANNIBAL AND FELDMANN
No. 42
by W. S. McKerrow, p. 125-145. MIT Press, Cam-
bridge.
Illustrated (fig. 36) Echinocaris [with a somewhat
disproportionately small carapace] in a reconstruc-
tion of the sedimentary environment in which por-
tions of the Upper Devonian Baggy Formation of
North Devon, England, were deposited. Suggested
(p. 140) Echinocaris to be associated with Lingula in
the bay and coastal lagoon facies. One of the echino-
caridids is shown (fig. 36) inside the burrow Diplo-
craterion. Suggested that Echinocaris was probably
epifaunal, based on the morphology of the carapace.
Goldring, R., and F. Langenstrassen. 1979. Open shelf
and near-shore clastic facies in the Devonian. Special
Papers in Palaeontology 23:81-97.
Reported (p. 89) phosphatized carapaces of Echino-
caris to be common in the Lingula facies of the
Baggy Sandstones (Famennian) of North Devon,
England.
Goldring, R., I. P. Tunbridge, A. Whittaker, et al. 1978.
North Devon. In A Field Guide to Selected Areas of
the Devonian of Southwest England , edited by C. T.
Scrutton, p. 8-27. Palaeontological Association, Lon-
don.
Specimens of Echinocaris , as phosphatised internal
molds, were reported (p. 23) to occur with Lingula
and Diplocraterion in a sequence of interbedded
shales and sandstones within the Famennian age
Baggy Sandstones at Path Cove, near Baggy Point,
England.
Goldring, W. 1929. Handbook of Paleontology for Be-
ginners and Amateurs. Part 1. The Fossils. New York
State Museum Handbook 9: 1-356.
Noted (p. 203) Echinocaris on exhibit in the New
York State Museum, Albany, and illustrated (fig.
56B) E. punctata (Hall), after [modified from]
Beecher, 1884 [the illustration is reversed],
Grabau, A. W. 1921. A Textbook of Geology. Part II:
Historical Geology. D. C. Heath and Co., Boston,
976 p.
In a discussion of the Devonian (p. 424) noted that,
“the fresh water (river and estuarine) deposits are
characterized by the remains of crustaceans (fig.
1336) and eurypterids.” Illustrated (Fig. 1336b)
Echinocaris punctata (Hall) [modified from
Beecher, 1884] from the Hamilton. [The illustration
is reversed.]
Grabau, A. W.,and H. W. Shimer. 1910. North American
Index Fossils: Invertebrates . Vol. II. A. G. Seiler and
Company, New York, 909 p.
Diagnosed Echinocaris (p. 376), noting that it had a
free rostrum and no posterolateral spines [no ros-
trum is known; most of the values of E. clarkei[-E.
clarkii ] Beecher, including their posterolateral por-
tions, are surrounded by spinose ornamentation]
and Eleutherocaris (p. 380). Briefly described: E.
punctata (Hall) (p. 376, fig. 1678) after [modified
from] Beecher, [1884] [the figure is reversed]; E. so-
cialis Beecher (p. 376-377, fig. 1679, after [modified
from] Beecher, [1902]); E. sub/aevis Whitfield [=£.
sublevis ] (p. 377, fig. 1680a&b) [modified from
Whitfield, 1890]; E. pustulosa Whitfield (p. 378, fig.
1680c) (modified from Whitfield, 1890]; and E. mul -
tinodosa Whitfield (p. 378, fig. 1 680d) modified from
Whitfield, 1890J. Diagnosed Eleutherocaris and
briefly described Eleutherocaris whitfieldi Clarke (p.
380). These species were also assigned (p. 686) to
faunal provinces.
Grasso, T. X. 1981. Stratigraphy, paleontology and pa-
leoecology of the Upper Hamilton Group (Middle
Devonian), in the Genesee Valley, Livingston County.
In Guidebook to Field Trips, Annual Spring Meeting —
National Association of Geology Teachers — Eastern
Section, June 5-7, 1981, edited by R. M. Liebe, p.
B1-B36.
Reported the presence of Echinocaris in the Ka-
shong Shale Member of the Moscow Formation ex-
posed at Little Beards Creek (Stop #4, p. B— 36) near
Leicester, New York. Brachiopods, bivalves, bry-
ozoans, crinoids, trilobites and Echinocaris are listed
(p. B— 8) as occurring in the Kashong fauna. The
Tropidoleptus — bivalve community, in which Echi-
nocaris occurs, is described (p. B— 1 6) as inhabiting,
“soft substrates in shallow water of low energy, rem-
iniscent of the Ledyard but containing higher levels
of dissolved oxygen.” Also illustrated (fig. B— 1 3)
Echinocaris punctata [after Beecher, 1884],
Gtirich, G. 1929. Silesicaris von Leipe und die Phylloka-
riden uberhaupt. Mitteilungen aus dem mineralogisch-
geologischen Staatsinstitut in Hamburg 1 1:21-90.
Described (p. 76-78) Echinocaris, commenting on
the following Devonian species: E. whidbornei Jones
and Woodward, E. punctata (Hall), E. socialis
Beecher, E. sublaevis Whitfield [=£. sublevis ], E.
condylepis Hall and Clarke, E. pustulosa Whitfield
and E. multinodosa Whitfield. Distinguished Echi-
nocaris from Ceratiocaris and Aristozoe. Noted (p.
82) that Pephricaris horripilata may belong to the ge-
nus Echinocaris. Found (p. 82) Ceratiocaris^ Cera-
tiocaris"\ longicauda Hall to be problematical. Re-
described (p. 86) and commented on the Lower
Carboniferous E. clarke i \-E. clarkii] Beecher, find-
ing the serrated margin of its carapace to be reminis-
cent of Pephricaris. Found (p. 86) the Lower Carbo-
niferous E. randallii Beecher to be little distinguished
from E. socialis. Also redescribed (p. 81-82) Eleu-
therocaris whitfieldi Clarke, distinguishing the genus
Eleutherocaris from Echinocaris . Illustrated E. punc-
tata (Text plate 7, fig. 4) [modified from Hall and
Clarke, 1888], E. socialis (Text plate 8, fig. 1) [modi-
fied from Beecher, 1902], E. sublaevis [=E. sublevis]
(Text plate 8, fig. 2) [modified from Hall and Clarke,
1987
ECHINOCARIS'. AN ANNOTATED BIBLIOGRAPHY
37
1888?], E. condylepis (Text plate 8, fig. 3) [modified
from Hall and Clarke, 1 888], E. clarkei (Text plate 9,
fig. 7) [modified from Beecher, 1902] and Eleuthero-
caris whitfieldi (Text plate 8, fig. 10) [modified from
Hall and Clarke, 1888],
Hall, J. 1863. Contributions to palaeontology, no. 6, on
the occurrence of crustacean remains of the genera Cer-
atiocaris and Dithyrocaris, with a notice of some new
species from the Hamilton Group and Genesee Slate.
16th Report of the New York State Cabinet of Natural
History , Appendix D:71-75.
Described Ceratiocaris armatus n. sp. (p. 72-73; PI.
I, figs. 1-3) [-Echinocaris punctata ( Hall)] based on
abdomen and telson material from the shales of the
Hamilton Group, in Ontario County, New York and
Ceratiocaris! punctatus n. sp. (p. 74; PI. I, fig. 8) [-E.
punctata (Hall)] based on the left half of a carapace
from the shales of the Hamilton Group on the east
shore of Cayuga Lake. [Although E. armatus should
have had priority over its synonym C. [-E.]punctata,
most subsequent authors, with a few exceptions, in-
cluding Jones and Woodward, 1884, have used the
name E. punctata to refer to this species. As most
authors, we refer to this species as Echinocaris punc-
tata.] Also described (p. 73, PI. I, fig. 4-7) Ceratio-
caris longicaudus n. sp. [-"Ceratiocaris" longicauda .
This taxon has been referred to as Echinocaris longi-
cauda in Beecher, 1884, and Miller, 1889, and listed
under Echinocaris in Clarke, 1892. Hall and Clarke
(1882) and Clarke (1892) noted that the specimens
illustrated as fig. 4-6 were other than arthropods.]
1876. Illustrations of Devonian fossils: Gaste-
ropoda, Pteropoda, Cephalopoda, Crustacea and cor-
als of the Upper Helderberg, Hamilton, and Chemung
groups. New York Geological Survey, Palaeontology.
39 pis.
Illustrated [after Hall, 1863] an abdomen and telson
referred to as Ceratiocaris armatus Hall \= Echi-
nocaris punctata (Hall)] (PI. XXIII, figs. 4-5) and a
carapace referred to as Ceratiocaris (Aristozoe)
punctatus [=£. punctata ] (PI. XXIII, fig. 7). Noted
(note accompanying PI. XXIII) that, “as this sheet is
going to press” a specimen was found with a cara-
pace like that of C. punctatus and an abdomen like
that of C. armatus.
1884. [Note accompanying PI. XV] Thirty-fifth
Annual Report on the New York State Museum of
Natural History 35: PI. XV.
Remarked on Equisetides wrightiana Dawson
[-IDunsopterus wrightianus ], referring to it as a
form not unlike Stylonurus. [See also Wright, 1884,
and Dawson, 1881b. ]
Hall, J., and J. M. Clarke. 1888. Trilobites and other
Crustacea of the Oriskany, Upper Helderberg, Hamil-
ton, Portage, Chemung and Catskill groups. New York
State Geological Survey, Palaeontology 7 : i— lxiv,
1-236.
Provided (p. liii) a synonymy for the genus Echino-
caris. Diagnosed (p. liv) Echinocaris and compared
the genus (p. liv— lv) to Aristozoe and Ptychocaris.
Included (p. liv) a sketch of an Echinocaris[E. punc-
tata, modified from Beecher, 1884], Described E.
condylepis n. sp. from greenish shales of the Che-
mung Group, Belmont (p. 173-174; PI. XXIX, figs.
14-17). Provided synonymies and described: Echi-
nocaris punctata (Hall) from the Hamilton Group of
New York State (including mandibles of Phyllocar-
ida found associated with E. punctata ) (p. 166-171,
PI. XXVII, fig. 10; PI. XXVIII, figs. 1-7 [figs. 6 and 7
modified from Hall, 1 863, figs. 3 and 5 modified from
Beecher, 1884], PI. XXIX, figs. 1-8 [figs. 1-2 modi-
fied from Beecher, 1 884] ); Echinocaris [= Eleuthero-
caris] whitfieldi Clarke from the lower beds of the
Portage (“Naples beds” of Clarke), Hatch Hill, Na-
ples, New York (p. 172-173; PI. XXIX, figs. 20-21)
[note the difference between fig. 21 and fig. 4 in
Clarke, 1885; see also Rolfe, 1969, fig. 142,4b]; E.
socialis Beecher from the base of the Chemung
Group as exposed at Warren, Pennsylvania (p.
1 74- 1 76; PI. XXX, figs. 1-12) [figs. 1 -6, 8-9, 11-12
modified from Beecher, 1884]; E. sublaevis Whitfield
[= E. sublevis ] from the Erie Shales [= Chagrin Shale]
at LeRoy, Ohio (p. 176-178; PI. XXIX, figs. 11-13);
E. pustulosa Whitfield from the Erie Shales at LeRoy
(p. 178-179; PI. XXIX, figs. 9-10); and E. multino-
dosa Whitfield from the Erie Shales at LeRoy (p.
180-181; PI. XXIX, figs. 18-19). Noted (p. 1 74) that
E. condylepis is “almost a miniature of E. punctata."
Provided (p. 160) a synonymy for and described (p.
160-162; PI. XXVII, figs. 7-9), as a possible stylo-
nurid, the sole specimen of Stylonurus(!) (Echino-
caris!) [=? Dunsopterus] wrightianus (Dawson). Also
provided synonymies and described Ceratiocaris
[-"Ceratiocaris"] longicauda Hall (p. 163-4, PI.
XXXI, fig. 1 ), noting that some specimens originally
referred to the taxa were, in fact, specimens of Coleo-
lus aciculum Hall. [Many of the illustrations in this
volume are similar to illustrations that have ap-
peared previously in the literature, but were likely
redrawn from original specimens or casts.]
Hannibal, J. T., and R. M. Feldmann. 1981. Arthropod
trace fossils, interpreted as echinocarid escape bur-
rows, from the Chagrin Formation (Late Devonian) of
Ohio. Geological Society of America Abstracts with
Programs 13(7):467.
[See Hannibal and Feldmann, 1983.]
1983. Arthropod trace fossils, interpreted as
echinocarid escape burrows, from the Chagrin Shale
(Late Devonian) of Ohio. Journal of Paleontology
57(4):705— 716.
38
HANNIBAL AND FELDMANN
No. 42
Discussed (p. 713-714) the possible relationships of
Echinocaris to the trace fossil Chagrinichnites os-
goodi n. sp., concluding (p. 713) that, “it is most
probable that the trace maker . . . was one or more
species of Echinocaris." Illustrated (fig. 7) a general-
ized Echinocaris associated with such a trace and
(fig. 8) an Echinocaris with an exploded block dia-
gram of the trace. [The trace had been described pre-
viously by Lesquereux, 1891.]
1985. A phyllocarid crustacean, Echinocaris au-
ricula, from the Late Devonian of West Virginia. Kirt-
landia 41:22-26.
Redescribed (p. 23-25) Echinocaris auricula Eller
based on the holotype and an additional specimen
from Upper Devonian Chemung rocks in Preston
County, West Virginia. Discussed variation in the
species and compared E. auricula to E. castorensis
Copeland and other species. Illustrated the holotype
(fig. 2A), its counterpart (fig. 2B), and the West Vir-
ginia specimen (fig. 2C).
1986. Late Devonian Crustacea, Palaeopalae-
mon newberryi and Concavicaris sp., from the Chagrin
Shale of northeastern Ohio and Alfred Shale of west-
ern New York state. Geological Society of America
Abstracts with Programs 18(4):291.
Noted that the Chagrin Shale of Ohio and the Alfred
Shale of New York contained a fauna of arthropods
dominated by Echinocaris.
Hecker. See Gekker.
Herries, R. S. 1896. Long excursion of West Somerset
and North Devon. Proceedings of the Geologist's As-
sociation 14( 10):433— 440.
Reported (p. 440) the collecting of part and counter-
part of a “fine specimen” of Echinocaris whidbornei
Jones and Woodward “with an impression of an-
other on the same slab” from the Baggy beds in Sloly
Quarry, Georgeham, England. [See Jones and Wood-
ward, 1899, p. 394 for a brief description of this
specimen.]
Hlavin, W. J. 1976. Biostratigraphy of the Late Devonian
black shales on the cratonal margin of the Appalachian
Geosyncline. Unpublished Ph.D. diss., Boston Univer-
sity, 194 p.
Listed (Table 3) the occurrence of “indet. phyllo-
carid”and cf. Echinocaris sp. in the upper five feet of
the Cleveland Shale on Townes Creek, Lorain Co.,
Ohio. [The specimens studied by Hlavin and depos-
ited in the Cleveland Museum of Natural History,
however, do not appear to include any Echinocaris .]
Hoover, K. V. 1960. Devonian-Mississippian shale se-
quence in Ohio. Ohio Department of Natural Re-
sources, Geological Survey Information Circular Num-
ber 27:1-154.
Listed the occurrence (Appendix, p. 144) of Echino-
caris multinodosa Whitfield, E. pustulosa Whitfield
and E. sublevis Whitfield in the Upper Devonian
Chagrin Shale of Ohio.
Jones, T. R. 1883. II — Palaeozoic phyllopoda; as re-
ported on to the British Association, Southport, 1883,
Section C. Geology. Geological Magazine , Decade 2,
10:461-463.
Listed (p. 463) Echinocaris as being leperditioid and
as having four spiny exposed abdominal segments
and three telson elements.
1898a. The fossil Phyllopoda of the Palaeozoic
rocks. Thirteenth report of the committee consisting of
Prof. T. Wiltshire (Chairman), Dr. H. Woodward, and
Prof. T. Rupert Jones (Secretary). Report of the sixty-
seventh meeting of the British Association for the Ad-
vancement of Science held at Toronto in August, 1897:
343-346.
Noted (p. 345-346) Whidborne’s (1896b) redescrip-
tion of Echinocaris whidbornei Jones and Wood-
ward and reported (p. 346) the discovery of two addi-
tional specimens of Echinocaris \= E. whidbornei ,
see Jones and Woodward, 1899] from the Sloly
Mudstone.
1898b. Ill — The fossil Phyllopoda of the Pa-
laeozoic rocks. Geological Magazine , New Series, De-
cade 4, 5:41-45.
Essentially a reprint of Jones, 1898a. The portion on
Echinocaris whidbornei Jones and Woodward (p.
44-45) is identical.
1 900. Fossil Phyllopoda of the Palaeozoic rocks —
fifteenth report of the committee . . . Report of the
sixty-ninth meeting of the British Association for the
Advancement of Science held at Dover in September,
1899 68:403-405.
Noted (p. 405) that one of the specimens of Echino-
caris whidbornei Jones and Woodward discussed in
Jones, 1898a and 1898b, was figured and redescribed
in Jones and Woodward, 1899.
Jones, T. R. See also Etheridge, Woodward, and Jones.
Jones, T. R., and H. Woodward. 1884. Notes on phyllo-
podiform crustaceans, referable to the genus Echinoca-
ris, from the Palaeozoic rocks. Geological Magazine ,
New Series, Decade 3, 1:393-396, PL 13.
Gave a history of the specimens referred to Ceratio-
caris armatus Hall [= Echinocaris punctata (Hall)]
(p. 393, PI. XII, fig. 2, after Hall [1863]) and its syn-
onyms, noting that the trivial name armatus deserves
priority [but see Hall, 1 863]. Reviewed (p. 394) Whit-
field, 1880, and illustrated, for the first time, E. sub-
laevis Whitfield [=£. sublevis] (PI. XIII, figs. 3-5)
and E. pustulosa Whitfield (PL XIII, fig. 6) from a
yet unpublished plate supplied by Whitfield. Recog-
nized (p. 394) only three “well-determined” species of
1987
ECHINOCA RIS: AN ANNOTATED BIBLIOGRAPHY
39
Echinocaris : E. armata (Hall), E. sublaevis Whit-
field, and E. pustulosa Whitfield. Redescribed (p.
395-396) the sole specimen of Equisetides wrighti-
ana Dawson (PI. XIII, fig. la-b; A and B on p. 395)
as representing a portion of the abdomen of a large
Echinocaris, E. wright iana (Dawson) [=? Dunsop-
terus wrightianus]. [See Dawson, 1881b.]
1888. A Monograph of the British Palaeozoic
Phyllopoda (Phyllocarida, Packard). Part 1. Ceratio-
caridae. The Palaeontographical Society, London, 72 p.
Noted (p. 3), in a table of genera of fossil Phyllocar-
ida, that Echinocaris was leperditioid, with spinous
segments, and had a pod-like, oculate, bivalved
carapace.
1889. I. On some new Devonian fossils. Geo-
logical Magazine, New Series, Decade 3, 6:384-388.
Described (p. 385-386) and illustrated (fig. 1) Echi-
nocaris whidbornei n. sp., from the “leaden-blue
shales of the Lingula squamiformis beds in a quarry
near Sloly, close to the three-milestone on the Barn-
staple and Ilfracombe road.” The authors distin-
guished the specimen from E. punctata ( Hall) (illus-
trated as fig. 2, after [modified from] Beecher [ 1884])
and other previously described species on the basis of
the presence of two parallel ridges on the carapace of
the species and other characters. [The holotype, H
419, is deposited in the Sedgwick Museum, Cam-
bridge University. See also Whidborne, 1896b.]
1899. II. — Contributions to fossil Crustacea.
Geological Magazine, New Series, Decade 4, 6:388-
395.
Briefly described (p. 393-394, PI. XV, fig. 6) speci-
mens of Echinocaris whidbornei Jones and Wood-
ward from the Sloly Quarry [near Barnstaple], [See
also Partridge, 1902.]
Jux, U. 1960. Montecaris lehmanni, a new crustacean
from the Rhenish Devonian and the problem of its sys-
tematic position. Journal of Paleontology 34(6): 1 1 29—
1152.
Noted (p. 1 147) the fossil assemblage containing
Echinocaris socialis Beecher, described in Hall and
Clarke, 1888, in a discussion of phyllocarid living
habits. Discussed (p. 1 147-1 148) various aspects of
the carapace of E. crosbyensis Eller, suggesting that
the “eye notches” of Echinocaris and other phyllo-
carids were “the original sites of eyes.” Also dis-
cussed (p. 1148) the mandibles of phyllocarids, in-
cluding those of Dilhyrocaris [that is, Echinocaris ].
Illustrated E. crosbyensis (Text fig. 4B) after [modi-
fied from] Eller, 1 937 [not 1 935] and a specimen of E.
punctata (Hall) (Text fig. 9F) after [modified from]
Hall and Clarke, 1888.
Kesling, R. V., and R. B. Chilman. 1975. Strata and mega-
fossils of the Middle Devonian Silica Formation. Uni-
versity of Michigan Museum of Paleontology Papers
on Paleontology, Number 8:1-408.
Gave (p. 155) a short synonymy for Echinocaris
punctata (Hall) and briefly described the species as
found in the Middle Devonian Silica Formation.
Noted that the species occurs at the Martin-Marietta
Quarry (in Washtenaw County, Michigan), as well
as Silica, Ohio. Illustrated specimens of E. punctata
( PI. 73, fig. 1 ; PI. 93, fig. 5; PI. 1 29, figs. 6-9) [some of
which were previously illustrated by Stumm and
Chilman, 1969] and provided a diagram of the spe-
cies (p. 155, after Stumm and Chilman, 1969).
Kjellesvig-Waering, E. N. 1961 . Eurypterids of the Devo-
nian Holland Quarry Shale of Ohio. Fieldiana, Geol-
ogy 1 4( 5 ): 79— 98.
Listed (p. 8 1) Stylonurus (?) [- 1 Dunsopterus ] wrigh-
tianus (Dawson) in a list of eurypterids from the
Devonian of North America. [See Dawson 1881b.]
1966. A revision of the families and genera of the
Stylonuracea (Eurypterida). Fieldiana, Geology 14(9):
169-197.
Recognized the species Stylonurus ? [=? Dunsop-
terus ] wrightianus (Dawson), as questionably be-
longing to the genus Stylonurus (p. 178). [See Daw-
son 1881b.]
Krestovnikov, V. N. 1960. Nadotryad Phyllocarida Pack-
ard, 1879. In Osnovv Paleontologii, 8, Trilobitoobv-
raznye i rakoobraznye, edited by N. E. Chernysheva, p.
425-429. Akademia Nauk USSR, Moscow.
Diagnosed the Echinocaridae [szc] and Echinocariss
(p. 426). Illustrated E. punctata (Hall) (fig. 1247)
[modified from] Hall and Clarke, 1888 [which is, in
turn, after Beecher, 1884], Mentioned (p. 427) Eleu-
therocaris Clarke as a member of the family. [In
Russian.]
1961. Novye rakoobraznye fillokaridy Paleozoya
Russkoi platformy, Urala, Timana i Donbassa.
Akademia Nauk SSSR, Trudy Geological Institute,
Vpy, 52:1-67.
Reviewed the occurrences of, and summarized strati-
graphic information on, Echinocaridae [sz'c], includ-
ing E. archae [=£■. arschae ] Tschernyshev, E. uralen-
sis Tschernyshev, E. (?) brevicarinata Tschernyshev,
E. tudrensis Tschernyshev, and various E. spp.
Tschernyshev, from the Upper Devonian deposits of
the Ural Mountains and the northwestern region of
the Russian Platform (p. 10, 33, Table 5). Noted the
similarity of Echinocaris sp. from the bank of the
Arsha to E. randalii Beecher [-E. randallii], Pro-
vided maps (figs. 1, 3) showing the distribution of
Echinocaris in the Soviet Union. Also provided a
chart (folded, between p. 8 and 9) showing the distri-
bution of Echinocaris (worldwide) as Middle Devo-
nian through Lower Carboniferous. [In Russian.]
40
HANNIBAL AND FELDMANN
No. 42
La Touche, T. H. D. 1913. Geology of the northern Shan
States. Memoirs of the Geological Survey of India
39(2): 1-379.
Included Echinocaris asiatica Reed in a list of fauna
from the Wetwin shales [in what is now Burma] (Ta-
ble 9) and identified, as its nearest ally, E. punctata
(Hall) from the Hamilton Group of North America.
Lesley, J. P. 1889-1890. A dictionary of the fossils of
Pennsylvania and neighboring states. Second Pennsyl-
vania Geological Survey Report P4: 1-1283.
Illustrated (p. 214) Echinocaris punctata (Hall)
[modified from] Zittel after [modified from] Beecher
[ 1 884] [the illustration is reversed] (and upside down
as noted by J . Hall in the Errata, p. xxi) and Echino-
caris [-Eleutherocaris] whitfieldi Clarke [modified
from] Clarke, 1885. Echinocaris socialis is listed (p.
214) with the note “see Appendix.” [We could not,
however, locate a reference to this taxon in any
appendix.]
Lesquereux, L. 1891. Remarks on some fossil remains
considered as peculiar kinds of marine plants. Proceed-
ings of the U.S. National Museum 13:5-12.
Described (p. 9-1 1, PI . 1 , figs. 4-9), as a fossil plant,
Physophycos bilobatus n. sp. from rocks of the Por-
tage Group [Chagrin Shale] along Lake Erie, near
Cleveland. [Later authors have overlooked this arti-
cle and the name has been unused subsequently.
Specimens have been redescribed as trace fossils
under the generic name Chagrinichnites. A request
to suppress the name P. bilobatus has been submit-
ted to the International Commission on Zoological
Nomenclature. Hannibal and Feldmann, 1983, in-
terpreted one of these forms as having been pro-
duced by Echinocaris .]
McLaren, D. J. 1955. Devonian formations in the Alberta
Rocky Mountains between Bow and Athabasca rivers.
Geological Survey of Canada Bulletin 35:1-59.
Reported (p. 47) well preserved carapaces of Echino-
caris sp., along with Aulopora, Schuchertella, Cyr-
tospirifer and other taxa, from a limestone unit in
Member B of the Upper Devonian Alexo Formation
at Beaver Ridge, Alberta. [The Echinocaris material
is described in Copeland, 1960a.]
1963. Southwestern Ellesmere Island between
Goose Fiord and Bjorne Peninsula. In Geology of the
North-Central Part of the Arctic Archipelago, North-
west Territories (Operation Franklin), by Y. O. Fortier
et al. Geological Survey of Canada Mermoir 320:
310-338.
Reported (p. 321) Echinocaris sp. [= Ptychocaris
novaki Copeland] from a stromatoporoid-coral bio-
herm within the limestone and shale (lower) member
of the Blue Fiord Formation (Middle Devonian) on
the south side of Eids Fiord, Ellesmere Island.
[Copeland (1960a) described the “ Echinocaris "
material reported here as Ptychocaris novaki n. sp.]
Miller, S. A. 1889. North American Geology and Pa-
laeontology for the Use of Amateurs, Students, and
Scientists. Western Methodist Book Concern, Cincin-
nati, 664 p. and Second Appendix, p. 719-793.
Diagnosed (p. 545) Echinocaris and listed (p.
545-546, 787) articles describing E. condylepis Hall,
E. longicauda Hall ( Ceratiocaris [=“Ceratiocaris”\
longicauda ), E. multinodosa Whitfield, E. punctata
(Hall), E. pustulosa Whitfield, E. socialis Beecher,
E. sublaevis (also, spelled correctly as E. sublevis)
Whitfield, E. [= Eleutherocaris] whitfieldi Clarke,
and E. wrightiana (Dawson)[-I Dunsopterus wright-
ianus (Dawson)]. Illustrated (fig. 1002) E. punctata
(Hall) [modified from Beecher, 1884; the figure is
reversed],
Monroe, C. E., and E. E. Teller. 1899. The fauna of the
Devonian formation at Milwaukee, Wisconsin. Jour-
nal of Geology 7:272-283.
Listed (p. 279) Echinocaris (= Ceraliocaris ) [s/c] sp.
as rare in the Devonian rocks in the lower 2 1 feet at
the cement quarry on the Milwaukee River, imme-
diately north of Milwaukee. [See Cleland, 191 1, fora
description of Echinocaris from this locality.]
Moore, R. C., and L. McCormick. 1969. General features
of Crustacea. In Treatise on Invertebrate Paleontol-
ogy, edited by R. C. Moore, Part R, Arthropoda 4(1),
p. R57-R 120. Geological Society of America and Uni-
versity of Kansas Press, Lawrence.
Included (p. R 1 1 3) the Echinocaridinae and the
Echinocarididae in an outline of classification of
crustaceans.
Moore, R. C., C. G. Lalicker, and A. G. Fischer. 1952.
Invertebrate Fossils. McGraw Hill, New York, 766 p.
Illustrated (fig. 14-17C) Echinocaris socialis Beecher
[modified from Hall and Clarke, 1888] from Upper
Devonian, Chemung, sediments of New York.
Moret, L. 1940. Manuel de Paleontologie Animate. Mas-
son et cie, Paris, 675 p.
Considered (p. 294) Echinocaris as a Devonian rep-
resentative of the Phyllocarida which typically in-
habited shallow marine habitats. The phyllocarids
were considered intermediate between entomostra-
cans and malacostracans.
1966. Manuel de Paleontologie Animate. Fifth
Edition. Masson et cie, Paris, 781 p.
Mentioned (p. 329) Echinocaris as a Devonian rep-
resentative of the phyllocarids. The phyllocarids
were considered intermediate between entomostra-
cans and malacostracans.
Morris, S. F. 1980. Catalogue of the type and figured
specimens of fossil Crustacea (excl. Ostracoda), Che-
licerata, Myriapoda and Pycnogonida in the British
Museum (Natural History). British Museum (Natural
History), London, 53 p.
Listed (p. 6) specimen 1.3945, Echinocaris whidbor-
nei Jones and Woodward, figured by Jones and
1987
ECHINOCA RIS: AN ANNOTATED BIBLIOGRAPHY
41
Woodward, 1899, p. 393, PI. 15, fig. 6, and its coun-
terpart, In. 38080, as being deposited in the British
Museum (Natural History).
Muller, A. H. 1963. Lehrbuch der Palaozoo/ogie. Band
II, Invertebraten. Teil 3, Arthropoda 2 — Stomochorda.
Gustav Fischer Verlag, Jena, 698 p.
Diagnosed (p. 92) Echinocaris and illustrated (fig.
1 14) E. socialis Beecher from Roger 1953 [which, in
turn, was modified from Beecher, 1902],
Muller [Mueller], K.. J. 1967. Devonian of Malaya and
Burma. In International Symposium on the Devonian
System , edited by D. H. Oswald, Vol. 1:565-568. Al-
berta Society of Petroleum Geologists, Calgary.
Noted (p. 567), after Pascoe (1959), that the presence
of Echinocaris [E. asiatica Reed] in the Devonian
Wetwin fauna of Burma “seems to indicate the prox-
imity of estuarine conditions.” [ Echinocaris , how-
ever, does not indicate such conditions.]
Murphy, J. L. 1972. Echinocaris punctata (Had) from the
Hamilton Group, Thedford, Ontario. Ohio Journal of
Science 72(3): 155- 157.
Described and illustrated (figs. 1, 2) two specimens
of Echinocaris punctata (Hall), USNM 170561 and
170562, from the Middle Devonian Hamilton Group
in, and near, Thedford, Ontario. Greenops hoothi
(Green), Ponderodictya punctulifera (Hall), Tor-
noceras uniangulare (Conrad), “ Chonetes" lepida
Hall, Styliolina fissurella (Hall) and Tasmanites sp.
are listed as being associated with the Echinocaris.
Newberry, J. S. 1873. Devonian System. Report of the
Geological Survey of Ohio Vol. I, Geology and Paleon-
tology. Part I, Geology: 140-167.
Noted (p. 166-167) that concretions occurring in the
Erie [=Chagrin] Shale, “in the bed of Paine’s Creek,
in the north part of Leroy township,” Ohio, con-
tained two crustaceans, one “probably a species of
Ceratiocaris ,” the other an allied but new genus.
[The crustaceans, Echinocaris and Palaeopa/aemon ,
were described by Whitfield in 1880.]
Nicholson, H. A., and R. Lydekker. 1889. A Manual of
Palaeontology for the Use of Students with a General
Introduction on the Principles of Palaeontology. Third
edition. Vol. 1. William Blackwood and Sons, Edin-
burgh and London, 885 p.
Noted (p. 5 14) that, “the curious genus Echinocaris ,”
occurs in the Devonian of North America and briefly
described its characters including, “an ovoid folded
carapace.”
O'Connell, M. 1916. The habitat of the Eurypterida. Bul-
letin of the Buffalo Society of Natural Sciences
XI(3): 1-277.
Concluded (p. 23) that the single specimen of Stylo-
nurus (?) wrightiana (Dawson) [=? Dunsopterus
wrightianus ( Dawson)] was part of a jointed append-
age. Provided a synonymy (p. 50) including Echi-
nocaris wrightiana Jones and Woodward = Stylonu-
rus (?) wrightianus (Dawson). [See Dawson 1881b.]
Oehlert, D.-P. 1889. Sur le Devonien des environs
d ’Angers. Bulletin Societe Geologique de France , Se-
ries 3, 17:742-791.
Reviewed (p. 770-771) the observations of Hall,
Whitfield (1880), and Jones and Woodward (1884)
that Echinocaris, Tropidocaris , and Elymocaris
could confidently be assigned to the phyllocarids and
were comparable to the living Nebalia and that Aris-
tozoe, Callizoe, and Orozoe were probably also phyl-
locarids, not leperditioid [ostracodes].
Olsson, A. 1912. New and interesting fossils from the
Devonian of New York. Bulletins of American Paleon-
tology 5(23): 1 -7.
Reported (p. 7; PI. 7, figs. 2-4) two specimens [one of
which, fig. 2, is now deposited in the collection of the
Paleontological Research Institute, Ithaca, num-
bered 28297] of Echinocaris punctata (Hall) col-
lected from the Ithaca Shale beds of the Portage
Formation in the McGraw, or University, Quarry on
the Cornell University campus. These specimens
were associated with Spirifer mesastrialis ( Hall) and
Cryptonella eudora Hall. Olsson noted, “its appear-
ance in higher beds [Frasnian], associated with re-
current Hamilton species, and as having changed but
little in the time interval.”
Osgood, R. G., Jr. 1976. Trace fossils of Chagrin Forma-
tion (Upper Devonian, Northeast Ohio). AAPG Bul-
letin 60(4):704.
Mentioned, “enigmatic resting and dwelling traces of
what may be a phyllocarid or eocarid crustacean.”
[The likely phyllocarid would have been Echino-
caris. These trace fossils were later assigned to the
genus Chagrinichnites\see Feldmann et al., 1978 and
Hannibal and Feldmann, 1983.]
Packard, A. S. Jr. 1882. The Palaeozoic allies of Nebalia.
American Naturalist 16:945-953.
Noted (p. 953) that, “ Echinocaris punctatus (Hall)
[ -E. . punctata ] must have been nearly a foot in
length, while the Echinocarides [s/o] . . . described
recently by Mr. R. P. Whitfield were considerably
smaller.” Illustrated an abdomen of E. punctatus
(Hall) [=E. punctata ] (fig. 12), [modified] from Hall
[1863], a carapace of E. multinodosus Whitfield [=E.
multinodosa ] (fig. 10), [modified from? Whitfield’s
unpublished plate], and an abdomen of E. sublevis
Whitfield (fig. 1 1) [modified from? Whitfield’s un-
published plate],
1883. A monograph of the phyllopod Crustacea
of North America, with remarks on the Order Phyllo-
carida. U.S. Geological and Geographical Survey of
the Territories, Twelfth Annual Report , Part I:
295-592.
Diagnosed (p. 450) Echinocaris , listed (p. 450-451)
42
HANNIBAL AND FELDMANN
No. 42
E. sublevis Whitfield (fig. 7 IB, after [modified
from?] Whitfield[’s unpublished plate]), E. punctatus
(Hall) [ -E . punctata] (fig. 70, [modified] from Hall
[1863]), E. armatus (Hall) = E. punctatus, E. pustu-
losus Whitfield [ -E . pustulosa] and E. multinodosus
Whitfield [-E. multinodosa ] (fig. 71 A, after [modi-
fied from?] Whitfield['s unpublished plate]), and
provided (p. 457) a bibliography of fossil species of
phyllocarids. [The figures are the same as those in
Packard, 1882.]
Partridge, E. M. 1902. Echinocaris whidbornei (Jones &
Woodward) and Echinocaris sloliensis, n. sp. Geologi-
cal Magazine , Decade 4, 9:307-308, PI. 17.
Described E. sloliensis n. sp. (p. 307-308, PI. XVII,
figs. 8, 9) and the abdomen of Echinocaris whidbor-
nei Jones and Woodward (p. 307, PI. XVII, fig. 7)
based on specimens from the Marwood beds of Slo-
ley Quarry, near Barnstaple, England. Compared (p.
308) E. sloliensis to E. socialis Beecher. [The deposi-
tory of these specimens is recorded in Butler, 1980.]
Pascoe, E. H. 1959. A Manual of the Geology of India
and Burma. Vol. 2. Government of India Press, Cal-
cutta, 2:485-1343.
Listed (p. 688, after Reed, 1908) Echinocaris asiatica
Reed as occurring in the Upper Devonian Wetwin
fauna of Burma and noted that, “ Echinocaris is
characteristic of brackish if not fresh water,” indicat-
ing, “the proximity of estuarine conditions.” [The
reference to salinity is incorrect.]
Prosser, C. S. 1 898. The classification and distribution of
the Hamilton and Chemung series of central and east-
ern New York, Part 1. New York State Museum,
Forty-ninth Annual Report of the Regents (New York
State Geologist Annual Report 15), 1895, 2:87-222.
Reported (p. 183) Echinocaris cf. [£.] punctata
(Hall) as occurring, very rarely, at a shaly arenaceous
horizon, “probably in the lower part of the Ithaca
formation,” near Stetsonville, New York.
Read, M. C. 1873. Geology of Lake County. Report of
the Geological Survey of Ohio , Vol. I, Geology and Pa-
leontology, Part I, Geology:5 10-519.
Noted (p. 519) the occurrence of a new crustacean in
nodules found in the Erie [Chagrin] Shale in Paine’s
Creek, LeRoy Township, Ohio [This crustacean was
probably Echinocaris ; see Whitfield, 1880.]
Reed, F. R. C. 1908. The Devonian faunas of the northern
Shan States. Palaeontologia Indica , New Series, 2(5):
1-183.
Described (p. 179-180) and illustrated (PI. XX, fig.
21) Echinocaris asiatica n. sp. from the Devonian
Wetwin shales near Wetwin, in what is now Burma.
The description is based on a portion of a single left
valve of a carapace. Reed also described (p. 180) and
illustrated (PI. XX, figs. 22-25) “elongate styliform
bodies” that he determined could be “regarded as the
cercopods of a species of Echinocaris ,” probably E.
asiatica. [These “cercopods” were redescribed by
Feldmann, Hannibal and Babcock, 1986 as
“Vermes.” Reed’s figured specimens are now depos-
ited in the Geological Survey of India, Calcutta, with
the type numbers 9335-9339; the carapace bears the
number 9335.]
Reimann, I. G. 1942. Hamilton phyllocarids in western
New York. Buffalo Society of Natural Sciences, Bul-
letin 1 7(3):48— 5 1 .
Reported (p. 48) the occurrence of Echinocaris punc-
tata (Hall) in the Ludlowville Shale at Wanakah and
in the Ledyard Shale at Alden and Cazenovia Creek.
Described (p. 49) as Echinocaris sp. nov.? a speci-
men, differing from E. punctata by the shape and
spinosity of the border of its carapace, from the Led-
yard Shale at Alden. Reported E. sp., a fragmentary
specimen, from the Ledyard Shale at Alden (p. 49)
and two mandibles of undetermined phyllocarids
from the Wanakah Shale (p. 48). Also noted (p. 48)
that Echinocaris occurs in the Ludlowville along
Lake Erie. Other phyllocarids were also reported.
Roger, J. 1953. Sous-classe des malacostraces (Malacos-
traca Latreille 1806). In Trade de Paleontologie , edited
by J. Piveteau, Tome 3. Onychophores, Arthropodes,
Echinodermes, Stomocordes, p. 309-378. Masson et
cie, Paris.
Diagnosed the family Echinocaridae [sic] and the
genera Echinocaris and Eleutherocaris (p. 312) and
illustrated a specimen of Echinocaris socialis Beecher
from the Devonian of Pennsylvania (PI. I, fig- 5)
[modified from Beecher, 1902],
Rolfe, W. D. I. 1962a. The cuticle of some middle Silurian
ceratiocaridid Crustacea from Scotland. Palaeontol-
ogy 5(2):30— 5 1 .
Noted (p. 48) Clarke’s (1921) figuring of borings,
“Clionolithes ,” on Echinocaris .
1962b. Grosser morphology of the Scottish Silu-
rian phyllocarid crustacean, Ceratiocaris papilio Salter
in Murchison. Journal of Paleontology 36(5): 9 1 2-932.
Noted (p. 917) that some Echinocaris species have
nodes along the hinge line that differ from the tuber-
cles on the valves of the carapace, referring to such
nodes on E. punctata (Hall) and E. socialis ( Beecher).
Also (p. 925) quoted remarks of Beecher (1884) on
the mandible of E. punctata.
1962c. A new phyllocarid crustacean from the
Upper Devonian of Ohio. Breviora 151:1-7.
Noted (p. 4) that the dorsal head of the telson style of
Ohiocaris wycoffi Rolfe is similar to that of Echi-
nocaris sublevis Whitfield and (p. 5) that the smooth
last abdominal segment of O. wycoffi differs from
every species of Echinocaris. Also noted (p. 6) that
the Harvard Museum of Comparative Zoology spec-
imen of O. wycoffi is part of a collection containing
E. multinodosa Whitfield and E. sublevis.
1987
ECHINOCA RIS. AN ANNOTATED BIBLIOGRAPHY
43
1969. Phyllocarida. In Treatise on Invertebrate
Paleontology, edited by R. C. Moore, Part R, Arthrop-
oda 4( 1 ), p. R296-R33 1 . Geological Society of Amer-
ica and University of Kansas Press, Lawrence.
Diagnosed, and corrected, when necessary, the names
of (p. R3 1 7- R3 1 8) the Echinocarididae, the Echino-
caridinae, Echinocaris and related taxa. Commented
on various aspects of phyllocarid biology and ecol-
ogy, including (p. R305) the mandibular palps of
Echinocaris and (p. R303) Beecher’s correlation of
carapace lobes to cephalic appendages, and (p.
R308) the lack of evidence for archaeostracans being
fluviatile or “continental.” Noted (p. R298) that
prismatic structures are found in the integument of
Echinocaris and provided a chart (fig. 122) of mor-
phological features observed in various phyllocarid
genera, including Echinocaris. Illustrated E. socialis
Beecher (fig. 141,1), E. punctata (Hall) (fig.
142,3a&b) [after Hall, 1888, which, in turn, were
modified from Hall, 1863], and E. randalliil Beecher
(fig. 128). Diagnosed (p. R319) Eleutherocaris and
illustrated (fig. 142,4a&b) rubber molds of the holo-
type of Echinocaris (= Eleutherocaris ) whitfieldi
(Clarke).
1981. Phyllocarida and the origin of the Mala-
costraca. Geobios 14:17-27.
Suggested (p. 18) homologies of the carapace grooves
of Echinocaris with those of Montecaris and deca-
pods, illustrating the carapace of Echinocaris multi-
nodosa Whitfield (Fig. 1 A, C, drawn from Sturgeon
et al., 1964). Reported (p. 21) and illustrated (fig. 3;
PI. IB) apertures for limb insertion on abdominal
somites 3 and 4 of Echinocaris punctata (Hall) from
Kashong Glen, New York. Suggested (p. 21) that the
large size, and large mandibles, of archaeostracans
indicate that they were not filter feeders, but were
probably macrophagous.
Rolfe, W. D. I., and R. H. Denison. 1966. The supposed
fish Pseudodontiehthys Skeels, 1962, is the phyllocarid
crustacean Dithyrocaris. Journal of Paleontology
40(0:214-215.
Concluded that mandibles erroneously identified as
belonging to a chondrichthyan by Skeels (1962) “are
identical with the ‘Mandibles of Phyllocarida’ fig-
ured by Hall and Clarke . . . although it is still un-
certain whether these belong to Dithyrocaris (= Me-
sothyra) or to unusually large Echinocaris .” [See
also Dunkle, 1965.]
Rolfe, W. D. I., and V. A. Edwards. 1979. Devonian Ar-
thropoda (Trilobita and Ostracoda excluded). Special
Papers in Palaeontology 23:325-329.
Discussed the stratigraphic value of phyllocarids,
providing (text fig. 2) a chart showing the geologic
age of species of Echinocaris, Eleutherocaris , and
other Devonian phyllocarids, as well as their inferred
phylogenetic links. The range of Echinocaris in the
Devonian was given as earliest Givetian to late Fa-
mennian. Also noted (p. 328) that revision of the
Russian species of Echinocaris is needed.
Rollins, H. B., N. Eldredge, and R. M. Linsley. 1972. Pa-
leontological problems of the Hamilton Group (Mid-
dle Devonian). In Field Trip Guidebook (44th Annual
Meeting, Hamilton) New York State Geological Asso-
ciation, edited by J. McLelland, p. F-l-F-28.
Listed (p. F — 7) Echinocaris sp. from the Solsville
Member, Marcellus Formation, at the Peterbo-
rough South Quarry, near Morrisville, New York,
and E. punctata ( Hall) from the Ludlowville Forma-
tion at the “Pierceville” (=Bradley Brook) Quarry,
Pierceville, New York (p. F— 18) and from the Mos-
cow Formation at the Deep Spring Quarry, Leb-
anon, New York (p. F-26).
Sartenaer, P. 1969. Late Upper Devonian (Famennian)
rhynchonellid brachiopods from Western Canada.
Geological Survey of Canada Bulletin, 169:1-269.
Lists (p. 4) Echinocaris sp. and other taxa as occur-
ing in the Eoparaphorhynchus zone (Lower Fa-
mennian) of western Canada. [The Echinocaris
material was described in Copeland, 1960a as E. cas-
torensis Copeland and E. consanquina Eller.].
Sass, D. B., and R. A. Condrate. 1985. Destruction of a
Late Devonian ophiuroid assemblage: a victim of
changing ecology at the Catskill delta front. In The
Catskill Delta , edited by D. L. Woodrow and W. D.
Sevon. Geological Society of America Special Paper
201:237-246.
Noted (p. 238) that the earliest occupants of the en-
vironment represented by the lower member of the
Upper Devonian Alfred Shale, exposed in the vicin-
ity of Alfred Station, N. Y., may have been malacos-
tracans, for example Echinocaris, and infaunal bur-
rowers. These were succeeded by a more diverse
fauna. Also illustrated phyllocarids [including Echi-
nocaris?\ in a reconstruction (fig. 6) of the sea bot-
tom during “Alfred Station time.”
Schram, F. R. 1986. Crustacea. Oxford University Press,
New York, 606 p.
Referred, after Krestovnikov, 1961, and Chlupaiuc,
to echinocaridines as, “benthic reef-dwelling forms”
(p. 328) and illustrated (fig. 26-4A) Echinocaris so-
cialis Beecher [incorrectly identified as E. punctata
(Hall)] (after Rolfe, 1969). Concurred (p. 329) with
Rolfe (1969) that large mandibles possessed by ar-
chaeostracans may not confirm a carnivorous habit
but that, alternatively, they may have been scav-
engers and detritus feeders.
Schuchert, C. 1943. Stratigraphy of the Eastern and Cen-
tral United States. Wiley, New York, 1013 p.
Included (p. 105) Echinocaris punctata (Hall) in a list
of taxa (after Clarke, 1905) from the Ithaca fauna
(Upper Devonian, Naples Stage) of New York. Re-
44
HANNIBAL AND FELDMANN
No. 42
ported (p. 106) Echinocaris [=Eleutherocaris] whit-
fieldi Clarke, and, after Clarke, 1904, Eleutherocaris
whitfieldi , from the Naples fauna of New York. Also
reported (p. 112, after Caster, 1934) Echinocaris
from the Devonian Conneaut beds of Pennsylvania.
Scott, W. B. 1907. An Introduction to Geology. Second
Edition. Macmillan, New York, 816 p.
Illustrated (PI. IX, fig. 19) Echinocaris punctatus
(Hall) [=£’. punctata ] on a plate illustrating Devo-
nian fossils. [The illustration is a combination of
figs. 1 and 2, PI. 28, in Hall, 1888, with a restored
abdomen added.]
Sepkoski, J. J., Jr. 1982. A compendium of fossil marine
families. Milwaukee Public Museum Contributions in
Biology and Geology , Number 51:1-125.
Noted (p. 60) the range of the Echinocarididae as be-
ing Siegenian through Tournaisian.
Shimer, H. W., and R. R. Shrock. 1944. Index Fossils of
North America. Technology Press, Massachusetts In-
stitute of Technology, Cambridge, 837 p.
Gave a diagnosis (p. 657) of Echinocaris and illus-
trated (PI. 279) E. socialis Beecher (figs. 32, 39) and
E. sublevis Whitfield [spelled incorrectly as E. sub-
laevis in the figure caption] (figs. 37, 38, 41) after
Hall and Clarke, 1888, and E. punctata (Hall) (fig.
42) after Hall and Clarke, 1888, which, in turn, is
modified from Beecher, 1884.
Skeels, M. A. 1962. Two new fishes from the Middle
Devonian Silica Formation, Lucas County, Ohio.
Journal of Paleontology 36(5): 1039- 1046.
Described (p. 1043-1045, text figs. 2A-D, 3), as a
chondrichthyan, Pseudodontichthvs whitei n. gen.
and sp. [This fossil was later identified as represent-
ing the mandible of a phyllocarid. Some authors
have considered it to be Echinocaris. See Dunkle,
1965 and Rolfe and Denison, 1966.]
Smith, B. 1935. Geology and mineral resources of the
Skaneateles quadrangle. New York State Museum Bul-
letin 300:1-120.
Included Echinocaris punctata (WaM) in a list (p. 106)
of conspicuous and characteristic Paleozoic fossils
of the Skaneateles quadrangle having been noted in
the Otisco Member of the Ludlowville Formation in
the platform which underlies the Staghorn Point
submember.
Spinar, Z. 1960. Zaklady Paleontologie Bezobratlych.
Ceskoslovenske akademie ved, Prague, 834 p.
Mentioned (p. 598) Echinocaris as a Paleozoic phyl-
locarid and illustrated (fig. X-85A) Echinocaris so-
cialis Beecher after [modified from] Roger, 1953
[which, in turn, was modified from Beecher, 1902],
Steinmann, G., and Doderlein, L. 1890. Elemente der
Palaontologie. Wilhelm Engelmann, Leipzig, 848 p.
Described (p. 501) Echinocaris and provided a dia-
gramatic illustration (fig. 596) [modified from
Beecher, 1884] of E. punctata (Hall).
Straelen, Van. See Van Straelen.
Stukel, D. J. II. 1986. Ichnology and environmental anal-
ysis of the Chagrin Shale (Famennian), Northeast
Ohio. Geological Society of America Abstracts with
Programs 18(4):326.
Commented on the occurrence of Zoophycos and
Chagrinichnites within the same horizon of, and
throughout, the Chagrin Shale of northeastern Ohio.
Suggested, based on the occurrence of Lingulichnus
and Chagrinichnites that the Chagrin represented a
shallow water environment, deepening to the west.
Stromer von Reichenbach, E. 1909. Lehrbuch der Pa-
laozoologie. I. Teil. Wirbellose Tiere. B. G. Teubner,
Leipzig, 342 p.
Briefly discussed (p. 286) the Archaeostraca, noting
that some forms probably had a double-valved shell
and a spiniform telson. Illustrated (fig. 359) Echino-
caris socialis Beecher, after Beecher, 1902, as an ex-
ample.
Stumm, E. C., and R. B. Chilman. 1969. Phyllocarid crus-
taceans from the Middle Devonian Silica Shale of
northwestern Ohio and southeastern Michigan. Con-
tributions from the Museum of Paleontology, Univer-
sity of Michigan 23(3):53— 7 1 .
Provided (p. 65-66) a short synonymy and a diagno-
sis for Echinocaris. Described E. punctata ( Hall) (p.
66, PI. 1, figs. 1-15; PI. 2, figs. 1 -5) from the Middle
Devonian Silica Shale of Ohio, and reported the spe-
cies as being abundant in the same beds as Rhino-
caris. Included a description (p. 66) of the mandibles of
E. punctata , with illustrations of a specimen from the
Windom Shale of New York ( PI. 2, fig. 6) as well as
the Silica Shale. Also described (p. 54, 60, 63, PI. 7,
figs. 10-15) mandibles, belonging either to Dithyro-
caris or Hebertocaris , some of which have been con-
sidered by other authors (e.g. Dunkle, 1965) to be-
long to, or possibly belong to, Echinocaris.
Sturgeon, M. T., W. J. Hlavin, and R. V. Kesling. 1964.
Rare crustaceans from the Upper Devonian Chagrin
Shale in northern Ohio. Contributions from the Mu-
seum of Paleontology, University of Michigan
19:47-64.
Reported, described or remarked on, and illustrated,
Echinocaris multinodosa Whitfield (p. 49-51, PI. I,
figs. 1 -5; PL II, figs. 1-4; PL V, fig. 10), Echinocaris
sp. cf. E. multinodosa (p. 51, PL V, figs. 7, 8), E. sub-
levis Whitfield (p. 52, PL V, figs. 1, 2), E. pulchra n.
sp. (p. 52-53, PL III, figs. 1 -5), E. ohioensis n. sp. (p.
53-55, PL IV, figs. 1-4), and E. sp. (p. 55-56, PL IV,
fig. 5; PL V, fig. 3). All of the taxa reported were
found in concretions from exposures in the Mill
Creek area, Ashtabula County, Ohio. Echinocaris
multinodosa was also reported from Euclid Creek,
Cuyahoga County, Ohio. A composite diagram (fig.
1 ) of Echinocaris was also presented and some other
fossils found in concretions in the Mill Creek area
1987
ECHINOCA RIS: AN ANNOTATED BIBLIOGRAPHY
45
were also described. [Weidner and Feldmann, 1985,
redescribed some of these taxa, synonymizing E.
pulchra with E. sublevis .]
Szmuc, E. J. 1970. The Devonian System. In Guide to the
Geology of Northeastern Ohio , edited by P. O. Banks
and R. M. Feldmann, p. 9-21. Northern Ohio Geolog-
ical Society.
Noted (p. 13) that the fossil assemblages of the Cha-
grin Shale are, “remarkable in that, except for iso-
lated occurrences of arthropod remains (Echinoca-
ris), they consist almost entirely of brachiopods . . .
and mollusks.”
Tchernychev. See Tschernyshev
Tesmer, I. H. 1975. Geology of Cattaraugus County, New
York. Buffalo Society of Natural Sciences Bulletin
27:1-105.
Illustrated (PI. 5, fig. A) a specimen of Echinocaris,
identified as E. cf. [if.] socialis Beecher, from undif-
ferentiated Canadaway in Freedom Township, Cat-
taraugus County, New York. The right valve of the
carapace and much of the abdomen can be seen in
Tesmer’s photograph.
Tschernyshev, B. I. 1928. Phyllocaridae from the Devo-
nian of the Urals. Annuaire de la Societe Paleontolo-
gique de Russie 7:132-135.
Described and illustrated (p. 132-133; PI. IX, figs.
1-3) Echinocaris arschae n. sp. and (p. 133-134, PI.
IX, figs. 4, 5) Echinocaris n. sp. from the Devonian,
“reef limestone overlying the Stringocephalus Bur-
tini beds on the Arsha River,” in the South Urals.
Briefly compared E. n. sp. to E. randallii Beecher and
noted that the specimens referred to as E. n. sp. may
be young individuals of E. arshae. [In Russian, with
an English summary.]
1933. Arthropods from the Urals and other re-
gions of the U.S.S.R. Paleontology and Stratigraphy
Magazine , Leningrad 1:15-24.
Described Echinocaris tudrensis n. sp. (p. 19, 23, PI.
I, figs. 5-5a) from the Devonian rocks on the right
bank of the M. Tudr River, at Bilovo Village. Com-
pared the new species with E. condylepis Hall and
Clarke, E. punctata (Hall), and E. socialis Beecher.
[In Russian, with an English summary. See also
Tschernyshev, 1941.]
1938. Some Phyllocarida of the Urals and the
north-western district. Materials of the Central Geo-
logical and Prospecting Institute , Leningrad ( Gen. Se-
ries) 3:71-79.
Discussed the occurrences of Echinocaris in the
Urals and northwestern district, postulating that
these phyllocarids lived in a district isolated from
other phyllocarids and that their fossils were al-
lochthonous. Discussed the genus Echinocaris (p.
73), mentioning the American species E. multino-
dosa&nd E. clarke\-E. clarkii ]. Noted (p. 73-74, 78)
the occurrence of E. arschae Tschernyshev (spelled,
incorrectly, E. aschae on p. 73). Described £. uralen-
sis n. sp. (p. 74, 78, PI. I, figs. 4-6) from Sukhoy
Spring, E. sp. 1 (p. 74-75, 78, PI. I, fig. 7) from oppo-
site Terebum Village along the Syas River, E. sp. II
(p. 75, 78, PI. I, fig. 10) from Bogoslovskaya Dacha,
E. sp. Ill (p. 75, 78, PI. I, fig. 1 1) from the Ura-tubin
southern slope, and E. (?) brevicarinata n. sp. (p. 75,
79, PI. I, figs, 8-9) from the Bogoslovskaya Dacha
and the Sungurduk River. [In Russian, with English
summary. Rolfe and Edwards (1979) noted that E.
uralensis and other Russian species were too poorly
illustrated for identification, even to the generic
level.]
1941. Phyllocarida of the Main Devonian Field.
In Fauna of the Main Devonian Field /, edited by R. F.
Gekker, p. 3 1 3-3 17. USSR Academy of Sciences, Pa-
laeontological Institute, Moscow.
Described the genus Echinocaris (p. 315) and de-
scribed E. tudr ensis Tschernyshev, 1933 (p. 315-316,
317, PI. II, fig. lOa&b) from the Bilovo Series (of the
Upper Variegated series, Famennian) near Bilova
and E. sp. (p. 316-317, PI. II, fig. 1 1) from the Pskov
beds (Frasnian) opposite Terebuny. Noted (p. 317)
that phyllocarid remains are extremely scarce in the
Main Devonian Field of the Russian Platform. [In
Russian, with English summary. See also Gekker,
1941.]
Twenhofel, W. H., and R. R. Shrock. 1935. Invertebrate
Paleontology. McGraw-Hill, New York, 51 1 p.
Noted (p. 444) that Echinocaris is a typical genus of
fossil Nebaliacea and illustrated (fig. 163D), after
Beecher, 1902, Echinocaris socialis Beecher.
Van Straelen, V., and G. Schmitz. 1934. Crustacea Phyl-
locarida (=Archaeostraca), pt. 64:1-246. In Eossi-
lium Catalogus 1, Animalia , edited by W. Quenstedt.
W. Junk, Berlin.
Listed (p. 88-94, 95, 179-180, 182-183) most arti-
cles published through 1930 that discussed, men-
tioned, and/or illustrated Echinocaris and species
and specimens that had been referred to the genus,
including Eleutherocaris ( Echinocaris ) whitfieldi
Clarke, “ Ceratiocaris ” longicauda Hall and “ Echi-
nocaris" ( StylonurusI ) [='1 Dunsopterus] wrightia-
nus (Dawson). Also corrected many names. [Many,
but not all, of the corrections are used by us in this
bibliography.] Articles are listed by taxon and syn-
onyms are given.
Vogdes, A. W. 1889. A catalogue of North American Pa-
laeozoic Crustacea confined to the non-trilobite genera
and species. New York Academy of Science Annals
5:1-38.
Provided (p. 16) a diagnosis of Echinocaris after Hall
and Clarke, 1888, and listed (p. 16- 18) articles refer-
ring to the following species of Echinocaris : E. con-
46
HANNIBAL AND FELDMANN
No. 42
dylepis Hall and Clarke, E. multinodosa Whitfield,
E. punctatus Hall (also spelled E. punctata ), E. pus-
tulosa Whitfield, E. socialis Beecher, E. sublevis
Whitfield (also spelled E. sublaevis ), and Echinoca-
ris \= Eleutherocaris] whitfieldi Clarke. Also listed (p.
35) articles referring to Stylonurus [ -1 Dunsopterus ]
wright ianus (Dawson). The names used are those of
the original authors. Entries for some forms referred
to Echinocaris were also included under other ge-
neric names, for instance, Ceratiocaris. Illustrated
(p. 16) Echinocaris \punctata (Hall), modified from
Hall and Clarke, 1888, which, in turn, was modified
from Beecher, 1884],
1890. A bibliography of Paleozoic Crustacea
from 1698 to 1889. U. S. Geological Survey Bulletin
63:1-177.
Diagnosed (p. 16) Echinocaris. Listed (p. 16-18) ar-
ticles referring to Echinocaris condylepis Hall and
Clarke, E. multinodosa Whitfield, E. (= Ceratiocaris ?
= Ceratiocaris - Ceratiocaris [Aristozoe]) punctatus
(Hall) ( -E . armatus) [=£. punctata ], E. pustulosa
Whitfield, E. socialis Beecher, E. sublevis Whitfield
( -E . sublaevis ), and E. whitfieldi Clarke. Also listed
articles (p. 19, 35) referring to Stylonurus (-Equise-
tides = Echinocaris = ? Echinocaris) wrightianus
(Dawson).
1893. A classed and annotated bibliography of
the Palaeozoic Crustacea, 1698-1892, to which is
added a catalogue of North American species. Occa-
sional Papers of the California Academy of Sciences
4:1-412.
In an annotated list of works arranged alphabetically
by author, noted species referred to Echinocaris (at
one time or another) as being included in Beecher,
1884 (p. 18), Clarke, 1885 (p. 37), Hall, 1863 (p. 84),
Hall, 1876 (p. 87), Hall and Clarke, 1888 (p. 88-89),
Jones and Woodward, 1884 (p. 110), Etheridge,
Woodward and Jones, 1885 [=1886] (p. 112) and
1889 [=1891] (p. 119), Packard, 1883 [= 1882] (p. 171),
Whitfield, 1880 (p. 235), and Woodward and Jones
[=Jones and Woodward], 1884, (p. 250). Also listed,
in a separate list arranged alphabetically by taxon,
articles mentioning E. condylepsis [ -E . condylepis ]
Hall and Clarke, E. multinodosa Whitfield, E. ( Cer-
atiocaris) punctatus ( Hall) [-E. punctata ], E. pustu-
losa Whitfield, E. socialis Beecher, E. sublevis Whit-
field (also spelled E. sublaevis), E. [-Eleutherocaris]
whitfieldi Clarke, E. wrightiana (Dawson) (p.
383-384), Equisetides [-1 Dunsopterus] wrightiana
(p. 385) and Stylonurus ( Equisetides ) [=? Dunsopte-
rus] wrightianus (Dawson) (p. 407). Synonyms were
supplied.
1917. Palaeozoic Crustacea. The publications
and notes on the genera and species during the past
twenty years, 1895-1917. Transactions, San Diego So-
ciety of Natural History 3( 1 ): 1 — 1 4 1.
Listed species referred to Echinocaris mentioned in
various articles as follows: Echinocaris punctata
(Hall) (p. 24) in Cleland, 1903; E. whidbornei Jones
and Woodward (p. 46) in Jones and Woodward,
1899; E. socialis Beecher (p. 66) in Partridge, 1902
[not 1912]; E. sublevis Whitfield, E. pustulosa Whit-
field and E. multinodosa Whitfield (p. 135) in Whit-
field, 1893; and Stylonurus [=1 Dunsopterus] wright-
ianus (Dawson) in Clarke and Ruedemann, 1912.
1925. Palaeozoic Crustacea. Part I, A bibliog-
raphy of Palaeozoic Crustacea, supplementing the au-
thor’s previous papers on the same subject. Transac-
tions of the San Diego Society of Natural History
4:5-88.'
Abstracted (p. 48) Reed’s 1909 description of Echi-
nocaris asiatica Reed and noted (p. 88) the inclusion
of Echinocaris in the 5th edition of Zittel’s Grund-
zuge der Palaeontologie.
Waterston, C. D. 1 968. Further observations on the Scot-
tish Carboniferous eurypterids. Transactions of the
Royal Society of Edinburgh 68:1-20.
Noted (p. 12) the similarity of the massive prosomal
appendages associated with Dunsopterus stevensoni
(R. Etheridge, Jr.) to “the controversial specimen
. . . still known as Stylonurus (?) wrightianus
(Dawson)” and, on this basis, placed (p. 18) the ge-
nus Dunsopterus in the Stylonuracea. Placed (Table
1) S’. (?) wrightianus questionably in Dunsopterus as
Dunsopterus (?) wrightianus (Dawson) and that ge-
nus tentatively in the Stylonuridae. [See also Daw-
son, 1881b.]
Weidner, W. E.,and R. M. Feldmann. 1983. Paleoecolog-
ical interpretation of echinocarid arthropod assem-
blages in the Late Devonian Chagrin Shale, northeast
Ohio. Geological Society of America Abstracts with
Programs 15(4):248-249.
[See Weidner and Feldmann, 1985.]
1985. Paleoecological interpretation of echino-
carid arthropod assemblages in the Late Devonian
(Famennian) Chagrin Shale, northeastern Ohio.
Journal of Paleontology 59(4):986— 1 004.
Described five echinocarid-bearing concretionary
horizons from two exposures of the Chagrin Shale
(Famennian) in Lake and Ashtabula counties, Ohio.
Remarked on the genus Echinocaris (p. 989) and on
E. multinodosa Whitfield (p. 989, figs. 3. 5-3.9), E.
ohioensis Sturgeon, Hlavinand Kesling(p. 989-990,
figs. 4.9-4.11), E. sublevis Whitfield (p. 990-992,
figs. 3. 1-3.4) and E. sp. (p. 992-995, figs. 4. 1-4.3).
Considered E. pulchra Sturgeon, Hlavin and Kes-
ling, 1964, to be a synonym of E. sublevis Whitfield,
1880, providing a synonymy of E. sublevis. Noted
the occurrence of a presumed echinocarid trace,
Chagrinichnites osgoodi (p. 995-996, figs. 5. 1-5.2)
from along the Chagrin River. Divided the upper
1987
EC H IN OCA RIS\ AN ANNOTATED BIBLIOGRAPHY
47
portion of the Chagrin Shale into two ichnofacies,
the Lingulichnus ichnofacies and the Chagrinich-
nites ichnofacies. Discussed the sedimentological
setting of Echinocaris in the Chagrin Shale and
characterized the crustacean (p. 1002-1003) as an
epifaunal detritus feeding arthropod, “well adapted
to the turbulent, periodically nutrient-poor . . .
offshore environment, owing to their mobility and
success in utilizing a broad range of food.”
Whidborne, G. F. 1896a. A preliminary synopsis of the
fauna of the Pickwell Down, Baggy, and Pilton beds.
Proceedings of the Geologist’s Association 14(9):
371-377.
Listed (p. 371) Echinocaris whidbornei Jones and
Woodward as occurring in the Marwood, Baggy and
Sloly beds. [See Whidborne, 1896b.]
1896b. A monograph of the Devonian fauna of
the south of England. Vol. 3, Part 1, The fauna of the
Marwood and Pilton beds of North Devon and
Somerset. London, Palaeontographical Society, 1 1 2 p.
Redescribed (p. 6-7) Echinocaris whidbornei Jones
and Woodward, 1889. Cited two specimens of the
species, the type specimen (PI. I, fig. 3) from the
Marwood series, near Sloly and another, fragmen-
tary, specimen from Pilton (in the Porter Collec-
tion).
Whitfield, R. P. 1880. Notice of new forms of fossil crus-
taceans from the Upper Devonian rocks of Ohio, with
descriptions of new genera and species. American Jour-
nal of Science , Third Series, 19:33-42.
Proposed (p. 33-34) the name Echinocaris for Cera-
tiocaris punctatus Hall, 1863, and three newly de-
scribed species from Ohio. Provided a diagnosis of
the genus (p. 34) and compared it to Ceratiocaris and
other genera then considered members of the Cerati-
ocaridae (p. 34-36). Described E. sublevis n. sp. (p.
36-37), designating it (p. 34) as the type species of the
genus, E. pustulosa n. sp. (p. 38), and E. multinodosa
n. sp. (p. 38-39) from calcareous concretions [the
concretions are actually phosphatic; there is little
calcareous material in them] in the Erie Shale
[=Chagrin Shale, Famennian] at Leroy, Ohio. Also
described the decapod Palaeopalaemon newberryi n.
gen. and sp. (p. 39-42), found associated with Echi-
nocaris. [Although this work was published without
illustrations, plates were evidently supplied with the
author’s special edition (fide Vogdes, 1893, p. 383)
and other authors copied some of the illustrations on
these plates. Feldmann and McKenzie ( 198 1, p. 386)
suggested that Indian Point, Lake County, Ohio,
may be Whitfield’s type locality. However Indian
Point is actually in Perry Township, whereas New-
berry ( 1873) indicated that the specimens were found
in northern Leroy township. The type locality is
probably somewhere along Paine’s Creek upstream
of Indian Point, perhaps at Hell Hollow (Hell Hole),
which is just to the east of the present town of Leroy
Center. Both Indian Point and Hell Hollow are Lake
County metropolitan parks. Permits are necessary to
collect fossils in them.]
1890. Contributions to invertebrate paleontol-
ogy [of Ohio], New York Academy of Science Annals
5:505-620.
Described Echinocaris. Reported and described (p.
560-573) various fossils, including E. sublevis Whit-
field (p. 565-567, PI. XII, figs. 12-14), E. pustulosa
Whitfield (p. 567, PI. XII, fig. 15), and E. multino-
dosa Whitfield (p. 568, PI. XII, fig. 16), from the Erie
Shales [=Chagrin Shale, Famennian] at Leroy, Ohio.
[Pages 562-572 of this article are essentially a reprint
of Whitfield, 1880, which was published without il-
lustrations. The specimens of Echinocaris described
and/ or figured in this article are now in the Ameri-
can Museum of Natural History, Columbia Univer-
sity Collection, numbered 12281 and 12282 (E. sub-
levis), 5512G (E. pustulosa ), and 551 1G, 12278 and
12280 (E. multinodosa ).]
1892. Discovery of a second example of the mac-
rouran decapod crustacean, Palaeopalaemon newber-
ryi. The American Geologist 9:237-238.
Described a specimen of Palaeopalaemon newberryi
that had been sent to Whitfield “under the name
Echinocaris sp.”
1 893. Contributions to the paleontology of Ohio.
Report of the Geological Survey of Ohio 7(2):407— 494.
The sections dealing with or mentioning Echinoca-
ris, “Fossils from the Erie Shales” (p. 452-462) and
“Species from the Huron Shales” (p. 462-464) are
reprints of p. 560-572 and p. 573-575 of Whitfield,
1890. The plate, PI. VIII, is a reproduction of PI. XII
in Whitfield, 1890; both plate numbers are indicated
at the top of the plate.
1 899. List of fossils, types and figured specimens,
used in the palaeontological work of R. P. Whitfield,
showing where they are probably to be found at the
present time. New York Academy of Sciences Annals
12:139-186.
Listed (p. 182) the type specimens of Echinocaris
multinodosa Whitfield, E. pustulosa Whitfield, and
E. sublaevis Whitfield [ =E . sublevis], as published in
Whitfield, 1890, as being located at Columbia Col-
lege. [These specimens are now at the American Mu-
seum of Natural History; see Whitfield, 1890.]
Willard, B. 1932. Devonian faunas in Pennsylvania. Penn-
sylvania Geological Survey, Bulletin, Fourth series,
G4: 1-43.
Reported Echinocaris sp. in a faunule 3.8 miles east
of Wellsboro, Tioga County (p. 28), E. cf. E. sublae-
vis [ -E . cf. E. sublevis ] Whitfield in a faunule found
in the Hollenback region, Bradford County (p. 31,
34), and E. cf. [£.] multinodosa Whitfield in a faunule
found in Monroe County (p. 37, 40), Pennsylvania.
48
HANNIBAL AND FELDMANN
No. 42
The age of the first two faunules is given as Che-
mung; the age of the third as Ithaca. The first and
third faunules are composed of marine elements, the
second, found in redbeds and associated strata, is
composed of marine and nonmarine elements, in-
cluding the fish Bothriolepis.
1935. Hamilton Group of central Pennsylvania.
Geological Society of America Bulletin 46: 195-224.
Listed ( p. 2 1 1 ) Echinocaris punctata ( Hall) as occur-
ring very rarely and E. sp. nov. as occurring rarely in
the Ludlowville faunal facies of the Mahantango
Formation of the Hamilton Group in Pennsylvania.
Willard, B., F. M. Swartz, and A. B. Cleaves. 1939. The
Devonian of Pennsylvania. Pennsylvania Geological
Survey, Bulletin , Fourth Series, G 19: 1-481.
Reported (p. 177) Echinocarisl from the Moscow
beds (Hamilton) north of Auburn, Pennsylvania.
Listed E. punctata (Hall) (p. 190, PI. 31, fig. 26) as
occurring, very rarely, and E. sp. nov. as occurring,
rarely, in the Ludlowville portion of the Mahan-
tango Formation (Hamilton). Listed (p. 213) Echi-
nocaris (?) sp. nov. as occurring in the Trimmers
Rock fauna (Portage Group) of south-central
Schuylkill County. Also, illustrated E. multinodosa
Whitfield (PI. 31, fig. 25) [probably -E. cf. E. multi-
nodosa of Willard, 1932] from the Trimmers Rock
Sandstone in Monroe County and E. sublaevis
Whitfield [=£■. sublevis , probably E. cf. E. sublaevis
in Willard, 1932] from the Canadaway in Bradford
County. [The specimens are too poorly illustrated to
judge the accuracy of the identifications of the echi-
nocaridids, even to the generic level.]
Williams, H. S., and E. M. Kindle. 1905. Contributions to
Devonian paleontology, 1903. Part 1. Fossil faunas of
the Devonian and Mississippian (“Lower Carbonifer-
ous”) of Virginia, West Virginia, and Kentucky. United
States Geological Survey, Bulletin 244, Part 1 :9-58.
Reported (p. 37; chart facing p. 55) the rare occur-
rence of Echinocaris sp. in a Devonian (Chemung)
faunule near White Sulphur Springs in southeastern
West Virginia. [The whereabouts of the document-
ing specimen(s) is unknown.]
Woods, H. 1909. Palaeontology: Invertebrate. Fourth
edition. The University Press, Cambridge, England,
388 p.
Mentioned (p. 323) that Echinocaris is found in the
Devonian and (p. 322) that the Leptostraca (Phyllo-
carida) “are all marine, and live mainly in shallow
water or at moderate depths.”
Woodward, H. P. 1943. Devonian System of West Vir-
ginia. West Virginia Geological Survey [ Report] 15:1-
655.
Listed (p. 366) Echinocaris punctata (Hall) as occur-
ring in Hamilton rocks elsewhere than West Vir-
ginia.
Wright, B. H. 1884. Notes on the geology of Yates
County, N.Y. Thirty-fifth Annual Report on the New
York State Museum of Natural History 35:195-206.
Described (p. 196-197), after Dawson, 1881b, and
illustrated (PI. XV, figs. 1-3, figs. 1 and 3 after [mod-
ified from] Dawson, 1881b) Equisetides wrightiana
Dawson \=r!Dunsopterus wrightianus (Dawson)].
[See also Hall, 1884.]
Zell, P. D. 1985. Paleoecology and stratigraphy of the
Middle Devonian Moscow Formation in the Che-
nango Valley, New York. Unpublished M. S. thesis,
University of Pittsburgh, 128 p.
Found phyllocarids, particularly Rhinocaris colum-
bina Clarke and Echinocaris punctata (Hall) to be
rare but persistent elements in the Moscow fauna of
the Chenango Valley (p. 55) with the latter, “more
common in the coarse siltstones in the upper, shal-
lower, portions of cycles.” Reported E. punctata as
occurring in a Devonochonetes- Mucrospirifer com-
munity (p. 44, Table 6) and a Tropidoleptus commu-
nity (p. 50, Table 8). Listed E. punctata (p. 83, 1 18,
122) from six localities in the Chenango Valley.
Commented on the life habits of E. punctata (p.
55-56), finding it to be, “an epifaunal scavenger/
predator on silty substrates with currents of moder-
ate energy.” Reported (p. 56) the epizoans IConcho-
trema (Appendix D, PI. 2.4) and Orbiculoidea doria
(Appendix D, PI. 1.3) associated with the carapace
of E. punctata.
Zittel, K. A. 1880-1885. Handbuch der Palaeontologie.
Part I. Palaeozoologie. Vol. II. Mollusca und Ar-
thropoda. R. Oldenbourg, Munich, 893 p.
Described (p. 658) Echinocaris and illustrated (fig.
846) E. punctata (Hall), Beecher, 1884. [The illustra-
tion is reversed. This same picture is reproduced in
the following Zittel entries.]
1887. Traite de Paleontologie. Tome II. Paleo-
zoologie. Partie I. Mollusca et Arthropoda. Octave
Doin Editeur, Paris, 897 p.
Described (p. 655) Echinocaris and illustrated (fig.
863) E. punctata ( Hall), [after an earlier Zittel which,
in turn, was] after [modified from] Beecher [1884],
[The illustration is reversed.] Echinocaris is included
in a table (p. 658) listing the geologic distribution of
phyllocarids.
1900. Text-book of Palaeontology .Vol. 1, edited
by C. R. Eastman. Macmillan and Co., New York,
706 p.
J. M. Clarke, in this work, named and diagnosed
the Echinocaridae [j/c] (p. 655) [see Chlupdc, 1963],
and diagnosed Echinocaris (p. 655-6) and p. 656)
Eleutherocaris [the latter for the first time, but
without an illustration or a mention of any species be-
longing to it; see Clarke, 1902.] Illustrated (fig. 1373)
1987
ECHINOCARIS : AN ANNOTATED BIBLIOGRAPHY
49
Echinocaris punctata (Hall), [after an earlier Zittel
which, in turn, was] after [modified from] Beecher,
1884. [The illustration is reversed.] Also illustrated
“gastric teeth” [mandibles] of E. punctata (fig.
1369a&b) [modified from Hall and Clarke, 1888,
which, in turn, were modified from Beecher, 1884],
1903. Grundziige der Palaontologie (Palaozool-
ogie). Second edition. I. Abteilung: Invertebrata. R.
Oldenbourg, Munich, 558 p.
Briefly described (p. 515) Echinocaris and illustrated
(fig. 1319) E. punctata (Hall), [after an earlier Zittel
which, in turn, was] after [modified from] Beecher,
1884. [The illustration is reversed.]
1913. Text-book of Paleontology, V ol. 1, Second
Edition, edited by C. R. Eastman. Macmillan and Co.,
London, 839 p.
J. M. Clarke, in this work, diagnosed (p. 751) the
Echinocaridae [szc], Echinocaris , and (p. 752) Eleuth-
erocaris. Illustrated (fig. 1455) E. punctata (Hall),
[after an earlier Zittel which, in turn, was] after
[modified from] Beecher, [1884], [The illustration is
reversed.] Also illustrated “gastric teeth” [mandi-
bles] of Echinocaris punctata (Figure 1451) [modi-
fied from Hall and Clarke, 1888], [At least one re-
print edition exists with a later date of publication;
that one is dated 1937.]
1915. Grundziige der Palaontologie (Palaozool-
ogie). Neubearbeitet von Ferdinand Broili. 1. Abtei-
lung: Invertebrata. R. Oldenbourg, Munich, 694 p.
Diagnosed the Echinocaridae [sz'r] and briefly de-
scribed Echinocaris ( p. 626). Illustrated (fig. 1368) E.
punctata (Hall), [after an earlier Zittel which, in
turn, was] after [modified from] Beecher, [1884],
[The illustration is reversed.]
1924. Grundziige der Palaontologie (Palaozool-
ogie). Neubearbeitet von Ferdinand Broili. Vol. 1, In-
vertebrata. Sixth and revised edition. R. Oldenbourg,
Munich, 733 p.
Diagnosed the Echinocaridae [szc] and Echinocaris
(p. 660) and illustrated (fig. 1 377) E. punctata ( Hall),
[after an earlier Zittel which, in turn, was] after
[modified from] Beecher [1884], [The illustration is
reversed.] Mentioned the presence (p. 659) of “eye
tubercles” on Echinocaris.
Index to Generic and Species Taxa Cited
The index lists citations, by author and date, for all ref-
erences to generic and species taxa of fossil arthropods
(excluding trilobites) referred to, or cited as being illus-
trated, in annotations [but not our bracketed comments]
in the bibliography. However, no citations to the genus
Echinocaris itself are listed, as such a list would be too
long to be useful.
In addition to the references to taxa assigned, at one
time or another, to Echinocaris , the index also lists cit-
ations to other arthropod taxa made in the original litera-
ture and mentioned in the annotations. The references to
these arthropods are not exhaustive but do, in many
cases, provide an entry into the literature on these forms.
The combinations and spellings presented in the index
are those of the original author and do not reflect the most
recent combinations or the correct spelling of the taxon,
necessarily. Current combinations, correctly spelled, are
indicated in the index by boldface type.
Aristozoe: Giirich, 1929; Hall and Clarke, 1888; Oehlert,
1889.
Callizoe: Oehlert, 1889.
Ceratiocaris: Giirich, 1929; Newberry, 1873; Vodges,
1889; Whitfield, 1880.
Ceratiocaris?: Carll, 1883: Claypole, 1903.
Ceratiocaris sp.: Claypole, 1903.
Ceratiocaris (Echinocaris?): Clarke, 1904.
Ceratiocaris armata Hall: Bigsby, 1878.
Ceratiocaris armatus Hall: Hall, 1863; Hall, 1876; Jones,
1884; Jones and Woodward, 1884; Packard, 1883.
Ceratiocaris beecheri Clarke: Clarke, 1885b; Clarke,
1892.
Ceratiocaris (?) beecheri Clarke: Chadwick, 1935; Clarke,
1892.
Ceratiocaris longicauda Hall: Giirich, 1924.
? Ceratiocaris longicauda Hall: Clarke, 1892; Hall and
Clarke, 1888.
“Ceratiocaris” longicauda Hall: Van Straelen and
Schmitz, 1934.
Ceratiocaris longicaudus Hall: Hall, 1863.
Ceratiocaris punctata Hall: Bigsby, 1878.
Ceratiocaris punctatus Hall: Vogdes, 1889; Whitfield,
1880.
Ceratiocaris ? punctatus Hall: Hall, 1863.
Ceratiocaris (Aristozoe) punctatus Hall: Hall, 1876.
Ceratiocaris whitfieldi Clarke: Clarke, 1902.
Chagrinichnites: Frey, Curran, and Pemberton, 1984;
Stukel, 1986; Weidner and Feldmann, 1983.
Chagrinichnites brooksi Feldmann et al.: Feldmann et
al„ 1978.
Chagrinichnites osgoodi Hannibal and Feldmann: Han-
nibal and Feldmann, 1983; Weidner and Feldmann,
1983.
Dithyrocaris: Stumm and Chilman, 1969.
Dithyrocaris (= Mesothvra): Rolfe and Denison, 1966.
Dunsopterus: Waterston, 1968.
Dunsopterus stevensoni (R. Etheridge, Jr.): Waterston,
1968.
Dunsopterus (?) wrightiamus (Dawson): Waterston, 1968.
Echinocaris (- Ceratiocaris [sic]) sp.: Monroe and Teller,
1899.
Echinocaris sp.: Allan, 1935; Baird, 1978; Beecher, 1884;
50
HANNIBAL AND FELDMANN
No. 42
Bolton, 1966; Buehler and Tesmer, 1963; Copeland,
1960a; Copeland and Bolton, 1985; Edmonds, Wil-
liams, and Taylor, 1979; Feldmann, Boswell, and
Rammer, 1986; Gekker, 1941; Goldring, 1971; Hlavin,
1976; Krestovnikov, 1961; McLaren, 1955; McLaren,
1963; Reimann, 1942; Rollins, Eldredge, and Linsley,
1972; Sartenaer, 1969; Sturgeon, Hlavin, and Kesling,
1964; Tschernyshev, 1928; Tschernyshev, 1938; Whit-
field, 1892; Willard, 1932; Willard, Swartz, and
Cleaves, 1939; Williams and Kindle, 1905.
Echinocaris? sp.: Fisher, 1951; Willard, Swartz, and
Cleaves, 1939.
Echinocaris archae Tschernyshev: Krestovnikov, 1961.
Echinocaris armata (Hall): Etheridge, Woodward, and
Jones, 1889; Jones and Woodward, 1884.
Echinocaris armatus (Hall): Packard, 1883.
Echinocaris arschae Tschernyshev: Tschernyshev, 1928;
Tschernyshev, 1938.
Echinocaris asiatica Reed: Chhibber, 1934; Feldmann,
Hannibal, and Babcock, 1986; La Touche, 1913; Pas-
coe, 1959; Reed, 1908; Vogdes, 1925.
Echinocaris asiaticus Reed: Feldmann, Hannibal, and
Babcock, 1986.
Echinocaris auricula Eller: Copeland, 1960a; Eller, 1935;
Feldmann, Boswell, and Kammer, 1986; Hannibal and
Feldmann, 1985.
Echinocaris? beecheri Clarke (nomen nudum): Clarke,
1891; Clarke, 1898a; Dana, 1895.
Echinocaris beecheri Copeland: Bolton, 1966; Copeland,
1960a.
Echinocaris? brevicarinata Tschernyshev: Krestovnikov,
1961; Tschernyshev, 1938.
Echinocaris castorensis Copeland: Bolton, 1966; Cope-
land, 1960a; Hannibal and Feldmann, 1985.
Echinocaris clarkei Beecher: Grabau and Shimer, 1910;
Gurich, 1929.
Echinocaris clarkii Beecher: Beecher, 1902; Caster, 1930;
Chadwick, 1935.
Echinocaris clarkii? Beecher: Castor, 1934.
Echinocaris condylepis Hall and Clarke: Chadwick, 1 935;
Clarke, 1892; Clarke and Ruedemann, 1903; Cope-
land, 1960a; Eller, 1935; Eller, 1937; Gurich, 1929; Hall
and Clarke, 1888; Miller, 1889; Tschernyshev, 1933;
Vogdes, 1889; Vogdes, 1890.
Echinocaris condylepsis Hall and Clarke: Vogdes, 1890;
Vogdes, 1893.
Echinocaris consanguina Eller: Bolton, 1966; Copeland,
1960a; Eller, 1935.
Echinocaris crosbyensis Eller: Eller, 1937; Jux, 1960.
Echinocaris longicauda (Hall): Beecher, 1884; Miller,
1889.
Echinocaris ? longicauda Hall: Clarke, 1904; Clarke and
Luther, 1904.
Echinocaris longicauda Hall (Ceratiocaris longicauda)-.
Miller, 1889.
Echinocaris multinodosa Whitfield: Beecher, 1884;
Clarke, 1892; Clarke, 1904; Clarke and Ruedemann,
1903; Feldmannand McKenzie, 1981; Feldmann et al.,
1978; Grabau and Shimer, 1910; Gurich, 1929; Hall
and Clarke, 1888; Hoover, 1960; Miller, 1889; Rolfe,
1962c; Rolfe, 1981; Sturgeon, Hlavin, and Kesling,
1964; Tschernyshev, 1938; Vogdes, 1889; Vogdes, 1890;
Vogdes, 1893; Vogdes, 1917; Weidner and Feldmann,
1983; Whitfeld, 1880; Whitfield, 1890; Whitfield, 1899;
Willard, Swartz and Cleaves, 1939.
Echinocaris cf. E. multinodosa Whitfield: Willard, 1932.
Echinocaris sp. cf. multinodosa Whitfield: Sturgeon,
Hlavin and Kesling, 1964.
Echinocaris multinodosus Whitfield: Packard, 1882; Pack-
ard, 1883.
Echinocaris multispinosis Feldmann and McKenzie, 1981:
Feldmann and McKenzie, 1981.
Echinocaris ohioensis Sturgeon, Hlavin, and Kesling:
Feldmannand McKenzie, 1981; Feldmann etal., 1978;
Sturgeon, Hlavin, and Kesling, 1964; Weidner and
Feldmann, 1983.
Echinocaris pulchra Sturgeon, Hlavin, and Kesling:
Feldmannand McKenzie, 1981; Feldmann et al., 1978;
Sturgeon, Hlavin and Kesling, 1964; Weidner and
Feldmann, 1985.
Echinocaris pulcra Sturgeon, Hlavin, and Kesling: Feld-
mann and McKenzie, 1981.
Echinocaris punctata (Hall): Beecher, 1884; Bernard,
1895; Bigsby, 1878; Brooks, 1957; Brown, 1956; Buehler
and Tesmer, 1963; Chadwick, 1935; Chamberlin and
Salisbury, 1905; Chamberlin and Salisbury, 1909;
Clarke, 1892; Clarke, 1902; Clarke, 1905; Clarke, 1921;
Clarke and Ruedemann, 1903; Cleland, 1903; Cleland,
1911; Copeland and Bolton, 1985; Dacque, 1921;
Dana, 1895; Dunkle, 1965; Eller, 1935; Etheridge,
Woodward, and Jones, 1889; Feldmann and Hannibal,
1985a; Fenton and Fenton, 1958; Goldring, 1929; Gra-
bau, 1921; Grabau and Shimer, 1910; Grasso, 1981;
Gurich, 1929; Hall and Clarke, 1888; Jones and
Woodward, 1889; Jux, 1960; Kesling and Chilman,
1975; Krestovnikov, 1960; La Touche, 1913; Lesley,
1889-1890; Miller, 1889; Murphy, 1972; Olsson, 1912;
Reimann, 1942; Rolfe, 1962b; Rolfe, 1969; Rolfe, 1981;
Rollins, Eldredge, and Linsley, 1972; Schram, 1986;
Schuchert, 1943; Shimer and Shrock, 1944; Smith,
1935; Steinmann and Doderlein, 1890; Stumm and
Chilman, 1969; Tschernyshev, 1933; Vogdes, 1889;
Vogdes, 1917; Willard, 1935; Willard, Swartz, and
Cleaves, 1939; Woodward, 1943; Zell, 1985; Zittel,
1880-1885; Zittel, 1887; Zittel, 1900; Zittel, 1903; Zit-
tel, 1913; Zittel, 1915; Zittel, 1924.
Echinocaris cf. punctata (Hall): Prosser, 1898.
Echinocaris punctatus (Hall): Packard, 1882; Packard,
1883; Scott, 1907; Vogdes, 1889.
Echinocaris ( Ceratiocaris ) punctatus (Hall): Vogdes,
1893.
Echinocaris (= Ceratiocaris ? = Ceratiocaris - Ceratiocaris
1987
ECHINOCA RIS: AN ANNOTATED BIBLIOGRAPHY
51
( Aristozoe )) punctatus (Hall) (= armatus ): Vogdes,
1890.
Echinocaris pustulosa Whitfield: Beecher, 1884; Clarke,
1892; Clarke, 1904; Clarke and Ruedemann, 1903;
Grabau and Shimer, 1910; Gtirich, 1929; Hall and
Clarke, 1888; Hoover, 1960; Jones and Woodward,
1884; Miller, 1889; Vogdes, 1889; Vogdes, 1890;
Vogdes, 1893; Vogdes, 1917; Whitfield, 1880; Whit-
field, 1890; Whitfield, 1899.
Echinocaris pustulosus (Hall): Copeland and Bolton,
1985.
Echinocaris pustulosus Whitfield: Packard, 1883.
Echinocaris randalii Beecher: Tschernyshev, 1928; Kres-
tovnikov, 1961 .
Echinocaris randalii Beecher: Caster, 1930; Caster, 1934;
Chadwick, 1935; Copeland, 1960a.
Echinocaris randallii Beecher: Beecher, 1902; Gurich,
1929.
Echinocaris randalliil Beecher: Rolfe, 1969.
Echinocaris sloliensis Coomarasway: Goldring, 1971.
Echinocaris sloliensis Partridge: Butler, 1980; Edmonds,
Williams and Taylor, 1979; Partridge, 1902.
Echinocaris sloliensis Partridge: Butler, 1980; Edmonds,
Williams and Taylor, 1979; Partridge, 1902.
Echinocaris sociales Beecher: Eller, 1935.
Echinocaris socialis Beecher: Beecher, 1884; Beecher,
1902; Case, 1982; Caster, 1930; Caster, 1934; Chad-
wick, 1935; Clarke, 1892; Clarke and Ruedemann,
1903; Dana, 1895; Easton, 1960; Eller, 1935; Eller,
1937, Grabau and Shimer, 1910; Gurich, 1929; Hall
and Clarke, 1888; Jux, 1960; Lesley, 1889-1890; Miller,
1889; Moore, Lalicker, and Fischer, 1952; Muller,
1963; Partridge, 1902; Roger, 1953; Rolfe 1962b;
Rolfe, 1969; Schram, 1986; Shimer and Shrock, 1944;
Spinar, 1960; Tschernyshev, 1933; Twenhofel and
Shrock, 1935; Vogdes, 1889; Vogdes, 1890; Vogdes,
1893; Vogdes, 1917.
Echinocaris socialis? Beecher: Castor, 1934.
Echinocaris cf. socialis Beecher: Tesmer, 1975.
Echinocaris sublaevis Whitfield: Clarke, 1892; Clarke,
1904; Clarke and Ruedemann, 1903; Eller, 1935; Feld-
mann and McKenzie, 1981; Grabau and Shimer, 1910;
Gurich, 1929; Hall and Clarke, 1888; Jones and
Woodward, 1884; Miller, 1889; Shimer and Shrock,
1944; Vogdes, 1889; Vogdes, 1893; Whitfield, 1899;
Willard, Swartz, and Cleaves, 1939.
Echinocaris cf. E. sublaev is Whitfield: Willard, 1932.
Echinocaris sublevis Whitfield: Beecher 1884; Feldmann
et al., 1978; Gurich, 1929; Hoover, 1960; Miller, 1889;
Packard, 1882; Packard, 1883; Rolfe, 1962c; Sturgeon,
Hlavin and Kesling, 1964; Vogdes, 1889; Vogdes, 1890;
Vogdes, 1893; Vogdes, 1917; Weidner and Feldmann,
1983; Whitfield, 1880; Whitfield, 1890.
Echinocaris tudrensis Tschernyshev: Gekker, 1941; Gek-
ker, 1983; Krestovnikov, 1961; Tschernyshev, 1933;
Tschernyshev, 1941.
Echinocaris turgida Eller: Eller, 1935; Eller, 1937.
Echinocaris uralensis Tschernyshev: Krestovnikov, 1961;
Tschernyshev, 1938.
Echinocaris whidbornei Jones and Woodward: Butler,
1980; Cleevely 1983; Copeland, 1960a; Edmonds, Wil-
liams, and Taylor, 1979; Eller, 1937; Goldring, 1971;
Gurich, 1929; Herries, 1896; Jones, 1898a; Jones,
1898b; Jones, 1900; Jones and Woodward, 1889; Jones
and Woodward, 1899; Morris, 1980; Partridge, 1902;
Vogdes, 1917; Whidborne, 1896a; Whidborne, 1896b.
Echinocaris whitfieldi Clarke: Clarke, 1885a; Clarke,
1885b; Clarke, 1891; Clarke, 1892; Clarke, 1898a;
Clarke, 1904; Dana, 1895; Hall and Clarke, 1888; Les-
ley, 1889-1890; Miller, 1889; Schuchert, 1943; Vogdes,
1889; Vogdes, 1890; Vogdes, 1893.
Echinocaris (= Eleutherocaris) whitfieldi Clarke: Rolfe,
1969.
Echinocaris wrightana (Dawson): Miller, 1889.
Echinocaris wrightiana (Dawson): Jones and Woodward,
1884; O’Connell, 1916; Vogdes, 1893.
“ Echinocaris ” wrightiana ( Dawson ){- Stylonurus ? wrigh-
tianus ): Copeland, 1960a.
"Echinocaris" ( Stylonurus ?) wrightianus (Dawson): Van
Straelen and Schmitz, 1934.
Echinocarys: Barrois, 1891.
Eleutherocaris: Clarke, 1902; Copeland, 1960b; Grabau
and Shimer, 1910; Gurich, 1929; Krestovnikov, 1960;
Roger, 1953; Rolfe, 1969; Rolfe and Edwards, 1979;
Zittel, 1900; Zittel, 1913.
Eleutherocaris whitfieldi Clarke: Chadwick, 1935;
Clarke, 1904; Clarke and Luther, 1904; Clarke and
Ruedemann, 1903; Dana, 1895; Grabau and Shimer,
1910; Gurich, 1929; Schuchert, 1943.
Eleutherocaris ( Echinocaris ) whitfieldi Clarke: Van
Straelen and Schmitz, 1934.
Elymocaris: Oehlert, 1889.
Equisetides wrightiana Dawson: Clarke, 1885b; Dawson,
1881b; Hall, 1884; Jones and Woodward, 1884; Vogdes,
1893; Wright, 1884.
Equisetites wrightiana Dawson: Clarke, 1885; Dawson,
1881a; Dawson, 1882.
Equisetites wrightianus Dawson: Dawson, 1881a.
Hebertocaris: Stumm and Chilman, 1969.
Montecaris: Rolfe, 1981.
Montecaris? sp.: Dzik, 1980.
Ohiocaris wycoffi Rolfe: Rolfe, 1962c.
Orozoe: Oehlert, 1889.
Palaeopalaemon newberryi Whitfield: Whitfield, 1880;
Whitfield, 1892.
Pephricaris: Gurich, 1929.
Pephricaris horripilata Clarke: Beecher, 1902; Clarke,
1898b; Gurich, 1929.
Physophycos bilobatus Lesquereux: Lesquereux, 1891.
52
HANNIBAL AND FELDMANN
No. 42
Pseudodontichthys whitei Skeels: Case, 1982; Skeels,
1962.
Ptychocaris: Copeland, 1960b; Hall and Clarke, 1888.
Ptychocaris novaki Copeland: Copeland, 1960b.
Rhinocaris: Stumm and Chilman, 1969.
Rhinocaris columbina Clarke, 1888: Clarke, 1902; Zell,
1985.
Silesicaris: Copeland, 1960b.
Stylonurus: Beecher, 1900; Clarke and Ruedemann,
1912; Hall, 1884.
Stylonurus (?) wrightiana (Dawson): O’Connell, 1916.
Stylonurus wrightianus (Dawson): Chadwick, 1935;
Clarke, 1904; Vogdes, 1917.
Stylonurus ? wrightianus (Dawson): Chadwick, 1935;
Clarke, 1904; Clarke and Luther, 1904; Clarke and
Ruedemann, 1903; Kjellesvig-Waering, 1961; Kjellesvig-
Waering, 1966; Waterston, 1968.
Stylonurus'? wrightianus (Dawson) (-Echinocaris wright-
ianus)-. Clarke and Ruedemann, 1912.
Stylonurus (?) ( Echinocaris ?) wrightianus (Dawson):
Beecher, 1900; Clarke, 1892; Hall and Clarke, 1888.
Stylonurus (= Equise tides - ? Echinocaris) wrightianus
(Dawson): Vogdes, 1890.
Stylonurus (Equisitides) wrightianus (Dawson): Vogdes,
1893.
Stylonurus (= Equisetides - Echinocaris - ? Echinocaris)
wrightianus (Dawson): Vogdes, 1890.
Trigonocarys lebescontei Barrois: Barrois, 1891.
Tropidocaris: Cleevely, 1983; Oehlert, 1889.
A BLUEGRASS NEW TO OHIO:
POA SALTUENSIS FERN. & WIEG.
JAMES K. BISSELL
Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Pasture bluegrass ( Poa saltuensis Fern. & Wieg.), a
northern species which ranges from western Ontario and
Minnesota to eastern Quebec, southward through north-
ern and western New England to northern Maryland and
West Virginia (Fernald 1950), has been discovered re-
cently in northeastern Ohio in Chardon Township,
Geauga County. P. saltuensis is a species of open woods,
thickets, and recent clearings (Fernald 1950). The Ohio
specimen (Bissell 1984: 23) was collected on 23 May 1984
just below a bluff rim of an eroding, south-facing clayey
slump along the west valley wall of Big Creek just north of
the corporate limits of Chardon. A duplicate of the mu-
seum specimen has been annotated by A. A. Reznicek at
the University of Michigan, Ann Arbor.
A diligent search for additional occurrences of P. sal-
tuensis was made on similar slumping valley walls of the
Chagrin River to the west of Chardon and slumping val-
ley walls of the Grand River to the east of Chardon during
1985 and 1986. However, no additional Ohio records for
P. saltuensis were discovered.
P. saltuensis is fairly common in adjacent Pennsylva-
nia, occurring across the northern portion of the state ex-
cept the extreme northwestern counties, and extending
south through the state on the Appalachian Mountains
(Wherry et al. 1979). This newly discovered Ohio locality
of P. saltuensis is about 90 miles west of its westernmost
occurrence in Pennsylvania (Wherry et al. 1979). Within
Michigan P. saltuensis is confined to the Upper Peninsula
and central and northern Lower Peninsula with a disjunct
county occurrence in Allegan County on Lake Michigan
in the southern lower Peninsula (Voss 1972).
P. saltuensis is uncommon at the single Ohio locality,
growing in partial shade on the upper slump face just
below a forested bluff rim. Several breaks occur on the
forest along the bluff rim. The most dominant tree along
the bluff rim is white oak (Quercus alba L.), with a few
scattered beech (Fagus grandifolia Ehrh.). Shrubs asso-
ciated with P. saltuensis include buffaloberry (shepherdia
candensis L.), leatherwood ( Dirca palustris L.), witchha-
zel ( Hamamelis virginiana L.) and maple-leaf viburnum
( Viburnum acerifolium L.). The most common herbace-
ous associates of P. saltuensis at the site are longstalked
sedge ( Carex pedunculata Willd.) and wreath goldenrod
(Solidago caesia L.).
P. saltuensis is similar to another species of bluegrass
(Poa languida Hitchc.) (Gleason and Cronquist 1963)
which does occur in northeastern Ohio (Braun 1967). The
length of the ligules in P. saltuensis runs between .6 and
1.5 mm (3 mm), while ligule length in P. languida ranges
between (2. 1 mm) 2.4 and 4 mm (Voss 1972). The margins
of the lemmas in both species are glabrous (Voss 1972). P.
saltuensis has acute or acuminate, distinctly 5-nerved
lemmas, while P. languida has obtuse, obscurely nerved
lemmas (Gleason and Cronquist 1963). Another species
of bluegrass (Poa cuspidata ), common throughout the
eastern half of Ohio (Braun 1967), grows in habitats sim-
ilar to the P. saltuensis habitat in Chardon. Many of the
northeastern Ohio P. cuspidata specimens in the CLM
herbarium contain the label comment: “in partial shade
on eroding upper slopes just below bluff rim.” The keels
of the lemmas of P. cuspidata are pubescent in contrast to
the glabrous keels of P. languida and P. saltuense (Fer-
nald 1950).
References
Braun, E. L. 1967. The Monocotyledonaea (of Ohio): Cat-tails
to orchids. With Gramineae by Clara G. Weishaupt. Colum-
bus: Ohio State University Press.
Fernald, M. L. 1950. Grays manual of botany. 8th ed. New
York: American Book Company.
Gleason, H. A., and A. Cronquist. 1963. Manual of vascular
plants of northeastern United States and adjacent Canada.
New York: D. Van Nostrand.
Voss, E. G. 1972. Michigan flora part E Gymnosperms and
Monocots. Cranbrook Institute of Science Bulletin 55.
Bloomfield Hills, Mich.
Wherry, E. T., J. M. Fogg, Jr., and H. A. Wahl. 1979. Atlas of
the flora of Pennsylvania. Philadelphia: The Morris Arbore-
tum. Univ. Penn.
Kirtlandia, No. 42
© by The Cleveland Museum of Natural History
Membership and Subscription
Membership in Kirtlandia Society is open to any member of the Cleveland Mu-
seum of Natural History. Current individual membership is $30.00 ($10.00 for stu-
dents); participating members, $100.00; institutional, $30.00; sustaining, $250.00.
Applications and fees should be sent to the Treasurer, Kirtlandia Society, Cleve-
land Museum of Natural History, Cleveland, Ohio 44106.
Membership includes subscription to Kirtlandia. There is a non-U. S. postal sur-
charge of $5.00. All subscriptions outside North America will be sent by air. Pay-
ment must be made in U.S. dollars. Payments not drawn on a U.S. bank will be
subject to bank collection charges.
Change of Address '. Please send to Museum six weeks prior to move; enclose
present mailing label with change of address.
Claims for missing issues: Claims cannot be honored beyond four months after
mailing date. Duplicate copies cannot be sent to replace issues not delivered be-
cause of failure to notify Museum of change of address.
Back issues: Individual issues of Kirtlandia are available from the Museum Pub-
lications Office. A complete list is available upon request.
CLEVELAND, OHIO
NUMBER 43
KIRTLANDIA
•NATURAL HISTORY*
PALEONTOLOGY
A New Species of Sauropod Dinosaur, Haplocanthosaurus delfsi
sp. nov., from the Upper Jurassic Morrison Fm. of Colorado
John S. McIntosh and Michael E. Williams 3
Isolated Tetrapod Remains from the Carboniferous of West Virginia
Stephen J. Godfrey 27
JULY 1988
KIRTLANDIA
The Scientific Publication of the Cleveland Museum of Natural History
David S. Brose, Editor
Joseph T. Hannibal, Assistant Editor
Brief History and Purpose
Kirtlandia is a publication of the Cleveland Museum of Natural History, an
organization , incorporated in 1922 whose predecessor, the Ark, was founded in
Cleveland by Jared Potter Kirtland in 1837. Kirtlandia is a continuation of the earlier
series. Scientific Publications volumes 1 to 7 (1928-1937), and new series volumes
2 to 4 (1962-1965), and the Survey of Ohio Fluted Points. Supported by the
Kirtlandia Society of the Cleveland Museum of Natural History, Kirtlandia is
devoted to the publication of scientific papers in the various fields of inquiry within
the Museum’s sphere of interest: Cultural and Physical Anthropology; Archaeology;
Botany; Geology; Paleobotany; Invertebrate and Vertebrate Paleontology; Systematic
Ecology; and Invertebrate and Vertebrate Zoology. Issues will vary from single
monographs to collections of short papers, review articles, and brief research notes.
Kirtlandia is distributed by The Kent State University Press, Kent, Ohio 44242.
Authorization to photocopy items for internal or personal use, or the internal or
personal use of specific clients, is granted for libraries and other users registered
with the Copyright Clearance Center (CCC) Transactional Reporting Service,
provided that the base fee of $00.50 per copy, plus .25 per page is paid directly to
CCC, 27 Congress Street, Salem, MA 09170. The fee code for users of the
Transactional Reporting Service is 0075-6245/88 $00.50+.25.
Kirtlandia is abstracted in Biological Abstracts and indexed in Bibliography and
Index of Geology and Zoological Record.
Kirtlandia No. 43
© 1988 by The Cleveland Museum of Natural History
KIRTLANDIA
THE CLEVELAND MUSEUM OF NATURAL HISTORY
Cleveland, Ohio July 1988 Number 43
PALEONTOLOGY
A New Species of Sauropod Dinosaur, Haplocanthosaurus delfsi
sp. nov., from the Upper Jurassic Morrison Fm. of Colorado
John S. McIntosh and Michael E. Williams
Isolated Tetrapod Remains from the Carboniferous of West Virginia
Stephen J. Godfrey
$M\lTHS6ivJ^
AUG 2 6 1988
3'fiRARIES
3
27
ISSN: 0075-6245
KIRTLANDIA
EDITOR
DAVID S. BROSE
Cleveland Museum of Natural History
ASSOCIATE EDITORS
MARY BAUM
Research Librarian
Cleveland Museum of Natural History
JAMES K. BISSELL
Curator of Botany
Cleveland Museum of Natural History
BRUCE LATIMER
Curator of Physical Anthropology
Cleveland Museum of Natural History
SONJA TERAGUCHI
Manager of Collections
Cleveland Museum of Natural History
JOSEPH T. HANNIBAL
Curator of Invertebrate Paleontology
Cleveland Museum of Natural History
MARTIN ROSENBERG
Editorial Assistant
Case Western Reserve University
EDITORIAL ADVISORY BOARD
RODNEY FELDMANN
Professor of Geology
Kent State University
MICHAEL C. HANSEN
Geologist
Ohio Geological Survey
RICHARD MEINDL
Associate Professor of Anthropology
Kent State University
G. MICHAEL PRATT
Associate Professor of Anthropology
Heidelberg University
DAVID H. STANSBERY
Director, Museum of Zoology
Ohio State University
FREDERICK H. UTECH
Curator of Botany
Carnegie Museum of Natural History
ED VOSS
Curator of the Herbarium
University of Michigan
ANDREW M. WHITE
Professor of Biology
John Carroll University
KIRTLANDIA
THE CLEVELAND MUSEUM OF NATURAL HISTORY
Cleveland, Ohio July 1988 Number 43:3-26
A NEW SPECIES OF SAUROPOD DINOSAUR,
HAPLOCANTHOSAURUS DELFSI SR NOV.,
FROM THE UPPER JURASSIC MORRISON FM. OF COLORADO
john s. McIntosh
Wesleyan University
Middletown, Connecticut 06457
and
MICHAEL E. WILLIAMS
Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Abstract
A skeleton mounted in the Cleveland Museum of Natural History is
assigned to the rare genus of sauropod dinosaur, Haplocanthosaurus , as a
new species, H. delfsi. The seventy-foot-long skeleton is the largest known
specimen of Haplocanthosaurus and the only mounted one. It was
collected between the years 1954 and 1957 by a museum party from the
clays of the lower part of the Upper Jurassic Morrison Formation on Oil
Creek in Fremont County, Colorado, north of Canon City.
4
mcintosh and williams
No. 43
Abstract
A skeleton mounted in the Cleveland Museum of Natural History is
assigned to the rare genus of sauropod dinosaur, Haplocanthosaurus , as a
new species, H. delfsi. The seventy-foot-long skeleton is the largest known
specimen of Haplocanthosaurus and the only mounted one. It was
collected between the years 1954 and 1957 by a museum party from the
clays of the lower part of the Upper Jurassic Morrison Formation on Oil
Creek in Fremont County, Colorado, north of Canon City.
Introduction
In 1901 the newly appointed curator of vertebrate pale-
ontology at the Carnegie Museum in Pittsburgh, John Bell
Hatcher, decided to reopen the Marsh-Felch quarry in
Garden Park, Colorado, about ten miles north of Canon
City. This quarry had produced the type specimens of a
number of species of dinosaurs including Diplodocus lon-
gus, Allosaurus fragilis, Ceratosaurus nasicornis, Stego-
saurus stenops , and “Morosaurus” agilis. W. B. Utterback
was sent to the field and during the summer’s operations,
collected two medium-sized, partial skeletons of a new
genus of sauropod dinosaur. The more complete of these,
CM 572, was subsequently described in a brief paper by
Hatcher (1903a) as Haplocanthus priscus. Four months
later he altered the name to Haplocanthosaurus
(1903b: 100) because Haplocanthus was “essentially preoc-
cupied” by a genus of fish named by Agassiz (1844). In
November 1903, Hatcher published a monograph on Hap-
locanthosaurus in which the second skeleton was also
described as a new species, H. utterbacki (CM 879). Since
the publication of Hatcher’s memoir, very little has been
added to our knowledge of this animal. In a review of the
sauropods, von Huene (1929) modified several of Hatcher’s
conclusions and suggested that the number of dorsal verte-
brae should be reduced from 14 to 12. In a paper redescrib-
ing “ Morosaurus ” agilis Marsh, 1889, Gilmore (1907)
conjectured that its type specimen, a cranial fragment and
cervicals 1 to 3, might belong to Haplocanthosaurus, and
this remains a distinct possibility as discussed below.
Finally, in 1981, McIntosh referred some limb bones from
the type locality to this form as probable.
No new material belonging to this genus was reported
until summer of 1954, when a field party from the Cleve-
land Museum of Natural History began excavating a large
sauropod skeleton on the east bank of Oil Creek (Four Mile
Creek, Nine Mile Creek) less than a mile south of the
historic Marsh-Felch Quarry 1 (Fig.l).
History of the CMNH Quarry (The “Sawropod Lode”)
Early in the summer of 1954, William E. Scheele, then
director of the CMNH, dispatched a small field party to
several western states with the expressed aim of finding a
mountable dinosaur for exhibit. The search initially cen-
tered around Vernal, Utah, where the crew was “shown the
(★), and the nearby Marsh-Felch (■) and the Cope-Lucas Quarries (•).
Cooper Mtr. Quadrangle, Colorado.
ropes” by Leroy “Pop” Kay, long-time curator at the
Carnegie Museum. Kay also provided leads to several
promising sites. Although several specimens were located,
none proved workable due to problems of obtaining permis-
sion or the difficulty of the required excavation.
The field party then split in two, one group headed by
director Scheele travelling northward into Wyoming and the
second moving eastward into Colorado. While camped at
the Colorado National Monument near Grand Junction, a
member of the Cleveland crew overheard a conversation
between students in a geological field party who had found
and collected a partial bone in a stream exposure near
Canon City. Dr. Carl Sanderson, a geologist at Louisiana
State University and leader of that University’s summer
field camp, kindly provided a map of the site. The bone
fragmnent was given to the CMNH crew, and later proved
to be the posterior end of the fourth cervical vertebra.
The first day’s excavation demonstrated the presence of a
considerable amount of bone, and it soon became apparent
that the major portion of a skeleton was present. The quarry
site (Figs. 1, 2A) is located just off the road, on the east
bank of Four Mile Creek in the NW 1/4, NW 1/4, Sec. 34,
T. 17S, R.70W, Fremont County, Colorado (Cooper Mt.
Quad.). Since this was during a major uranium boom, it was
deemed necessary to protect the site from potential pros-
pecting damage by filing a claim with the Fremont County
recorder. A letter from the County Clerk and Recorder’s
office, dated October 21, 1954, states “The name of this
lode is Sawropod, and the reception number is 289758.”
Kirtlandia, No. 43, July 1988 © by the Cleveland Museum of Natural History
1988
NEW SPECIES OF SAUROPOD DINOSAUR
5
Fig. 2A The CMNH Quarry from the road leading north from Canon City. Early summer, 1954.
The bones were recovered from the lower part of the
Morrison Formation, in a light gray clay layer, bound both
above and below by massive sandstone ledges (Figs. 2B,
2C, 3). The upper sand was five- to eight-feet thick and was
overlain by some eight feet of sand and gravel. Quarrying
operations during the first field season consisted of tunnel-
ing under the overlying sandstone which was shored up with
vertical timbers. By the end of the season, the practical
limits of this technique had been reached and the remaining
excavation, in 1955 and 1957, (Fig. 4) was accomplished by
removing the overburden with a bulldozer and blasting away
the sandstone ledge.
Each of the three years’ excavations was led by Edwin
Delfs, then an undergraduate biology major at Yale Univer-
sity. Other field crew members were high school and
college students Wesley Williams, William West, and
Richard Jones (1954), Wesley Williams and Joseph Hurley
(1955), and Ralph Wrisley (1957). A local rancher, Joe
Rhode, of Garden Park, Canon City, did the bulldozer work
and provided invaluable advice on construction matters and
in handling the large and often very heavy blocks. Members
of the “permanent” field crew were supplemented from
time to time by a number of Cleveland Museum staff
members and volunteers, most notably Mary Flahive,
Elizabeth Olmstead, David Roberts, William Scheele, Dan
Snow, and Ellen Walters.
The skeleton was lying on its left side and was largely
articulated. As so often happens with sauropod skeletons,
the neck was drawn sharply backward and the skull had
snapped off and was not found. The anterior dorsal verte-
brae and all but the first four cervicals had been eroded
away by the stream. The posterior two-thirds of the tail was
also missing. The ribs and girdle bones of the under (left)
side were in place, but only the ilium remains of the right
side. The only limb bones preserved were the left femur and
the heads of two bones restored as the left tibia and fibula.
Other vertebrate remains found with the skeleton include a
number of turtle fragments, an isolated theropod tooth, and
6
McIntosh and williams
No. 43
Fig. 2B. Installing timbers, early summer, 1954. Left to right, Wesley Williams, Richard Jones, William West.
the skull and partial skeleton of a new goniopholid croco-
dile, Eutretauranosuchus delfsi (Mook 1967).
The members of the field party are to be congratulated
for their skill in exhuming the huge but often very fragile
bones, particularly the vertebrae. Some of the packages
were opened in the winter of 1954-55 (Delfs 1961) and the
skeleton was determined to be Haplocanthosaurus. Further
preparation and restoration of the missing parts prior to
mounting the skeleton were accomplished at the American
Museum of Natural History in New York under the super-
vision of veteran preparator George Whitaker. The speci-
men was first displayed in Kirtland Hall at the Cleveland
Museum of Natural History in 1961, with the body resting
on the ground. It was remounted in the upright pose shown
in Figure 5 two years later. Although brief notices have
appeared in the popular press (Anonymous 1959: Delfs
1961; Pearl 1975; Piel 1963; Anonymous 1966; Williams
1982), the skeleton has not been described.
Abbreviations
AMNH American Museum of Natural History
CM Carnegie Museum of Natural History
CMNH Cleveland Museum of Natural History
OUM Oxford University Museum
USNM National Museum of Natural History
YPM Yale Peabody Museum
Systematic Paleontology
Class Reptilia
Subclass Archosauria
Order Saurischia Seeley
Suborder Sauropoda Marsh
Family Cetiosauridae Seeley
Genus Haplocanthosaurus (Hatcher 1903b)
Diagnosis — Cervical vertebrae of only moderate length,
pleurocoels prominent but simple, neural spines of poste-
rior cervicals and anterior dorsals not divided. Dorsal
1988
NEW SPECIES OF SAUROPOD DINOSAUR
7
Fig. 2C. Quarry site, late summer, 1954, showing the removal of the first of 3 large plaster jackets, the
2500 lb. “Iceberg." Ed Delfs (by jacket), Joe Rhode (wearing hat) and unidentified truck driver.
centra relatively small, all containing prominent pleuro-
coels; dorsal arches high with diapophyses extending up-
ward at 45 degrees as well as outward, spines short and
broad. Sacrum consisting of the usual dorso-sacral, three
primary sacrals and a caudo-sacral with centra coosified, as
are the sacral ribs, to form a yoke; small pleurocentral
cavities in at least some of the centra; spines relatively low
with a tendency toward coalescence of all, but particularly
numbers one to three. Caudal centra amphicoelous, short,
and without pleurocentral cavities; chevron facets very
prominent and give the underside of the centrum a sculp-
tured appearance; caudal spines slender and curved back-
ward in the anterior region; they are of moderate height
anteriorly and low further back. Distal end of scapula thin
and broadly splayed; proximal plate relatively smaller than
in most sauropods. Sternal plates large and subquadrangu-
lar. Proximal part of ischium relatively small, shaft straight,
distal end slighty broadened but not thickened. Femur
neither overly slender nor stout, the latero-medial diameter
of the shaft significantly exceeding the antero-posterior one
as in Brachiosaurus. Other possible significant generic
characters are discussed below in connection with speci-
mens probably, but not certainly, belonging to Haplocan-
thosaurus.
Haplocanthosaurus priscus (Hatcher 1903b)
Haplocanthus priscus Hatcher 1903a
Haplocanthosaurus priscus (Hatcher 1903b)
Haplocanthosaurus utterbacki (Hatcher 1903c)
Holotype - CM 572 (Hatcher 1903a)
Horizon and locality — Upper Jurassic Morrison Fm.,
Marsh-Felch Quarry, Garden Park, Colorado
Amended specific diagnosis — Medium sized Haplocantho-
saurus with comparatively slender femur and pelvic girdle.
Distal ends of ischia narrowed, rotated inward and fused to
their opposite in the midline.
Type species Haplocanthus priscus Hatcher 1903a
Haplocanthosaurus delfsi sp. nov.
Holotype - CMNH 10380
Horizon and locality — Upper Jurassic Morrison Fm. East
bank of Four Mile Creek NW 1/4, NW 1/4, Sec. 34,
T. 17S., R.70W. , Fremont County, Colorado (Cooper Mt.
Quad.).
8
McIntosh and williams
No. 43
Fig. 3. Dick Jones exposing a section of rib, early summer 1954.
Material — CMNH 10380 cervicals 1-4, nine posterior
dorsals with ribs of the left side, five sacrals, caudals 1-14,
several chevrons, shaft and distal end of left scapula,
fragmentary coracoid?, right sternal plate, proximal end of
left radius, proximal end of left ulna, both ilia, left pubis,
left ischium, left femur.
Diagnosis — Very large Haplocanthosaurus with most
measurements 35-50% greater than that of the fully adult
holotype of H. priscus (CM 572). Girdle bones and femur
more robust than in H. priscus. Pubis in particular, much
heavier distally. Distal end of ischium broader, not rotated
inward and not fused to its mate. V-shaped, anterolaterally
projecting laminae present on neural spines of the middle
dorsals, and greater developement of median laminae on the
posterior dorsal spines than in H. priscus.
Designation of the specimen
As mentioned in the introduction, Hatcher changed the
generic name from Haplocanthus to Haplocanthosaurus
because Haplocanthus was “essentially preoccupied” by a
genus of fish named by Agassiz in 1844. Agassiz used the
spelling Haplocanthus. A similar situation exists with the
stegosaur originally named Kentrosaurus by Hennig (1915)
but altered to Kentrurosaurus (Hennig 1916) because
Lambe had used Centrosaurus for a genus of ceratopsian.
Following Romer (1966), Galton (1982) has recently re-
turned to the use of Kentrosaurus , arguing the “two generic
names cannot be considered homonyms even if it is only one
letter which is different (International Code of Zoological
Nomenclature 1961 Article 56a).” This discussion has been
challenged by Anderson (1982) who argued that Hennig
made the correct decision based on the rules in 1915. In the
present case, Haplocanthosaurus is clearly the valid form,
because the senior synonym Haplocanthus must be consid-
ered a forgotton name (nomen oblitum) due to lack of use
(ICZN, 1961 art. 23b). The most recent edition of the Code
(ICZN, 1985), however, does not use the term nomen
oblitum and requires that the current usage be maintained
while the matter is referred to the commission for a ruling
(art. 23b).
Skull and mandible
No part of the skull, mandible, hyoid bones, proatlas, or
any portion of the dentition has been found with any of the
three partial skeletons of Haplocanthosaurus. The skull on
the mounted skeleton is modelled. However, see comments
concerning Morosaurus agilis below.
Vertebrae
Cervicals. The number of cervical vertebrae in Haplo-
canthosaurus is not known. In CM 572 only the last two
were preserved. In CM 879 there were seven complete
cervicals, an additional centrum, an arch, and a fragmen-
tary arch. The atlas and axis are not represented. From this
material Hatcher surmised that the total number of cervical
vertebrae was fifteen, the same as in Diplodocus, the only
sauropod in which the number was known with certainty at
the time. Since then, a number of sauropod genera have
been found to have fewer than fifteen cervicals, and indeed,
von Huene (1929) revised Hatcher’s figure down to thir-
teen. The Cleveland skeleton does not help resolve this
question, but by providing the atlas and axis (Fig. 6), which
are missing in the Pittsburgh material, it does add signifi-
cantly to our knowledge of the animal.
Although most of the neural arch of the atlas is missing,
what remains shows that it was firmly coalesced to the
intercentrum as in all other adult sauropods. The odontoid
is firmly united with the anterior end of the axis and extends
straight forward, tapering and ending in a blunted point.
The intercentrum is relatively longer than in Apatosaurus ,
but as in the atlas of the latter, well-developed articular
facets for a single headed cervical rib occur on the posterior
part of the lateral face. The ribs mounted on the atlas are
both plaster, however, the cervical rib attached to the left
side of the axis is real, and is clearly single-headed, lacking
a dorsal tubercular process. The capitular end is consider-
ably expanded and conforms well to the articular surface of
the atlas. The distal end of the rib is rounded and restored
in plaster. It seems likely that it was displaced from the atlas
and that the axis bore a double-headed rib as is typical of
sauropods.
10 ft
NEW SPECIES OE SAUROPOD DINOSAUR
Fig. 4. Quarry map redrawn from field notes. The extent of the three summers’ activities are indicated
by dashed lines. Jacket no. 2, containing portions of rib and apparently the neural spine of the dorsal
vertebrae in PK 13, was lost in an early flash flood. (Not to scale.)
10
McIntosh and williams
No. 43
Fig. 5. Haplocanthosaurus delfsi, mounted skeleton on display in Kirtland Hall, Cleveland Museum of
Natural History.
The axis (Fig. 6) is complete except for the lower front
part of the arch, which would have included the articulation
with the neurapophysis of the atlas. A moderate depression
occupies a large part of the lateral face of the centrum rather
than a true pleurocoel. The anterior part of this depression
is deepened into a small round cavity that penetrates the
medium septum. Below and slightly anterior to this opening
is a well developed, dorso-ventrally elongate, articular
facet, the parapophysis. The transverse process arises on
the rear of the lower part of the arch, however, the distal
portions of both processes are restored.
As in other sauropod axes, the arch and spine sweep back
and upward from front to rear and the postzygapophyses lie
directly beneath the high point of the spine and directly
above the rearmost part of the centrum.
In cervical three the lateral depression occupies about
half of the side of the centrum. The vertebra is strongly
opisthocoelous and the prezygapophyses extend beyond the
anterior ball of the centrum. The diapophyses lie further
forward than in the axis, and both the neural spine and the
postzygapophyses are higher. As the distal end of the
cervical rib is not preserved, it cannot be ascertained
whether or not it extends beyond the back end of the
centrum.
The fourth cervical is considerably larger than the third
and is also strongly opisthocoelous. The edges of the lateral
depression are sharper but do not yet define a true pleuro-
coel. The zygapophyses resemble those of cervical three,
but the postzygapophyses are placed higher. The undivided
spine has assumed a triangular shape. Again nothing can be
said concerning the extent of the cervical rib.
The fragment identified by Hatcher as a postzygapophy-
sis of cervical three in CM 879 is too fragmentary to be of
any comparative value. Unfortunately the centrum of cer-
vical 4? of CM 879 has been damaged since Hatcher’s day
and only the rear half remains. Its lateral depression is
similar to that of CMNH 10380. The arch is much dis-
torted, but differs in no important way from that of the
Cleveland specimen, so while Hatcher’s identification of it
as the fourth may be correct this cannot be verified with
certainty.
A more significant comparison can be made between the
cervicals of CMNH 10380 and those of a specimen found
directly beneath a left femur associated with USNM 4275,
discussed below as probably belonging to this genus. The
specimen USNM 5384, consists of the braincase, proatlas,
atlas, axis, and cervical 3. It was described briefly by
Marsh (1889) as Morosaurus agilis sp. nov., and in detail
1988
NEW SPECIES OF SAUROPOD DINOSAUR
11
Fig. 6. First four cervical vertebrae of Haplocanthosaurus delfsi, CMNH 10380. Most of the arch of
the atlas and the cervical ribs are restored. Left lateral view.
by Gilmore (1907), who suggested that it might belong to
Haplocanthosaurus since it came from the same quarry as
the holotype. A direct comparison of this specimen with the
Cleveland skeleton appears to bear out Gilmore’s conjec-
ture. The Washington specimen has been crushed as are
most of the specimens from the Marsh-Felch quarry, but
when one takes into account 1) Gilmore’s observation that,
“the spinous process of cervical 3 has been crushed
forward somewhat from its normal position” and that 2)
“the transverse process, postzygapophysial lamina and
postzygapophyses are wanting,” the two specimens appear
quite similar. This is not only apparent in the general shapes
of the first three cervicals but more specifically in the
simple characters of their pleurocentral depressions. These
are not crossed by laminae nor puctuated with subsidiary
cavities as in many of the more advanced sauropods, and
their margins lack the sharp lips characteristic of a true
pleurocoel.
It will not be our intent here to discuss the skull fragment;
suffice it to say that it differs significantly from that of
Camarasaurus ( Morosaurus ) and the other Morrison sau-
ropods. The presence of complete right and left halves of
the proatlas in position in USNM 5384 is of great interest,
since this element is rarely reported in the sauropods.
Although it seems likely that Morosaurus agilis is syn-
onymous with Haplocanthosaurus priscus, we refrain from
invoking the law of priority at this time, because of the
distorted and incomplete nature of USNM 5384, and
because the first three cervicals are not among the most
diagnostic elements of the sauropod skeleton.
Dorsal Vertebrae. Nine dorsal vertebrae are preserved in
CMNH 10380. The first six of these appear to have been in
articulation, but the seventh had been displaced upward and
rotated onto its side (see additional comments concerning
the numerical sequence, below). The eighth dorsal was
displaced further back and lay near the ilium. The ninth had
been carried still further back and was found beyond the
fourteenth (last preserved) caudal vertebra. As it turned out,
collecting ended in 1954 with the seventh dorsal in the
series (Fig. 4). Collecting resumed in 1955 with the eighth
dorsal and the sacrum, but the widely displaced ninth dorsal
was not recovered until 1957. Field notes indicate that the
latter was displaced at least twenty feet from the articulated
series. Thus, the question arises as to whether additional
dorsals between the seventh and ninth may have also been
displaced and lost. For a number of reasons, we believe that
this is not the case, and that the series which is mounted as
the nine dorsals anterior to the sacrum is correctly restored.
Unlike most of the other Morrison sauropods, in Haplocan-
thosaurus the variation from one posterior dorsal to another
is minor.
Using the two Pittsburgh skeletons, Hatcher determined
the number of dorsal vertebrae in Haplocanthosaurus as
fourteen. In CM 572 there were three articulated vertebral
segments, the first of which he took to be the last two
cervicals and first dorsal. The second segment consisting of
nine dorsals, he took to be the last nine, while the third
segment represented the sacrum and tail. In CM 879 the
presacral vertebrae lay in approximate order, but only a
series of six posterior dorsals were actually articulated. He
12
McIntosh and williams
No. 43
TABLE 1
Measurements of Cervical Vertebrae, CMNH 10380
No.
Centrum or Intercentrwn
Height
Spread
length
anterior
posterior
overall
pre-
post-
Diapo-
breadth
height
breadth
height
height
zygap.
zygap.
physes
1
65
66
65
63
85
e200
—
2
192
70
98
80
97
277
194
113
3
237
106
102
55
72
282
186
203
172
4
300
123
123
72
71
305
197
232
197
All measurements here and elsewhere in this paper are in millimeters, e - estimated
took thirteen of these to be dorsals two through fourteen;
dorsal one represented in CM 572 being absent. Some years
later von Huene (1929) used Hatcher’s figures to reinterpret
these assignments and concluded that Hatcher’s dorsal one
was really the penultimate cervical and that his dorsal two
was actually the last cervical, thus reducing the dorsal count
to twelve. Briefly stated, the reasons for the change were the
occurrence of the parapophysis below the pleurocentral
cavity and the fact that Camarasaurus had recently been
shown to have twelve dorsals. The other two sauropod
genera with known dorsal counts at the time of von Huene’s
paper were Diplodocus and Apatosaurus both of which had
ten. Subsequently the Chinese genus Euhelopus was shown
to possess fourteen, and the prosauropod Plateosaurus had
fifteen, including the last one which is taken into the
sauropod sacrum as a dorso-sacral. Thus it would not be
surprising if a relatively primitive sauropod like Haplocan-
thosaurus did indeed possess fourteen dorsals. Theoretical
arguments aside, the empirical evidence also favors Hatch-
er’s interpretation. As students of the sauropods know well,
the transition from cervicals to dorsals in this group is quite
gradual, and the determination of where the change occurs
is based on the ribs. The transition from the last cervical rib
to the first thoracic rib is abrupt. In Diplodocus the
parapophysis drops significantly on the arch of the fifth
dorsal from the position seen in subsequent dorsals. In the
fourth and third it is lower on the centrum in front of the
pleurocentral cavity, and in dorsals one and two it lies
beneath this cavity. Dorsal one resembles the last cervical
more closely than it does a typical dorsal, but it bears a true
thoracic rib. There is another difference: in mature sauro-
pods the cervical ribs are invariably coalesced to their
vertebrae whereas the thoracic ribs are not. In CM 879 the
neural arches of most of the presacrals and sacrals are not
fused to their centra, and the cervical ribs are not fused to
their vertebrae, thus indicating an immature individual. In
CM 572, however, all the arches are firmly fused to their
centra, no trace of the line of fusion remaining. Likewise
the cervical ribs on both sides of the two anteriormost
vertebrae of segment one are firmly fused to their vertebrae,
while the ribs of the third are free (and have been lost).
Furthermore, the parapophyses lie beneath the pleurocen-
tral cavity at the base of the centrum, which is to be
expected in the first dorsal. It would appear to us that the
only possibility for reducing the number of dorsals from
fourteen would be if Hatcher’s dorsal two of CM 879
corresponds to his dorsal one of CM 572. This would
reduce the number to thirteen, and might conceivably be
attributed to individual variation, although not even von
Huene suggested this possibility. While admitting that
future discoveries might bear out this possibility, we believe
that based on current evidence, it is more prudent to stick
with Hatcher’s original determination of fourteen dorsals.
What then are the locations in the series of the CMNH
10380 dorsals? It seems reasonably certain that the first six
and probably the first seven occurred in serial order. The
first package taken out, PK 1, was a 750 lb. block said to
contain “assorted bones’’ (Fig. 5). A later, typed version of
the field notes lists PK 1 as containing “vertebra in scrappy
condition.’’ Package no. 2, which was immediately behind
PK 1, contained the neural arch of another dorsal, labelled
PK 13 on the quarry diagram. Unfortunately, PK 2, plastic
jacket and all, was lost in a flood early in the excavation
(Fig. 7).
In the mounted skeleton, two dorsals were placed anterior
to the one missing the neural arch (Fig. 8), suggesting that
the first two were in PK 1 . Evidence tending to confirm this
is the considerable difference in the anterposterior expanse
of the neural spines of the first two vertebrae. This is in
contrast with the remaining dorsals, whose neural spines
are essentially uniform in this respect, as are the rearmost
dorsals of both CM 879 and CM 572.
The two displaced dorsals have been mounted as the two
between the sacrum and the former group and they seem to
fit well in this series (Fig. 9). A hyposphene-hypantrum
articulation is present in the dorsals of both Carnegie
specimens from the sixth to the last. These articulations are
likewise present in the dorsals of CMNH 10380, save for
the first and third where they are restored in plaster. The
hyposphene on the first dorsal may be partially real, but the
presence of a hypantrum on the second demonstrates its
presence nonetheless. It would seem, therefore, that the
series is correctly restored, and represents dorsals six
through fourteen (dorsals four through twelve in the
mount).
One apparent difference, however, is the position of the
capitular articulation on the sixth dorsal, which in CMNH
10380 is at the same level as in succeeding vertebrae, but
1988
NEW SPECIES OF SAUROPOD DINOSAUR
13
Fig. 7. One of several flash floods that plagued the excavation. Joseph Hurley is seen standing atop a
large plaster jacket. Summer 1955.
distinctly lower and greatly enlarged on the sixth dorsal of
both Carnegie Museum specimens.
Detailed descriptions of the dorsals are unnecessary since
Hatcher’s suffice, but several points should be noted con-
cerning the new specimen. First, all the neural arches are
firmly fused to their respective centra. The centra are
relatively small, as in CM 572 and CM 879, with large,
sharpley defined pleurocentral cavities high up on the
centra. The anterior and middle centra are strongly opistho-
coelous, the posterior ones less so (Fig. 8). Indeed, the
anterior ball of the last three are only slightly convex to
nearly flat. The centra of dorsals 13 and 14 are noticeably
shorter than the others, but number 13 has clearly received
some antero-posterior flattening (crushing) so this feature
may be exaggerated. The neural arches and spines are, for
the most part, in accord with the Carnegie Museum
specimens. The arch is high, the spine short and broad and
the diapophyses are directed strongly upward as well as
outward. The arches of dorsals 6 and 7, the first bones
collected and those just in front of the segment of the
column eroded away by the stream, have required some
restoration (e.g. the left diapophysis of dorsal 6 has been
restored in plaster), but they are for the most part intact. As
noted above, the arch and spine of dorsal eight were lost in
a flood and have been completely restored. The remaining
six dorsals are essentially complete, and with the exception
of the minor crushing already noted in dorsal 13, they are
little distorted. Finally, J. F. Bonaparte (personal commu-
nication) has pointed out some differences in the develop-
ment of the laminae of the dorsal arches; in particular, the
presence of V-shaped antero-laterally projecting laminae on
the neural spines of the middle dorsals and a greater
14
mcintosh and williams
No. 43
Fig. 8. Dorsal vertebrae of CMNH 10380 (seen from right side) in the process of being mounted at the
AMNH in New York. A cervical rib was later added to the second (modeled) vertebrae in the series,
leaving 12 dorsal vertebrae in the mount. The last 9 dorsals are completely or in part real and are herein
regarded as the posteriormost 9 dorsals of the animal (i.e. 6-14). See text for additional comments.
Photo courtesy of the American Museum of Natural History.
development of the median laminae on the posterior dorsal
spines of CMNH 10380. He believes these characters may
indicate a new genus “related to Haplocanthosaurus but
different. We believe these differences indicate at most, a
separate species.
Sacrum. As far as can be determined the sacrum is in
almost complete agreement with that of CM 572. The
dorso-sacral and caudo-sacral centra are totally fused to
those of the three primary sacrals. Furthermore, the arches
of all five vertebrae are firmly fused to one another and to
the sacral ribs. As in CM 572 the short to moderately long
spines of sacrals one through three are firmly united
throughout, and those of sacrals four and five are united to
midheight. In CM 572 spines four and five are free, but this
is probably an individual character because in the even
younger CM 879 all five spines are united. As mounted it is
not possible to determine the existence of pleurocentral
cavities. As in CM 572, the spine of the dorso-sacral is
strongly inclined backward and that of the caudosacral
strongly inclined foreward.
Caudal Vertebrae. The first fourteen caudal vertebrae
were found articulated with the sacrum. Other than appear-
ing slightly heavier than those of CM 572 and CM 879 (a
probable age character), they resemble the latter closely
(Fig. 10). The centra are short, amphicoelous, lack side
cavities, and bear prominent chevron facets, particularly on
the posterior end of the centrum. In the first two vertebrae
the diapophyses (more correctly caudal ribs) bear a hint of
the wing-like processes seen on the anterior caudals of the
diplodocids (in contrast to those of Camarasaurus and
Brachiosaurus). Further back they are simple and extend
horizontally, diminishing in size from number one to
fourteen (the last preserved). The prominence of the diapo-
physes on the fourteenth caudal suggests that they were
1988
NEW SPECIES OF SAUROPOD DINOSAUR
15
TABLE 2
Measurements of the Dorsal Vertebrae
No.
length
anterior
breadth
height
Centrum
posterior
breadth
height
overall
height
pre-
zygap.
Height
post-
zygap.
6
e255
158
172
193
238
795
—
—
1
e245
172
193
190
236
875
465
—
8
230
182
215
213
248
—
—
—
9
e255
191
233
214
242
840
—
—
10
e248
211
263
237
274
913
—
—
11
e243
245
263
271
288
935
—
—
12
201
319
254
370
313
915
—
—
13
144
334
298
349
337
1040
—
—
14
170
338
278
349
293
1010
—
590
Note: These measurements were taken with some difficulty from the mounted skeleton. Minor errors may be present.
present on several succeeding vertebrae. This is in contrast
to the situation in CM 572, where the diapophyses have all
but disappeared on caudal thirteen. The first caudal bears a
typical spine which is directed first upward and then curves
around toward the back, resembling that of CM 879 closely.
In CM 572, caudal one was displaced and badly crushed.
When restored by the artist Sydney Prentiss, this curvature
was not indicated. Behind caudal number one the spines are
relatively slender, straight, and directed upward and back-
ward.
Fig. 9. Last 6 dorsal vertebrae of CMNH 10380 as seen from the left
side. Photo courtesy of the American Museum of Natural History.
Ribs. In the mounted skeleton the left ribs attached to
dorsals six through thirteen are at least partially bone. The
other ribs have been restored in plaster. Field diagrams
show that the left ribs nine, ten, and eleven were articulated
TABLE 3
Measurements of the Sacrum of CMNH 10380
Length of five centra 1090
Length of centrum of Sacral 1 202
Anterior breadth of same 301
Length of centrum of Sacral 2 234
Height of Sacral 1 1020
Height of Sacral 2 1010
Height of Sacral 3 1000
Height of Sacral 4 910
Height of Sacral 5 854
Distance between distal ends of ilia externally 640
with their respective vertebrae and that twelve was little
displaced. These ribs are all characteristic of sauropod
dinosaurs and do not exhibit any unusual features.
Chevrons. Several incomplete anterior chevrons were
found with the tail, but were displaced above it. They
appear to be relatively long and straight with the bifurcate
heads relatively short. The heads resemble those of Cama-
rasaurus, in having no bridge of bone above the haemal
canal, and are in contrast to the situation in Diplodocus and
Apatosaurus , where a transverse “crus” exists.
Sternum
One of the most important discoveries with the Cleveland
specimen was that of a complete sternal plate (Fig. 11),
previously unknown in Haplocanthosaurus. Considerable
controversy has existed in the literature concerning the
positioning of these sternal plates. It was partly resolved
—with the discovery in Camarasaurus USNM 13786 and
Alamosaurus USNM 15560— that the heavy pointed end of
the plate was directed forward (Gilmore 1946). This being
the case the element preserved with CMNH 10380 would be
the right sternal plate. It is roughly quadrangular with the
length greatly exceeding the breadth. It is very gently
convex downward. The lateral and medial margins are
parallel and very gently curved to bulge outward from the
16
mcintosh and williams
No. 43
TABLE 4
Measurements of Caudal Vertebrae of CMNH 10380
No.
length
anterior
breadth
Centrum
posterior
height breadth height
overall
height
pre-
zygap.
Height
post-
zygap.
Spread
Diapo-
physes
1
176
346
370
270
370
865
—
—
—
2
155
294
355
226
330
770
—
—
—
3
154
266
340
231
258
728
515
460
—
4
157
289
300
231
280
680
446
430
377
5
165
257
273
220
269
620
424
390
389
6
166
257
272
191
271
578
423
386
392
7
155
217
251
203
244
547
372
372
379
8
155
216
232
210
237
530
369
356
362
9
163
313
239
213
217
491
336
349
*306
10
163
227
204
194
201
454
348
328
—
11
162
207
192
205
193
415
305
278
278
12
163
192
182
188
187
430
279
310
223
13
174
187
175
197
164
359
279
271
e247
14
167
178
147
175
161
323
261
257
231
* = distorted
midline. The medial margin is smooth, but the lateral one is
irregularly rugose for the attachment of the cartilaginous
sternal ribs. The posterior margin also has indentations and
trends slightly forward from the outside in. The thickened
anterior margin is least straight of all. It curves forward to
form a blunted point and then recedes to the median margin.
The plate is relatively much longer than in any other
sauropod except for Alamosaurus and Opisthocoelicaudia.
If and how the sternal plate articulates with the coracoid in
sauropods has not been determined. The sternal bone of
CMNH 10380 more nearly resembles that of Camarasaurus
lentus USNM 13786 than any other, but sauropod sternal
plates show great variation within the same genus, so this
similarity may be fortuitous. This bone does not resemble
the sternal of Camarasaurus grandis YPM 1901. After
many attempts to place the sternal plates reasonably, the
AMNH preparator George Whitaker was unable to settle on
an arrangement that he regarded as satisfactory, and in the
end, they were not placed on the mounted skeleton.
TABLE 5
Measurements of the Sternum of CMNH 10380
Length of sternal plate 755
Breadth of same 358
Pectoral girdle
Scapula. The shaft and distal end of the left scapula are
preserved in CMNH 10380. This is to be compared to the
scapula of CM 879 and another left scapula-coracoid, CM
33995. The latter was miscatalogued CM 94, the number of
the cotype of Diplodocus carnegii. It does not belong to that
animal but rather to Haplocanthosaurus. Both scapulae of
CM 94 are preserved. All the bones found in Quarry D,
Sheep Creek, Wyoming (from which it derives) are repre-
sented on an excellent quarry map, and all are accounted
for. In addition, the color and preservation of CM 33995 is
distinctive of the Garden Park quarry and quite unlike those
of bones from the Sheep Creek quarries. Finally, the quarry
map of the Garden Park quarry (Hatcher 1903c) shows a
complete scapula-coracoid among the bones of CM 572. It
is shaded to show that it belonged to a different genus, but
it is not present elsewhere in the collections in Pittsburgh. It
appears to us that this bone was somehow misplaced and
very likely belongs to CM 33995. If so, Hatcher’s failure to
refer it to the H. priscus skeleton is puzzling. Perhaps he
considered it a bit too large.
The incomplete Cleveland specimen shows the same
widely flared distal end exhibited by CM 879 (Fig. 12). The
general appearance of the bone as restored would appear a
bit more massive than the latter, but this may be a matter of
restoration or possibly of age. The proximal (anterior) plate
is restored in plaster to resemble the “normal” sauropod
scapula. Haplocanthosaurus , like Cetiosaurus, differs from
the norm in having a much less developed anterior plate.
Coracoid? An incomplete girdle bone found near the
sternal plate may be an incomplete coracoid. If so, it is of
little diagnostic value as none of the characteristic features
—the coracoid foramen, the thickened glenoid surface,
etc.— is preserved. It is also possible that the element is the
posterior portion of the left sternal plate. One of the borders
does have a scalloped outline similar to the lateral margin of
the right sternal plate. On the other hand, it would appear to
be slightly wider and perhaps a little thicker than the latter.
Fore Limb
Radius and Ulna. All that remains of the fore limb are the
proximal ends of the left ulna and radius. They were
1988
NEW SPECIES OF SAUROPOD DINOSAUR
17
Fig. 10. Caudal vertebrae of CMNH 10380 seen from the right side. The transition from the first 14
(real) caudals to the modeled ones behind is marked by a dimunition in size of the vertical support rods
beneath each vertebra.
mistaken for parts of the left tibia and fibula and were
incorporated into these restored elements in the mounted
skeleton. These pieces are the only fore limb bones known
to belong to Haplocanthosaurus , but they have lost much of
their value because of their fragmentary condition. The
proximal end of the radius differs from that of most other
18
McIntosh and williams
No. 43
Fig. 11. Right sternal plate of Haplocanthosaurus delfsi. A) Ventral
View B) Dorsal View.
sauropods in that its greatest diameter (345mm) greatly
exceeds its perpendicular one. Viewed from the front, the
medially-directed pointed process is prominent. Little can
be said about the imperfectly preserved head of the ulna
except that the anterior trough, which cradles the end of the
radius, is shallow. As in other sauropods, the proximal end
is V-shaped when viewed from above, but the corners of
both legs of the V are incomplete, particularly that of the
medial branch. This apparently led the restorers to mistake
it for the cnemial crest of the tibia.
Pelvis
The bones of the pelvis and the femur of CMNH 10380
show more differences from those of the type of H. priscus,
CM 572, than do any other of the preserved parts.
Ilia. Both ilia are present and the left one is virtually
complete (Fig. 13). It is more robust than the illium of CM
572. In front of the pubic peduncle, the anterior tip of the
blade is deflected outward in a manner reminiscent of the
Upper Cretaceous titanosaurids, where this feature is
greatly accentuated. No such deflection is observed in
either the left or right ilium of CM 572. Both of the latter
have been crushed but in different directions, and this may
account for the apparent difference between the ilia of the
two animals. The crushing is most severe in the left bone of
CM 572 causing the lobe in front of the pubic peduncle to
appear much longer than it was.
Pubis. The left pubis is nearly complete, and while
resembling that of CM 572 it appears somewhat more
robust, particularly at the distal extremity. A swelling on the
upper portion of the anterior margin represents the attach-
ment for the ambiens muscle, but there is no indication of
the hook-like process that develops here in the diplodocids.
As preserved the pubic foramen was open posteriorly,
leaving an incomplete border for articulation with the
ischium. This is almost surely the result of incomplete
preservation rather than immaturity. In all known sauropod
genera (certainly in all adult specimens), the foramen is
ringed with bone. It has been restored to indicate the closed
condition. The shaft of the pubis is stocky and the distal end
is moderately expanded to meet its opposite member in the
midline, although the two bones are not coalesced.
TABLE 6
Measurements of the Pectoral Girdle
CMNH
10380
CM
879
CM
33995
Length of scapula
el 290
800
1063
Breadth, proximal
--
396
440
Breadth of shaft
—
137
154
Breadth, distal end
497
372
410
Length of coracoid ?
—
295
320
Breadth, greatest
—
350
360
Ischium. The left ischium is complete and, except for the
distal end, it resembles that of CM 572. The pubic articular
margin is arced forward, more so than in other sauropod
genera. The process for articulation with the ilium is
typically long. The shaft is slender but only a little twisted.
It broadens slightly at the distal end while not becoming any
thicker. In CM 572 the broadening is less and the shaft is
more twisted so that the two ischia meet at the distal end
edge to edge as shown in Figure 14. In fact, in CM 572, the
shafts of the two ischia are firmly coosified at their distal
ends and for a considerable distance forward. An identical
arrangement is found in USNM 4275 from the same quarry
as the type and clearly referrable to Haplocanthosaurus. As
both the ischia in CMNH 10380 and CM 572 were found
articulated there can be no question that, indeed, each did
belong to its assigned skeleton. Careful examination of the
distal end of the left ischium of the Cleveland specimen
shows that it was not coosified with that of the missing right
one. Various misalignments of both the pubes and ischia of
this specimen cause the distal ends of the ischia to meet in
a sharp “V” rather than the shallow trough described by
Hatcher (see additional comments in the final section
describing the mounted skeleton). As in Camarasaurus and
Brachiosaurus , none of the three specimens shows any
thickening of the distal end in the perpendicular direction so
typical of the diplodocids, particularly Apatosaurus (Fig.
15).
There is in the collection of the National Museum of
Natural History an articulated sauropod hind limb of the left
side with part of the foot and the associated right femur and
both ischia, USNM 4275. Only the ischia bear the cata-
logue number USNM 4275, but the other elements clearly
1988
NEW SPECIES OF SAUROPOD DINOSAUR
19
Fig. 12. Left scapula-coracoid of CMNH 10380. Only the shaft and upper end of the scapula are bone.
belonged to the same individual and will henceforth be
referred to by the same number. The specimen was found in
the East End of the Marsh-Felch quarry whence came the
types of Haplocanthosaurus priscus and Diplodocus lon-
gus. Marsh arbitrarily assigned the ischia to the D. longus
and figured them as such (Marsh 1896). Recognizing that
these ischia clearly had nothing to do with Diplodocus ,
Gilmore (1907) referred them, “to some large species of
the Morosauridae,” and refigured them as Morosaurus.
They are coalesced for half their length as in CM 572 and
resemble it very closely. Although bearing some resem-
blance to those of Camarasaurus ( Morosaurus ) these bones
certainly belong to Haplocanthosaurus.
Hind Limb
The hind limb in CMNH 10380 is represented by the left
femur only (Fig. 16).
Femur. At first glance, the femur of CMNH 10380 would
appear to show marked differences from that of the holotype
of H. priscus, CM 572. Aside from its smaller size (see
Table 8), the Pittsburgh specimen appears to be more
slender and to have a more circular shaft, although it has
suffered from latero-medial crushing; whereas the Cleve-
land specimen has, if anything, undergone some antero-
posterior flattening.
TABLE 7
Measurements of the Pelvis
CMNH
10380
CM
572
USNM
4275
Length of ilium
1315
827
—
Height of pubic peduncle
955
512
—
Breadth of acetabulum
370
—
—
Length of pubis
1100
693
—
Breadth, proximal end
504
—
—
Breadth, distal end
453
—
—
Length of ischium, distal end to pubic
1000
790
800
articulation
Length of ischium, distal end to iliac
995
—
—
articulation
Least breadth of shaft
144
—
—
Breadth, distal end
236
85
—
Length of articular surface with pubis
e440
—
—
The femur of CMNH 10380 is nearly complete although
an area on the medial side of the lower half of the shaft has
been restored in plaster. It is a straight heavy bone expanded
at both ends, in which the latero-medial diameter consider-
ably exceeds the antero-posterior one. The cross-section of
the shaft is thus a flat oval, most closely resembling
Brachiosaurus in this respect. The fourth trochanter is
20
McIntosh and williams
No. 43
Fig. 13. Left illium and sacrum of CMNH 10380.
located at mid-length on the postero-medial border of the
shaft. The slight swelling on the upper part of the lateral-
border of the shaft, which Galton (1981) has interpreted as
the final vestige of the lesser trochanter, is very weak, unlike
that of Brachiosaurus. The proximal end is broad with the
head directed nearly at right angles to the shaft, and not
rising much above the remnant of the greater trochanter.
The horizontal projection of the head into the acetabulum is
greater than in most sauropods, but that may be due to a
slight deformation of the Cleveland specimen and is not
observed in CM 572. The distal end is moderately expanded
with the tibial condyle exceeding the fibular one in extent.
The comparative measurements given below may be some-
what misleading due to the crushing of CM 572 noted
above.
The hind limb associated with the ischia USNM 4275 has
peculiarities that preclude its reference to any of the other
five Morrison sauropods: Diplodocus , Barosaurus , Apato-
saurus, Camarasaurus , and Brachiosaurus. The femora are
in general agreement with that of CMNH 10380 although
the comparative measurements indicate a slightly more
robust form. All four limb bones have been crushed; the left
femur antero-posteriorly to flatten it, the right one latero-
medially; although they are identical in length and were
lying across one another. The left tibia and fibula were in
position at the lower end of the femur, and the ischia were
also in position. The worth of USNM 4275, important in
itself in providing another specimen of this rare animal, is
further enhanced by providing information about the lower
segment of the leg described below.
Tibia, Fibula, and Pes. No trace of the lower segment of
the hind limb or foot have been preserved with CMNH
10380. As noted previously the proximal ends of the left
radius and ulna have been incorporated into the restored left
tibia and fibula of the mounted skeleton.
In USNM 4275 the left tibia, fibula, and astragalus are
complete but somewhat crushed. Their most notable feature
is their massiveness. The short stocky tibia is expanded at
both ends, but the cnemial crest is only moderately devel-
oped. The distal end is greatly expanded, more so than in
any other Morrison sauropod except Brachiosaurus. The
latero-medial diameter of the shaft greatly exceeds the
antero-posterior one, but this may have been exaggerated by
crushing. The astragalus is in place at the end of the tibia,
but the matrix between the bones has not been removed. As
presently visible, the astragalus presents no noteworthy
features. The left fibula is also stocky. Its proximal end is
unexpanded, the antero-posterior diameter exceeding the
1988
NEW SPECIES OF SAUROPOD DINOSAUR
21
Fig. 14. A) Pelvis of the holotype of Haplocanthosaurus priscus (CM
572) as seen from behind. Note the fusion of the distal ends of the ischia
in the midline. (From Hatcher, 1903c).
Fig. 14. B) Left ishium of Haplocanthosaurus delfsi CMNH 10380,
Length = 1000 mm.
latero-medial one as usual. The lower end is slightly
expanded, the two breadths being subequal. The tibial
articular scar on the medial face at the proximal end is
Fig. 15. Lateral Views of Left Ischia. A) Cetiosaurus, B) Haplocantho-
saurus priscus, C) H. delfsi, D) Titanosaurus , E) Brachiosaurus , F)
Camarasaurus , G) Diplodocus, H) Dicraeosaurus , I) Apatosaurus.
Fig. 16. Left femur of CMNH 10380, posterior view, with Gil Stucker,
Martin Cassidy and George Whitaker. Courtesy of American Museum of
Natural History.
prominent. It terminates below at a bulge on the anterior
margin of the bone. In addition to the stockiness of the tibia
and fibula the most striking feature of the USNM 4275 limb
22
mcintosh and williams
No. 43
is the small tibio-femoral length ratio, .52. Among the other
five North American Jurassic sauropods Brachiosaurus is
closest to this with .59, but the others are over .60. It may
be added that an almost identical right tibia, fibula, and
astragalus (CM 2043) were collected in the Marsh-Felch
Quarry at the same time, but some little distance to the east
of CM 572 and CM 879. These have been assigned to
Haplocanthosaurus by McIntosh (1981). The tibia has a
length of ,627m., which is too short for CM 572, although
it may belong to CM 879.
TABLE 8
Measurements of the Hind Limb
CMNH
10380
CM
572
USNM
4275
left right
Length of femur
1745
1275
1110
1105
Breadth, proximal end
555
353
350
370
Breadth of shaft
—
207
200
195
Breadth, distal end
545
309
315
245
Thickness, tibial condyle
360
—
215
275
Thickness, fibular condyle
245
—
185
255
Least circumference of shaft
755
518
491
502
Ratio, circumference: length
.43
.41
.44
.45
Length of tibia
580
Breadth, proximal end
240
Breadth of shaft
130
Breadth, distal end
256
Least circumference of shaft
321
Ratio, circumference: length
.55
Length of fibula
610
Breadth, proximal end
—
Breadth of shaft
79
Breadth, distal end
142
Least circumference of shaft
218
Ratio, circumference: length
.35
Ratio of lengths, tibia: femur
.52
Ratio of lengths, fibula: femur
.55
Note: caution should be observed in using these measure-
ments as most of the bones have suffered distortions. The
lengths are probably little affected however.
The Species of Haplocanthosaurus
Hatcher referred his two partial skeletons to separate
species based solely on the state of coosification of the
sacral spines. H. priscus CM 572 is about 5% larger than
H. utterbacki CM 879 and as noted previously has all its
vertebral arches firmly coosified to their respective centra,
even the suture line being obliterated. In H. utterbacki
almost all the arches are separate from their respective
centra; the scapula and coracoid are likewise not coalesced.
These characters indicate that CM 879 is a younger animal
than CM 572. However, in CM 572 the spines of the first
three sacrals only (i.e. , the dorso-sacral and primary sacrals
one and two) are coosified for their entire lengths, those of
the last two sacrals (primary sacral three plus the caudo-
sacral) remaining free. In the younger animal, on the other
hand, the spines of sacrals one to four are fused from top to
bottom and even that of number five is fused to the spine in
front of it at the top and bottom. Mook (1917) studied the
problem of speciation in sauropods. In his terminology the
fusion problem is one of “acceleration” in CM 879. In
some unpublished notes he accepted Hatcher’s criterion for
separating the species and noted further a second supposed
difference, namely, the lack of the hyposphene in dorsals
eight through eleven of CM 572 and the presence in CM
879. In their present state of preservation and restoration,
dorsals eight through eleven of CM 572 do indeed appear to
lack a hyposphene and have been so drawn by Sydney
Prentiss, but one must remember that six through fourteen
of CM 572 were found tightly articulated, and the process
of disarticulation, particularly of the hyposphene-
hypantrum, was most difficult. After examining these
specimens, we are convinced that these vertebrae originally
possessed a normal hyposphene. Hatcher himself stated,
“All the dorsals in this region, nos. 6-14 exhibit the
hyposphene-hypantrum articulation” (Hatcher 1903c). The
separation of the species must therefore stand or fall on
Hatcher’s original criterion regarding the union of the
sacral spines. Riggs (1903) made a detailed study of the
sacra of a number of individuals of Apatosaurus (Bronto-
saurus) and Morosaurus (= Camarasaurus) in an attempt to
show that the use of the number of centra fused together in
the sacrum, whether three, four, or five, was an age
character rather than a generic one as employed by Marsh
and others. He also discussed the union of the spines. It is
now generally accepted that he made his point at least at the
generic level. A great deal more study is needed before
distinctive criteria can be stated for the separation of species
in the dinosaurs in general and the Sauropoda in particular.
This must await the full preparation and study of large
population samples from quarries like those at Dinosaur
National Monument, the Cleveland-Lloyd Quarry, and
Como Bluff Quarry 13. There are very few such quarries,
and the problems relating to preparation and study are vast.
Clearly it will be some time before the necessary criteria are
available. As to the problem at hand, we believe that the
variations observed in the coalescence of the various ele-
ments of the sacrum of different individuals showing no
other characters worthy of specific differentiation are inad-
equate grounds for separating H. priscus and H. utterbacki.
The fact that the two skeletons were found only a few feet
from one another in the same quarry at the same level in the
same stratum and are so similar in all other characters adds
weight to this conclusion.
The question remains as to whether CMNH 10380
belongs to H. priscus in light of the apparent differences in
the pelvis and femur discussed above. There is also the
question of the great disparity in size between the two adult
animals CM 572 and CMNH 10380. There are a number of
bones in the National Museum from the younger, Marsh-
1988
NEW SPECIES OF SAUROPOD DINOSAUR
23
Felch Quarry at Garden Park, Colorado, which probably
belong to different individuals of Haplocanthosaurus (evi-
dence to be presented elsewhere). All of these indicate a
species relatively small by sauropod standards, none of
them approaching in any way the size of the gigantic
Cleveland skeleton. As pointed out by Mook (1917), size
may indeed be significant in determining species. There are
also the special characters of the laminae in the dorsal
arches noted earlier. We conclude that these differences,
together with the overall differences in size and robustness
as well as the unfused ischia, indicate that CMNH 10380,
from an older horizon, represents a distinct species, which
we hereby designate H. delfsi after Dr. Edwin Delfs.
Relationships of Haplocanthosaurus to other Sauropods
Noting that the division of the Sauropoda by Marsh
(1896) into six families would probably have to be reduced,
Hatcher ( 1903c) indicated that he accepted at least the three
families Atlantosauridae, Diplodocidae, and Morosauridae,
and that it was to the latter family that Haplocanthosaurus
belonged. He then compared that genus with three British
sauropods and concluded that its relationship was closest to
Cetiosaurus. The following year Riggs (1904) erected a
new family for his recently discovered Brachiosaurus and
assigned Haplocanthosaurus to the Brachiosauridae. The
family characters he chose were 1) fore limbs longer than
hind; 2) vertebral spines simple throughout; and 3) number
of dorsals more than ten. Of course evidence regarding 1)
was not available for Haplocanthosaurus. Some years later
in reviewing the sauropods, von Huene (1929) referred both
Haplocanthosaurus and Brachiosaurus to the Cetiosau-
ridae, which had been established by Lydekker in 1888.
Later classification schemes often place Cetiosaurus and
Haplocanthosaurus together in one subfamily, the Cetio-
saurinae, and Brachiosaurus in a second, the Brachiosau-
rinae, of a single family sometimes called Cetiosauridae,
sometimes Brachiosauridae. We agree that Haplocantho-
saurus should be grouped with Cetiosaurus in the Cetio-
sauridae, but believe Brachiosaurus and its allies have
advanced sufficiently to be grouped in a separate family, the
Brachiosauridae.
To compare Haplocanthosaurus and Cetiosaurus di-
rectly, we note that the skull is not known in either, at least
not the complete one. The cranial fragment USNM 5384
which likely belongs to Haplocanthosaurus cannot be
compared to Cetiosaurus because the corresponding frag-
ment referred by von Huene (1906, 1932) to Cetiosaurus
OUM 13596 does not belong to a sauropod. No teeth are
known in Haplocanthosaurus , but the fragmentary tooth
described by Phillips (1871) found with the Cetiosaurus
oxoniensis skeleton is of the broad spatulate type. The
heart-shaped teeth named by Owen (1840-45) Cardiodon
rugulosus were considered to belong to Cetiosaurus by
Marsh (1896)— a judgment augmented by the discovery of
very similar teeth in the Middle Jurassic Argentine cetio-
saurid Amygdalodon (Cabrera 1947). Similarities between
Haplocanthosaurus and Cetiosaurus suggest that the former
will also be found to have teeth of this sort. Comparison of
the cervicals of Haplocanthosaurus with those of C. oxo-
niensis is difficult because of the incomplete condition of
the latter, but the partial skeleton of Cetiosaurus from
Rutland, England, in the Leicester Museum, reported by
Jones (1970) seems to agree closely with Haplocanthosaur-
us in the simplicity of the lateral cavities and the undivided
neural spines. Further preparation and a detailed study of
the Leicester specimen is being pursued by John Martin and
when completed will allow a more critical comparison. The
dorsal vertebrae of the two genera are also similar, the one
complete dorsal of C. oxoniensis exhibiting the same high
arch, short spine, and having diapophyses directed outward
and upward at 45° (Fig. 17 A, B). The sacrum is unknown
Fig. 17. Anterior Views of Antero-median Dorsal Vertebrae. A) Cetio-
saurus, B) Haplocanthosaurus, C) Brachiosaurus, D) Titanosaurus , E)
Dicraeosaurus, F) Camarasaurus , G) Apatosaurus , H) Diplodocus.
in Cetiosaurus, but the caudals are very similar to those of
Haplocanthosaurus , one complete centrum from Bucking-
hamshire (OUM 13876) showing the same greatly enlarged
chevron facets. The scapula of Cetiosaurus has a reduced
proximal plate and a broadly expanded distal end; however,
the expansion is not as extreme as in that of its American
counterpart. Both genera have relatively broad pubes and
ischia with distal ends not greatly expanded (Fig. 15 A, B).
The ischia of Cetiosaurus from Gloucestershire, England in
the Stroud Museum (Reynolds 1937) compare very favor-
ably with those of CM 572, but less so with that of CMNH
10380 where the distal end is more broadly expanded. The
femora of both animals are moderately broad, show little
evidence of a lesser trochanter, and the head of each rises
little above the great trochanter. Finally, the tibio-
femoralratio is only .6 in Cetiosaurus, which is signifi-
cantly greater than the .52 of the referred specimen of
Haplocanthosaurus. Other genera of the Cetiosauridae
include Rhoetosaurus Longman, Amygdalodon Cabrera,
Patagosaurus Bonaparte, and Shunosaurus Dong, Zhou,
24
mcintosh and williams
No. 43
and Zhang. Preliminary notices relating to the latter two
genera (Bonaparte 1979 and Dong, Zhou, and Zhang 1983)
suggest more primitive forms but a full comparison must
await more detailed descriptions of the vertebrae in those
forms.
Comparison of Haplocanthosaurus with other Morrison
genera shows that it agrees with Brachiosaurus in the
undivided presacral spines (Fig. 17 B, C), but differs in
having shorter cervicals with less complicated pleurocentral
cavities, dorsals with smaller centra, shorter spines in the
anterior region, and differently directed diapophyses. The
ilium is lower in Haplocanthosaurus. With the diplodocids
Diplodocus , Apatosaurus , and Barosaurus the contrast is
very great. These taxa have presacrals with deeply cleft
spines in the shoulder region (Fig. 17 G, H), fewer dorsals
with lower arches, much higher spines, particularly in the
rear, and diapophyses directed horizontally. Their sacra
have much higher spines with the dorso-sacral spine not
coalesced with primary sacral 1. Their caudal vertebrae
contain pleurocentral cavities anteriorly, have far less prom-
inent chevron facets, and these vertebrae are much more
numerous and more elongate in the median and posterior
regions, resulting in a vastly longer tail. The two heads of
the anterior chevrons bridge across in the diplodocids but
not in Haplocanthosaurus, and in the former the median
and posterior chevrons exhibit the peculiar fore and aft
expansion which reaches its extreme in Diplodocus ; those in
Haplocanthosaurus are simple throughout. The scapulae
are quite different in the diplodocids and the distal end of
the ischium is expanded greatly both in breadth and thick-
ness (Fig. 15 G, H, I); finally the tibio-femoral ratio is well
over .60.
The comparison with Camarasaurus is somewhat closer
but here again the divided presacral spines in the latter (Fig.
17 F) are in sharp contrast. Its cervicals have more
complicated pleurocentral cavities than those of Haplocan-
thosaurus; it has two fewer dorsals, 12 instead of 14, its
dorsal centra are larger; the arches are somewhat lower and
the diapophyses are directed outward; however the neural
spines of the posterior dorsals are similar. The caudals in
the two genera resemble one another, but the chevron facets
are much less developed in Camarasaurus , and the poste-
rior caudals are noticeably shorter in Haplocanthosaurus
than in the latter. The chevrons themselves are quite similar,
with no transverse bridge above the haemal canal and no
diplodocoid fore and aft expansion. Both genera have
scapulae with expanded distal ends, and both have ischia
with distal ends little expanded (Fig. 15) and meeting one
another edge to edge. The tibio-femoral ratio in Camara-
suarus is a bit larger than .60; in Haplocanthosaurus it is
apparently smaller.
The Mounted Skeleton
As restored, the skeleton is fully seventy feet long, a large
sauropod (Williams 1982). All preserved elements of
CMNH 10380 have been incorporated into the mounted
skeleton except the sternal plate, the supposed coracoid, and
a few chevrons (Figs. 18, 5). The missing portions have
been modelled or cast from specimens in the American
Museum of Natural History in New York. The skull and
mandible have been modelled in generalized sauropod
fashion. The neck has been restored to contain fourteen
cervicals of which nos. one through four are bone whereas
the other ten are modelled from the Pittsburgh skeletons.
Twelve dorsals have been assigned to the thorax, the first
three modelled from Pittsburgh specimens and the last nine
real. Left thoracic ribs four through eleven as mounted are,
at least in part, real. The others have been restored in
Fig. 18. Diagram of the mounted skeleton. The shaded parts are bone.
plaster. The sacrum and first fourteen caudals are real.
These are followed by several modelled vertebrae and the
tail is completed with casts of an articulated series of
thiry-nine caudals of Camarasaurus AMNH 825. The first
two chevrons are real, all others restored. The left scapula,
excepting the proximal plate, is bone. The right one is
modelled after it and both coracoids are restored. The fore
limbs and feet have been completely restored from speci-
mens in the American Museum, based largely on Apato-
saurus. The complete left pelvis and right ilium as well as
the sacrum are real, the right pubis and ischium having been
modelled from their counterparts. The left femur is real and
the right one modelled from it. The tibiae, fibulae and pedes
have been modelled, the (real) heads of the left ulna and
radius having been incorrectly incorporated into the left
tibia and fibula.
For the most part the pose of this skeleton follows the
pattern of traditional sauropod mounts. The placing of the
scapula was the subject of considerable debate among
sauropod scholars until Gilmore’s (1925) paper on Cama-
rasaurus CM 11338, where the scapula was found articu-
lated and in position for the first time. In the present mount
the scapula is placed somewhat higher on the rib cage, and
its orientation more horizontal than in CM 11338. Unlike
many sauropod mounts the restored fore and hind feet are
constructed with the properly reduced carpus and tarsus and
the single claw on digit I of the manus. The only possible
criticism with the manus is that the five metacarpals are
mounted side by side instead of in circular fashion where the
first and fifth almost meet (Gilmore 1936). In addition,
1988
NEW SPECIES OF SAUROPOD DINOSAUR
25
Gilmore (1932) has shown that in Diplodocus, and probably
the other sauropods as well, the tail proceeds straight out
from the sacrum for a considerable distance before it begins
to descend. As mounted in this specimen, it begins to
descend almost at once and continues to do so sharply. One
final criticism concerns the misorientation of the pubes and
the ischia. As mounted, the medial borders of the pubes are
directed caudally and do not meet ventrally in the midline.
This causes them to appear much broader in side view than
would normally be the case and also causes the ischia to
meet in a sharp “V” rather than a shallow, nearly horizon-
tal trough. In addition, the acetabular borders of the pubes
are not in line with those of the ischia, resulting in an open
acetabulum which appears much larger than it was in life.
This skeleton is one of only two sauropod skeletons to be
mounted in the United States in the last twenty-five years,
the other being the Diplodocus in the Houston Museum of
Natural Science.
Acknowledgments
We wish to thank Dr. Wann Langston and Dr. Philp Currie for
having carefully read the manuscript and for numerous, valuable
suggestions, and Dr. J. F. Bonaparte for valuable observations
concerning the laminae of the dorsal vertebrae. We also wish to
thank Bruce Frumker, staff photographer of the Cleveland
Museum of Natural History and his assistant, Toni Hutton, for
their fine photographic work.
References
Agassiz, L. 1844. Monographien des poissons fossiles du Vieux
Gres Rouge ou systeme Devonien (Old Red Sandstone) des lies
brittaniques et de Russie. Neuchatel xxvi + 171 pp.
Anderson, H. J. 1982. Kentrosaurus Oder Kentrurosaurus. Eine
bemerkung zur Nomenklatur. Geologica et Palaeontologica
15:147.
Anonymous. 1959. Dinosaur remains main guest at museum
dinner. Science News Letter, Sept. 21, 1959, 76:341.
Anonymous. 1966. After 120 million years-on its feet again.
Explorer 8:2, 11-13.
Bonaparte, J. F. 1979. Dinosaurs: a Jurassic assemblage from
Patagonia. Science 205:1377-1379.
Cabrera, A. 1947. Un sauropodo nuevo del Jurasico de Patagonia.
Notas Museo de La Plata 12 Paleontology 95, 1-17.
Delfs, E. 1961. Operation dinosaur. Explorer 3:1,22-24.
Dong, Z. M., S. W. Zhou, and Y. H. Zhang. 1983. The
dinosaurian remains from Sichuan basin, China. Palaeontolo-
gia Sinica Whole No. 162, C No. 23:136 pp. 44 pis.
Galton, P. M. 1981. A juvenile stegosaurian dinosaur, “Astrodon”
pusiilus from the upper Jurassic of Portugal. Journal of Verte-
brate Paleontology 1:3-4, 245-256.
Galton, P. M. 1982. The postcranial anatomy of the stegosaurarian
dinosaur Kentrosaurus from the Upper Jurassic of Tanzania,
East Africa. Geologica et Palaeontologica 15:139-160, 6 pis.
Gilmore, C. W. 1907. The type of the Jurassic reptile Morosaurus
agilis with a note on Camptosaurus. Proceedings of the United
States National Museum XXXII: 151-165, pis. XII-XIII.
Gilmore, C. W. 1925. A nearly complete articulated skeleton of
Camarasaurus. Memoirs of Carnegie Museum X:347-384, pis.
xii-xvii.
Gilmore, C. W. 1932. On a newly mounted skeleton of
Diplodocus in the United States National Museum. Proceedings
of the United States National Museum 81:1-21, 6 pis.
Gilmore, C. W. 1936. Osteology of Apatosaurus. Memoirs of the
Carnegie Museum XI, 4:300 pp. 35 pis.
Gilmore, C. W. 1946. Reptilian fauna of the North Horn Forma-
tion of Central Utah. United States Geology Sur\>ey Professional
Papers 210-C:29-53, pis. 3-14.
Hatcher, J. B. 1903a. A new sauropod dinosaur from the Jurassic
of Colorado. Proceedings of the Biological Society of Washing-
ton XVI: 1-2.
Hatcher, J. B. 1903b. A New Name for the dinosaur Haplocanthus
Hatcher. Proceedings of the Biological Society of Washington
XVI: 100.
Hatcher, J. B. 1903c. Osteology of Haplocanthosaurus, with
description of a new species. Memoirs of the Carnegie Museum
II, 1:1-72, 5 pis.
Hennig, E. 1915. Kentrosaurus aethiopicus, der Stegosauridae
des Tendaguru. Sitzungsberichte der Berliner Gesellschaft
naturforschender Freunde. Berlin. 6:219-247.
Hennig, E. 1916. Zweite Mitteilung iiber den Stegosauriden des
Tendaguru. Sitzungsberichte der Berliner Gesellschaft naturfor-
schender Freunde. Berlin. 6:175-182.
Huene, F. von. 1906. Uber das Hinterhaupt von Megalosaurus aus
Stonesfield, Neues Jahrbuch f ir Mineralogie, Geologie and
Palaeontologie 1:1-12.
Huene, F. von. 1929. Los Saurisquios y Ornitisquoios del Creta-
ceo Argentino. Anales del Museo de La Plata 2a, 3: 196 pp. 44
pis.
Huene, F. von. 1932. Die fossile Reptil-Ordnung Saurischia.
Monographien zur Geologie and Palaeontologie 1,4:361 pp. 56
pis.
Jones, M. D. 1970. Cetiosaurus oxoniensis, Phillips, J.. Middle
Jurassic sauropod from Rutland, England. Transactions of the
Leicester Literary and Philosophical Society 64:144-160.
Lydekker, Richard. 1888. Catalogue of the fossil Reptila and
Amphibia in the British Museum (Natural History). Pt. 1:309
pp.
McIntosh, J. S. 1981. Annotated Catalogue of the Dinosaurs
(Reptilia, Archosauria) in the Collections of the Carnegie
Museum of Natural History No. 18:65 pp.
Marsh, O. C. 1889. Notice of new American Dinosauria. The
American Journal of Science (3) XXXVIL33 1-336.
Marsh, O. C. 1896. The dinosaurs of North America. Sixteenth
Annual Report of the United States Geological Survey 1895, pp.
133-414, pis. II-XXXV.
Mook, C. C. 1917. Criteria for the determination of species in the
Sauropoda, with description of a new species of Apatosaurus.
Bulletin of the American Museum of Natural History
XXXVIIL355-360.
Mook, C. C. 1967. Preliminary description of a new goniopholid
crocodilian. Kirtlandia. 2:1-10.
Owen, R. 1840-1845. Odontography, London.
Pearl, R. M. 1975. The dinosaur menagerie. Earth Science. 28:6,
299-301.
Phillips, J. 1871. The Geology of Oxford. Oxford, Clarendon
Press.
Piel, G. 1963. Last word from the dinosaur. Explorer 5:6, 14-19.
26
McINTOSH AND WILLIAMS
No. 43
Reynolds, S. H. 1936. On a collection of reptilian bones from the
Oolit of Stow-on-the-Wold, Gloucestershire. The Geological
Magazine 76:193-214.
Riggs, E. S. 1903. Structure and relationships of opisthocoelian
dinosaurs. Part I. Apatosaurus. Field Columbian Museum
Publications 82, Geological Series II, 4:165-196, pis. xlv-liii.
Riggs, E. S. 1904. Structure and relationships of opisthocoelian
dinosaurs. Part II. The Brachiosauridae. Field Columbian
Museum Publications 94, Geological Series II, 6:229-247, pis.
lxxi-lxxv.
Romer, A. S. 1966. Vertebrate Paleontology. University of Chi-
cago Press: Chicago and London.
Williams, M. E. 1982. Paleontological displays of the Cleveland
Musuem of Natural History. Fossils Quarterly 1:20-26.
KIRTLANDIA
THE CLEVELAND MUSEUM OF NATURAL HISTORY
Cleveland, Ohio July 1988 Number 43:27-36
ISOLATED TETRAPOD REMAINS FROM THE CARBONIFEROUS
OF WEST VIRGINIA
STEPHEN J. GODFREY
Redpath Museum, McGill University
859 Sherbrooke St. West,
Montreal, Quebec, Canada
H3A 2K6
Abstract
In addition to Greererpeton (Amphibia; Temnospondyli) and Protero-
gyrinus (Amphibia; Anthracosauria), Carboniferous tetrapod fossils from
Greer, West Virginia, include fragmentary remains attributed to Crassigy-
rinus, Eoherpeton, and at least one other taxon, a large temnospondyl
amphibian.
The description of additional tetrapods from Greer demonstrates that in
terms of its taxonomic diversity it resembles other Euramerican Carbonif-
erous localities which span the Visean-Namurian boundary.
28
GODFREY
No. 43
Abstract
In addition to Greererpeton (Amphibia; Temnospondyli) and Protero-
gyrinus (Amphibia; Anthracosauria), Carboniferous tetrapod fossils from
Greer, West Virginia, include fragmentary remains attributed to Crassigy-
rinus, Eoherpeton, and at least one other taxon, a large temnospondy!
amphibian.
The description of additional tetrapods from Greer demonstrates that in
terms of its taxonomic diversity it resembles other Euramerican Carbonif-
erous localities which span the Visean-Namurian boundary.
Introduction
The first vertebrate remains recovered from Carbonifer-
ous deposits within a limestone quarry operated by the
Greer Limestone Company near Greer, West Virginia, were
discovered in 1948 by an amateur, L. R. Collins, (Hotton
1970). The quarry lies just north of State Route 7, about
10.5 km southeast of Morgantown, Monongalia County
(Universal Transverse Mercator Grid, Zone 17 NJ 991809).
More detailed descriptions of the locality are provided by
Hotton (1970) and Romer (1969, 1970). In July of 1968, J.
J. Burke and W. E. Moran discovered, in a slab which had
fallen from the quarry wall, the incomplete remains of a
new temnospondyl amphibian, Greererpeton burkemorani
Romer, 1969. By 1969, this quarry was recognized as
yielding some of the oldest non-ichthyostegalian tetrapods.
In 1970, A. S. Romer and N. Hotton III published
descriptions of the new anthracosaurian amphibians from
Greer, Proterogyrinus scheelei and Mauchchunkia bassa
respectively. Because no consistently unique features dis-
tinguish the latter from the former. Holmes (1984) con-
cluded that Machchunkia was the junior synonym of Pro-
terogyrinus.
Between 1969 and 1973, field parties from the Cleveland
Museum of Natural History under the direction of D. H.
Dunkle recovered a large number of vertebrate skeletons
from a restricted bed within the quarry (Fig. 1). The main
bone-bearing layer, exposed for some 15m along the vertical
face of the quarry, contained a nearly solid mass of fish and
amphibian remains. Greererpeton, represented by at least
60 individuals, was by far the most common vertebrate
recovered. Several virtually complete specimens were pre-
served literally head-to-tail (Godfrey 1986). This deposit
also produced articulated skeletons of the lungfish Tranodis
castrensis Thomson and articulated skeletons of the anthra-
cosaur Proterogyrinus scheelei Romer.
In addition to these spectacular finds, the Cleveland
Museum of Natural History (CMNH) and the Museum of
Comparative Zoology (MCZ) recovered numerous spines
of the acanthodian Gyracanthus, one shark spine (M. E.
Williams, CMNH, personal communication), disarticu-
lated remains of palaeoniscoid fishes, isolated elements of a
large rhizodont crossopterygian, cf. Strepsodus, and iso-
lated postcranial remains of several heretofore undescribed
tetrapods. This paper deals with only the undescribed
tetrapod remains.
Vertebrate remains at Greer were recovered from the
Bickett Shale (Bluefield Formation), which forms the lower
subdivision of the Mauch Chunk Group (Busanus 1974).
The Bickett Shale is lithologically the most variable unit
within the Bluefield Formation, exhibiting rapid lateral
changes in lithology. This nonmarine unit consists primarily
of red and green mudstones but medium-grained sandstones
—which have been interpreted as fluvial point bar and
overbank deposits— are also present (Busanus 1976). The
tetrapod remains (which include fragments of Greererpeton
and Proterogyrinus ) recovered by the Museum of Compar-
itive Zoology were apparently found about lm below the
main vertebrate bearing bed quarried by the Cleveland
Museum (C. Scaff, MCZ, personal communication). It is
not known whether the fragmentary tetrapod remains col-
lected by the Cleveland Museum were removed from the
main layer or from the ‘bone bed’ below it.
Fig. 1. Quarrying operations by members of the Cleveland Museum of
Natural History field party in a section of the Greer Limestone Company
quarry, Monogalia County, West Virginia. Fossiliferous rock within the
Bickett Shale is being removed from the West wall of the quarry. Circa
1970.
Kirtlandia, No. 43, July 1988 © by the Cleveland Museum of Natural History
1988
TETRAPOD REMAINS FROM GREER, W.VA. 29
Because Visean and Namurian tetrapods are exceedingly
rare, a description of this material is important for it
substantiates a more diverse assemblage than previously
recognized at Greer. Recent descriptions of Scottish Car-
boniferous tetrapods by Panchen (1985) and Smithson
(1985a, 1985b) permit a more precise taxonomic assign-
ment for some of the tetrapod remains from Greer. Com-
paring taxonomic diversity indicates that Euramerican
Lower Carboniferous and basal Upper Carboniferous tetra-
pod assemblages are similar.
I have opted to refrain from erecting formal taxonomic
names because to do so on such fragmentary material would
be premature. Although it seems unlikely, the isolated
remains could represent elements from five different
species.
Description
Vertebrae
CMNH 11239 (Figs. 2-5) consists of two large rhachit-
omous vertebrae. Both vertebrae conform to the typical
rhachitomous pattern, being composed of a relatively mas-
sive intercentrum and paired pleurocentra that bear large
facets to support the neural arch. These vertebrae are
clearly distinct from the essentially embolomerous verte-
brae of Proterogyrinus, and their size and proportions
preclude the possibility that they pertain to Greererpeton.
The three intercentra preserved in CMNH 11239 are
massive and remarkably similar to those of Eryops
(Moulton 1974) (Fig. 2, A-D). Ventromedially, a flat
longitudinal ridge is only weakly developed and may
represent the zone of fusion between paired anlagen. In
lateral view, the ventral margin of the intercentrum is gently
concave. The median ridge is flanked by two shallow
concavities formed by periosteal bone that is deflected along
the anterior and posterior margins of the intercentrum
(Fig. 2, A and D). The periosteal bone is perforated by
minute pits.
The concave ventrolateral surfaces vanish dorsally as the
periosteal bone gathers to form a small semicircular facet
that articulates with the capitulum of the rib (Fig. 2A). The
parapophyses project little beyond the margin of the bone
but can be seen in anterior or posterior view. A narrow strip
of finished bone curves around each parapophysis to termi-
nate on the apex of the wedge. On the two anterior-most
intercentra, the facet for the capitulum of the rib is very
small and only weakly developed (Fig. 2D).
The anterior, posterior, and dorsal (notochordal) surfaces
are rough and unfinished. Because the intercentrum is very
thick mediolaterally, the notochordal notch is relatively
small. In typical rhachitomous fashion, the massive
intercentra from adjacent segments abut ventromedially
(Fig. 5 A).
Fig. 2. CMNH 11239, elements of the centrum. A-C, a thoracic
intercentrum in left lateral, anterior, and ventral views respectively; D, a
cervical intercentrum in left lateral view; E-F, a pleurocentrum half in
posterior and anterior views respectively. Scale bar equals 1 cm.
In life, the pleurocentra of CMNH 11239 were situated
above the notochord and the main ossified portion of the
intercentrum (Figs. 2 E-F and 5 A-B). In lateral view, each
pleurocentrum half is an elongated oval that bears a protu-
berant semicircular articular facet on its anterodorsal mar-
gin that received the pedicle of the neural arch. This facet of
unfinished bone forms an angle of approximately 60 ° with
a plane tangential to the dorsolateral curvature of the
external surface of the pleurocentrum. The external surface
below and around the posterior margin of the facet is
finished with smooth bone. Ventrally, the pleurocentrum
ends in a blunt point. In anterior or posterior view the
pleurocentrum is gently curved to surround the noto-
chordal space. The internal surface is unfinished. Opposing
antimeres of one pleurocentrum probably abutted above
the notochord and immediately below the spinal cord
(Fig. 5 B).
30
GODFREY
No. 43
Fig. 3. CMNH 11239, A-C, specimen drawings of a ‘thoracic’ neural
arch in left lateral, anterior, and dorsal views respectively. Scale bar equals
1 cm.
The two neural arches preserved in CMNH 11239 differ
in the structure of the neural spine and pedicle (Figs. 3 and
4). The distal end of one of the neural spines is greatly
expanded laterally and markedly rugose (Fig. 4), whereas
the other is much more slender and laterally compressed
(Fig. 3). Based on the structure of the neural arches in
Greererpeton (Godfrey 1986), laterally expanded neural
spines and poorly developed diapophyses occur in the
cervical series, whereas narrow, laterally compressed
spines and longer transverse processes characterize more
posterior vertebrae. Because both spines are unfinished
dorsally, each was probably capped by cartilage in life.
One of the more remarkable features of both arches is the
retention of a large supraneural space (Figs. 3 B and 5 B).
The space remains open for half the height of the spine as
measured from the base of the postzygapophyses. A supra-
neural space (canal) is primitive for tetrapods and is thought
to have accommodated a supraneural ligament (Smithson
1985a). In anterior view, the neural spine is roughly the
shape of an inverted V. On the posteromedial surfaces of
each spine, immediately above the postzygapophysis, is a
conspicuous groove that passes dorsally to terminate where
the arch halves meet above the supraneural space.
Both pre- and postzygapophyses are roughly ovoid in
outline and inclined very little from the horizontal. Imme-
diately below each prezygapophysis is a deep pit which may
have led to a foramen (Fig. 3 B). This feature is not known
in any other early tetrapod.
Both arches possess large, broadly oval to rectangular
facets on their ventromedial surfaces to receive the articular
facets of the pleurocentrum. Immediately above the lateral
margin of the pedicle is a narrow strip of periosteal bone.
Above this strip the transverse process is poorly developed.
In the cervical neural arch, the diapophyses are almost
non-existent and probably did not hold the tuberculum of a
rib in life (Fig. 4 A and B). On the other neural arch,
protuberant flange-like diapophyses are developed. Judging
from the relatively small size of these narrow, anteroven-
trally directed facets, the tuberculum must have been small.
Fig. 4. CMNH 1 1239, A and C, specimen drawings of a cervical neural
arch in left lateral and dorsal views respectively; B and D, restorations of
the arch in lateral and dorsal views respectively. Scale bar equals 1 cm.
To summarize: although the vertebrae of CMNH 11239
resemble superficially those of Eryops, they differ from that
genus in the possession of the following features: 1) a large
supraneural space; 2) a relatively short neural spine; 3)
widely separated contralateral pre- and postzygapophyses;
and 4) a deep pit or foramen below each prezygapophysis.
These vertebrae are probably derived from a large rhachi-
tomous temnospondyl amphibian.
1988
TETRAPOD REMAINS FROM GREER, W.VA.
31
Fig. 5. Restoration of vertebrae based on CMNH 11239; A, left lateral view; and B, posterior view.
Scale bar equals 1 cm.
Ribs
CMNH 1 1230 consists of a consecutive series of 10 or 11
large but incompletely preserved ribs lying on the visceral
surface of an articulated series of ventral scales (Fig. 6).
Although no rib heads are preserved, the morphology of the
gastralia indicates that the right side of the body is repre-
sented and that the narrow end of the slab in Figure 6 is the
anterior end. The ribs lack characteristic features of cervi-
cal, sacral, or caudal series and thus probably represent a
segment from the midtrunk region. Although all the ribs are
broken or variably crushed, they retain much of their
original curvature. Because the radius of curvature of the
eighth rib is just over 7 cm, the trunk diameter at that point
was probably 15 cm or more if rib curvature followed that
of the body. All the ribs are ornamented with longitudinal
striations as in Crassigyrinus (Panchen 1985). The rib
shafts lack an uncinate process or flange found in some
temnospondyls. The large size of the ribs and the morphol-
ogy of the gastralia indicates that they do not pertain to
Proterogyrinus.
Although some of the scales are scattered, articulated
series show that the original pattern was en chevron (Fig. 7
A). The scales are shaped much like an elongate tear drop.
The anterolateral margin of each scale is more conspicu-
ously convex than the trailing posteromedial edge. Their
ventral surface is smooth and convex whereas their visceral
surface is concave. A rounded ridge (Fig. 7 B) runs the
length of the scale along the posterodorsal (internal) sur-
face. In ventral view, the medial margin of each scale
underlaps the trailing edge of its medial neighbor. The
scales in CMNH 11230 resemble those of Crassigyrinus.
Pending further discoveries, CMNH 11230 is tentatively
assigned to the genus Crassigyrinus.
Scapulocoracoid
MCZ 8950 (Fig. 8) is a single incompletely preserved
right scapulocoracoid. The thin coracoid plate appears to be
virtually complete, whereas the margin of the scapular
blade anterodorsal to the supraglenoid buttress is poorly
represented. The position of foramina and its overall pro-
portions preclude it from belonging to either Greererpeton
or Proterogyrinus (Fig. 8, D-F).
The glenoid is fairly well developed and forms a concave
horizontal strap of unfinished bone that shows very little
32
GODFREY
No. 43
Fig. 6. CMNH 11230, 10 or 11 incomplete ribs and ventral scales from the right side of the body in
dorsal view. Two ribs, 8 and 9 in the series (counting from the narrow end of the slab), were fractured
in life and show a swollen mass of ankylosed bone distally. Scale bar equals 1 cm.
torque along its surface. Except for the anterior one-third of
the glenoid, which is directed postero-ventrolaterally, the
fossa is directed laterally.
Above the anterodorsal margin of the glenoid on the
posterodorsal-facing margin of the ascending scapular
blade, a small boss is developed that is visible in lateral
view (Fig. 8 A and B). Extending posteriorly from the base
of this small tuberosity, along the surface immediately
medial to the glenoid, is a small ridge which ends on the
posterior margin of the bone. A second, more prominent
ridge lies medial and parallel to this ridge and continues
from the posterior margin of the bone anterodorsally onto
the supraglenoid buttress, where it loses prominence.
In medial view, the supraglenoid buttress thins rapidly
anterior to the infraglenoid buttress. The ascending scapular
moiety is not preserved. The supraglenoid buttress is not
perforated by the supraglenoid foramen, but a C-shaped
notch (Fig. 8 A), immediately anterior to the supraglenoid
buttress and level with the glenoid articulation, may repre-
sent the remnants of this foramen. The foramen passes
through the 3 mm thick bone in a ventrolateral direction.
Alternatively, this foramen might represent the supracora-
coid foramen, and the supraglenoid foramen may have been
lost. A second large foramen is visible laterally, just
Fig. 7. Restoration of gastralia in ventral view based on CMNH 11230
(the narrow end of each scale points medially and the anterior end of the
body lies towards the top right-hand corner of the page); and B, one scale
enlarged, in dorsal view. Scale bar equals 1 cm.
1988
TETRAPOD REMAINS FROM GREER, W.VA.
33
beneath the anterior margin of the glenoid. It passes
through the scapulocoracoid in a dorsomedial direction to
open on the medial surface at the base of the supraglenoid
buttress. The homology of this foramen is difficult to
determine but, as in Eoherpeton (Smithson 1985a), it could
represent the common opening of the supracoracoid and
glenoid foramen.
The unfinished bone forming the thin ventromedial
margin of the coracoid plate was probably finished by
cartilage in life. Both medial and lateral surfaces of the plate
are perforated by numerous ‘nutritive’ foramina. The
antero-ventromedial portion of the scapulocoracoid is per-
forated by another fairly large foramen which passes
vertically through the bone. Its location is almost identical
to the anterior coracoid foramen seen in Protorothyris
archeri (Clark and Carroll 1973, Figs. 5 A and 6 F).
Humeri
MCZ 8951 (Fig. 9) is an incomplete right humerus that
possesses several features that are not present in Greerer-
peton or Proterogyrinus.
On the dorsal surface of the humerus immediately above
the proximal articulation is a conspicuous boss or tubercle.
Immediately behind the tubercle is a concavity which may
have marked the insertion of m. subcoracoscapularis. The
processus latissimus dorsi is damaged in MCZ 8951. The
ectepicondyle ridge begins at the base of the processus
latissimus dorsi and broadens as it curves gently towards the
anterodistal corner of the bone. Distally, the dorsal margin
of the condyle was damaged and, therefore, the exact height
of the ridge is not known.
A small posteriorly directed flange of bone is present on
the postaxial surface of the humerus (Fig. 9 A and C).
Aside from Eoherpeton (Smithson 1985a), this flange has
not been described on the humerus of other early tetrapods;
consequently, its function remains unclear. Between the
aforementioned flange and the processus latissimus dorsi is
a shallow longitudinal trough which deepens distally and
marks the postaxial base of the ectepicondyle.
The deltopectoral crest in MCZ 8951 is remarkably
large. Anteroventrally, the crest exhibits a shallow pit,
above which lies a large tubercle. The pit and tubercle mark
the major insertion of m. pectoralis and m. deltoideus
respectively. A postaxially directed ridge extends from the
deltoid tubercle across the dorsal surface of the humerus to
merge with the anterior end of the ectepicondyle ridge at the
base of the processus latissimus dorsi (Fig. 9 A). As in
some early amniotes (Romer 1956), this ridge presumably
marks the insertion of m. scapulohumeralis anterior.
The proximoventral surface of MCZ 8951 displays two
prominent areas of muscle attachment. A pronounced
preaxial tubercle marks the insertion of m. supracoracoid-
eus and, more posteriorly, a broad concavity marks the
insertion of m. coracobrachialis. The area of insertion of
these and other muscles on MCZ 8951 is based on the
description of other more completely known tetrapods
Fig. 8. MCZ 8950, A-C, specimen drawings of an incomplete right
scapulocoracoid in lateral, medial, and posterior views respectively.
Partially restored right scapulocoracoid of D, Greererpeton burlcemorani;
E, Proterogyrinus scheelei; and F, MCZ 8950, all drawn to the same scale.
P. scheelei after Holmes (1980). Scale bars equal 1 cm. Abbreviations:
anterior coracoid foramen, f ant cor; supracoracoid foramen, f scor;
supraglenoid buttress, sgl bt; supraglenoid foramen, f sgl.
(Romer 1956; Holmes 1980; Smithson 1985a).
The supinator flange and entepicondyle are incomplete in
MCZ 8951. The preserved section of the entepicondyle
differs from that seen in Eoherpeton in that it extends
distally as a smooth, broadly rounded flange beyond the
trochlea (ulnar articulation), much as it does in Eryops
(Miner 1925). The distal and postaxial margins of the
entepicondyle are finished with periosteal bone that is
rounded from dorsal to ventral surfaces. The entepicondyle
foramen passes from the proximal anterodorsal surface of
the entepicondyle in a ventrodistal direction. In ventral
view, the foramen exits below a distinct ridge that sweeps
34
GODFREY
No. 43
Fig. 9. MCZ 8951, A-D, specimen drawings of an incomplete right
humerus in dorsal, preaxial, ventral, and proximal views respectively.
Scale bar equals 1 cm. Abbreviations: deltopectoral crest, delt cr;
ectepicondyle, ect; entepicondyle foramen, f ent; m. coracobrachialis, cbl;
m. scapulohumeralis anterior, scha; m. subcoracoscapularis, sbcsc; m.
supracoracoideus, spec; processus latissimus dorsi, pr lat d; supinator
flange, sup.
postaxially along the proximal margin of the entepicondyle.
As the entepicondyle is incomplete, the extent of this ridge
is unknown. A shallow groove extends in an anterodistal
direction across the ventral surface of the entepicondyle
from the entepicondyle foramen to vanish below the ulnar
articulation (Fig. 9 C). Radial and ulnar condyles are
poorly defined. In MCZ 8951, the long axis of the proximal
articulation forms an angle of 45-50 ° with the plane of the
entepicondyle (Fig. 9 D).
Although somewhat smaller, MCZ 8951 bears a striking
similarity to the humerus of Eoherpeton as described by
Smithson (1985a). Minor differences between the two may
be attributed to size. Among Lower Carboniferous tetra-
pods, only Eoherpeton approaches MCZ 8951 in the large
number of conspicuous areas for muscle attachment. MCZ
8951 is provisionally assigned to this genus.
Fig. 10. MCZ 8952, A-E, specimen drawings of a left humerus in
preaxial, dorsal, slightly ventral of the postaxial surface, ventral, and
proximal views respectively. Scale bar equals 1 cm. Abbreviations as in
Fig. 9.
Although the second undescribed humerus from Greer
(MCZ 8952, Fig. 10) is of the tetrahedral type common to
most early tetrapods, it is unlike other humeri from this
locality in that it is a more robust element with a higher
degree of torsion and is without most of the tuberosities or
ridges marking the origin or insertion of limb musculature.
1988
TETRAPOD REMAINS FROM GREER, W.VA.
35
If the plane of the entepicondyle is positioned horizon-
tally, the unfinished convex strap of bone forming the
proximal articulation is nearly vertical and remarkably
compressed anteroposteriorly (Fig. 10 E). The articular
surface is poorly defined. A short distance behind the
postaxial margin of the proximal articulation is a small
tuberosity, probably marking the insertion of m. subcora-
coscapularis. This tuberosity forms a ridge that continues
posterodistally to form the anterodorsal margin of the
entepicondyle. On the preaxial side of this tuberosity, a
shallow groove passes distally and becomes progressively
more pronounced as it undercuts the ridge that forms the
ectepicondyle. The posterodorsally projecting ectepi-
condyle is unfinished.
The deltopectoral crest in MCZ 8952 is poorly devel-
oped, and the insertions of m. pectoralis and m. deltoideus
are not defined. A small ridge passes postaxially from the
posteroventral corner of the deltopectoral crest onto the
entepicondyle above the entepicondyle foramen. Anterior to
this ridge the proximoventral surface of the bone in MCZ
8952 shows two V-shaped concavities that are separated by
a second low, rounded ridge. The anterior-most concavity is
larger and lies immediately postaxial to the deltopectoral
crest. The second concavity, with the top of its V-shaped
margin directed proximally, lies adjacent to the small
tuberosity that may represent the insertion of m. subcora-
coscapularis. The deltopectoral crest and the supinator
process are continuous, forming a large preaxial flange.
The dorsal preaxial surface of the supinator flange bears a
bulbous tubercle that may have held the origin of m.
supinator.
The dorsal surface of the entepicondyle bears a broad but
low ridge that extends proximodistally. This ridge increases
substantially the thickness of the entepicondyle distally and
presumably accommodated the radial, ulnar, or both radial
and ulnar condyles.
MCZ 8952 displays a remarkably high degree of longi-
tudinal torsion. The long axis of the proximal articulation
forms an angle of about 64° with the plane of the entepi-
condyle.
Discussion
The description of several additional tetrapods from
Greer indicates that this assemblage is more diverse than
previously recognized. In terms of its taxonomic comple-
ment, the Greer tetrapod fauna resembles two localities
from the Midland Valley of Scotland and three North
American localities (Table 1). The faunal similarities be-
tween Greer, Gilmerton (Smithson 1985b), Keokuk County
(McKay et al. 1987), Cowdenbeath (Smithson 1985b,
1986), Point Edward (Carroll et al. 1972), and Hinton
(Smithson 1982 and personal communication) are not all
that surprising. All Visean and Namurian tetrapods are
derived from a narrow equatorial strip along the southern
region of ‘pre-drift’ Euramerica (Milner et al. 1986). These
sites extend from present-day Iowa in the west to the Federal
Republic of Germany in the east and demonstrate that
known localities are remarkably restricted latitudinally.
Their faunas were dominated by large, primarily aquatic
tetrapods.
TABLE 1
Distribution of Euramerican
Amphibians across the Visean-Namurian Boundary
STAGE
<
X
LOCALITY eS
Crassigyrinus
Loxomma
Spathicephalus
Eoherpeton
Doragnathus
Proterogyrinus
Pholiderpeton
Undescribed
Anthracosaur
Pholidogaster
Greererpeton
Large
Temnospondyl
Adelogyrinidae
c
.2
Hinton, West Virginia
*9
*?
*?
*9
*
p
£
Point Edward, Nova Scotia
*
*
*9
p
Z
Cowdenbeath, Fife Region
*
*
*
*
*
*
G
03
Greer, West Virginia
*9
*9
*
*
*
-
CO
Keokuk County, Iowa
*
*?
>
Gilmerton, Lothian Region
*
*
*
*
The ventral surface of the entepicondyle is conspicuously
concave, with its deepest point marked by the opening of the
entepicondyle foramen. The foramen passes from the ven-
tral suface in a posterodorsal and slightly medial direction
and opens near the proximal margin of the entepicondyle.
Acknowledgments
I am indebted to Michael E. Williams (Cleveland Museum of
Natural History) and Farish A. Jenkins and Charles R. Schaff
(Museum of Comparitive Zoology, Harvard University), for the
36
GODFREY
No. 43
loan of specimens that formed the basis of this paper. Collection
and initial preparation of the material from the Cleveland Museum
were supported by the U.S. National Science Foundation under
grant G.B.— 35474.
Many comments from R. W. Hook (Austin, Texas), J. R. Bolt
(Field Museum of Natural History), anonymous reviewers, T. R.
Smithson (University of Newcastle-upon-Tyne), R. L. Carroll
(Redpath Museum, McGill University), and D. W. Dilkes (Uni-
versity of Toronto) improved the contents of this paper. The
assistance of D. Baird (Princeton University) and the late J.
Moulton in the early stages of this work is also gratefully
acknowledged.
This research was supported by the Natural Sciences and
Engineering Research Council of Canada, and the National
Science Foundation of the United States of America.
References
Busanus, J. W. 1974. Paleontology and paleoecology of the
Mauch Chunk Group in Northwestern West Virginia. Unpub-
lished M. S. thesis. Bowling Green State University, Bowling
Green Ohio.
Busanus, J. W. 1976. Faunal distribution within the Greenbrier-
Mauch Chunk transition (Chesterian: Elviran) along the Chest-
nut Ridge Anticline. Geological Society of America, Abstracts
with Programs 8(2): 143.
Carroll, R. L., E. S. Belt, D. L. Dineley, D. Baird, and D. C.
McGregor. 1972. Vertebrate Paleontology of Eastern Canada.
Excursion A59, 24th International Geological Congress.
Clark, J., and R. L. Carroll. 1973. Romeriid reptiles from the
Lower Permian. Bulletin of the Museum of Comparative Zool-
ogy 144:353-407.
Godfrey, S. J. 1986. The skeletal anatomy of Greererpeton
burkemorani Romer 1969, an upper Mississippian temno-
spondyl amphibian. Unpublished Ph.D. dissertation, McGill
University, Montreal.
Holmes, R. 1980. Proterogyrinus scheelei and the early evolution
of the labyrinthodont pectoral limb. In The Terrestrial Environ-
ment and the Origin of Land Vertebrates. Edited by A. L.
Panchen. Systematics Association Special, 15:351-376. Lon-
don, Academic Press.
Holmes, R. 1984. The Carboniferous amphibian Proterogyrinus
scheelei Romer, and the early evolution of tetrapods. Philosoph-
ical Transactions of the Royal Society of London, (B) 306:431-
527.
Hotton, N. 1970. Mauchchunkia bassa, gen. et. sp. nov. , an
anthracosaur (Amphibia; Labyrinthodontia) from the Upper
Mississippian. Kirtlandia 12:1-38.
McKay, R. M., B. J. Witzke, M. P. McAdams, and J. R. Bolt.
1987. Early tetrapods and deposition of the Upper St. Louis
Formation (Mississippian; Visean) in Keokuk County, Iowa.
No. 129575. Geological Society of America, Abstracts with
Programs 19(4):233.
Milner, A. R., T. R. Smithson, A. C. Milner, M. I. Coates, and
W. D. I. Rolfe. 1986. The search for early tetrapods. Modem
Geology 10:1-28.
Miner, R. W. 1925. The pectoral limb of Eryops and other
primitive tetrapods. American Museum of Natural History,
Bulletin 51:145-312.
Moulton, J. M. 1974. A description of the vertebral column of
Eryops based on the notes and drawings of A. S. Romer.
Breviora 428:1-44.
Panchen, A. L. 1985. On the amphibian Crassigyrinus scoticus
Watson from the Carboniferous of Scotland. Philosophical
Transactions of the Royal Society of London, (B) 309:505-568.
Romer, A. S. 1956. Osteology of the Reptiles. The University of
Chicago Press, Chicago and London.
Romer, A. S. 1969. A temnospondylous labyrinthodont from the
Lower Carboniferous. Kirtlandia 6:1-20.
Romer, A. S. 1970. A new anthracosaurian labyrinthodont,
Proterogyrinus scheelei, from the Lower Casrboniferous. Kirt-
landia 10: 1-16.
Smithson, T. R. 1982. The cranial morphology of Greererpeton
burkemorani Romer (Amphibia: Temnospondyli) . Zoological
Journal of the Linnean Society 76:29-90.
Smithson, T. R. 1985a. The morphology and relationships of the
Carboniferous amphibian Eoherpeton watsoni Panchen. Zoo-
logical Journal of the Linnean Society 85:317-410.
Smithson, T. R. 1985b. Scottish Carboniferous amphibian local-
ities. Scottish Journal of Geology 21:123-142.
Smithson, T. R. 1986. A new anthracosaur amphibian from the
Carboniferous of Scotland. Palaeontology 29:603-628.
Membership and Subscription
Membership in Kirtlandia Society is open to any member of the Cleveland
Museum of Natural History. Current individual membership is $30.00 ($10.00 for
students); participating members, $100.00; institutional, $30.00; sustaining,
$250.00. Applications and fees should be sent to the Treasurer, Kirtlandia Society,
Cleveland Museum of Natural History, Cleveland, Ohio 44106.
Membership includes subscription to Kirtlandia. There is a non-U. S. postal
surcharge of $5.00. All subscriptions outside North America will be sent by air.
Payment must be made in U.S. dollars. Payments not drawn on a U.S. baank will be
subject to bank collection charges.
Change of Address: Please send to Museum six weeks prior to move; enclose
present mailing label with change of address.
Claims for missing issues: Claims cannot be honored beyond four months after
mailing date. Duplicate copies cannot be sent to replace issues not delivered because
of failure to notify Museum of change of address.
Back issues: Individuaal issues of Kirtlandia are available from the Museum
Publications Office. A complete list is available upon request.
•NATURAL HISTORY*
Anthropology
A Method for Making Three-dimensional Reproductions of Bones and Fossils 3
Jenny A. Smith and Bruce M. Latimer
Archaeology
The Squaw Rockshelter (33CU34):
A Stratified Archaic Deposit in Cuyahoga County 17
David S. Brose
Skeletal Remains from Squaw Rockshelter 55
Fred Prior
Knotted Cordage from Squaw Rockshelter (33CU34),
Aurora Run, Cuyahoga County 59
R. L. Andrews and J. M. Adovasio
MARCH 1989
KIRTLANDIA
The Scientific Publication of The Cleveland Museum of Natural History
David S. Brose, Editor
Joseph T. Hannibal, Assistant Editor
Brief History and Purpose
Kirtlandia , a publication of The Cleveland Museum of Natural History, is named
in honor of Jared Potter Kirtland, a noted nineteenth-century naturalist who lived
in the Cleveland, Ohio area. It began publication in 1967 and is a continuation of
the earlier series Scientific Publications volumes I to 10 (1928-1950), and new
series volumes 1 to 4 (1962-1965). Supported by the Kirtlandia Society of The
Cleveland Museum of Natural History, Kirtlandia is devoted to the publication of
scientific papers in the various fields of inquiry within the Museum’s sphere of
interest: Cultural and Physical Anthropology; Archaeology; Botany; Geology;
Paleobotany; Invertebrate and Vertebrate Paleontology; Systematic Ecology; and
Invertebrate and Vertebrate Zoology. Issues will vary from single monographs to
collections of short papers, review articles, and brief research notes.
Authorization to photocopy items for internal or personal use, or the internal or
personal use of specific clients, is granted for libraries and other users registered
with the Copyright Clearence Center (CCC) Transactional Reporting Service,
provided that the base fee of 50 cents per copy, plus 25 cents per page is paid
directly to CCC, 27 Congress Street, Salem, MA 09179. The fee code for users of
the Transactional Reporting Service is 0075-6245/89 $00.50+. 25.
Kirtlandia is abstracted in Biological Abstracts and indexed in Bibliography and
Index of Geology and Zoological Record.
Kirtlandia No. 44
© 1989 by The Cleveland Museum of Natural History
Cleveland, Ohio
KIRTLANDIA
The Cleveland Museum of Natural History
March 1989 Number 44
Anthropology
A Method for Making Three-dimensional Reproductions of Bones and Fossils 3
Jenny A. Smith and Bruce M. Latimer
Archaeology
The Squaw Rockshelter (33CU34):
A Stratified Archaic Deposit in Cuyahoga County 17
Day id S. Brose
Skeletal Remains from Squaw Rockshelter
Fred Prior
55
Knotted Cordage from Squaw Rockshelter (33CU34),
Aurora Run, Cuyahoga County 59
R. L. Andrew’s and J. M. Adovasio
KIRTLANDIA
Editor
Dr. David S. Brose
Assistant Editor
Joseph T. Hannibal
The Cleveland Museum of Natural History
Associate Editors
James K. Bissell, Curator of Botany
The Cleveland Museum of Natural History
Bruce Latimer, Curator of
Physical Anthropology
The Cleveland Museum of Natural History
Martin Rosenberg, Editorial Assistant
Case Western Reserve University
Sonja Teraguchi, Manager of Collections
The Cleveland Museum of Natural History
Editorial Advisory Board
Rodney Feldmann, Professor of Geology
Kent State University
Michael C. Hansen, Geologist
Ohio Geological Survey
Richard Meindl, Associate Professor
of Anthropology
Kent State University
G. Michael Pratt, Associate Professor
of Anthropology
Heidelberg University
David H. Stansbery, Director, Museum
of Zoology
Ohio State University
Frederick H. Utech, Curator of Botany
Carnegie Museum of Natural History
Ed Voss, Curator of the Herbarium
University of Michigan
Andrew M. White, Professor of Biology
John Carroll University
KIRTLANDIA
The Cleveland Museum of Natural History
March 1989 Number 44:3-16
A Method for Making
Three-dimensional Reproductions
of Bones and Fossils
Jenny A. Smith and Bruce M. Latimer
Department of Physical Anthropology
The Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Abstract
The production and dissemination of high quality casts is an integral part of the
paleontological and anatomical sciences. However, detailed descriptions of the techniques for
making high resolution, three dimensional reproductions are not readily available.
The casting laboratory at the Cleveland Museum of Natural History has developed a
technique for making detailed, dimensionally stable reproductions of fossil and skeletal
materials. Experimentation with molding compounds has shown Dow Corning Silastic, RTV
E, to be an excellent material for our purpose, in spite of its extreme sensitivity to certain
organotin compounds. Proper preparation of the original specimens prior to molding is
therefore crucial in the molding procedure. The original specimen is placed in a clay base and
is covered with a number of thin coats of deaired silastic. A plaster jacket is formed over the
flexible rubber layer to prevent angular distortion. The half mold is inverted, the clay
removed, a thin coat of parting compound painted over the exposed silastic and the process is
repeated. Casts are made in urethane plastic or dental stone. Hollow casts are made by
partially filling and rotating the mold. These easily learned techniques allow the production of
three dimensional casts of exceptional quality.
Introduction
The casting facilities of the Department of Physical
Anthropology at the Cleveland Museum of Natural History
were developed in 1975. At that time, the primary function
of the casting laboratory was to duplicate the three-to-four-
million-year-old hominid fossils that Donald Johanson had
recently discovered in the Hadar region of Ethiopia. The
techniques developed to mold and cast these fossils are
now being used to replicate select specimens from the
Hamann-Todd Osteological Collection, a large assemblage
of recent human and nonhuman primate skeletal materials
which is curated by the Museum’s Department of Physical
Anthropology. We believe that these procedures produce
accurate, stable reproductions of unequaled quality in both
4
Smith and Latimer
No. 44
plaster and plastic. They are, moreover, easily learned and
are readily adaptable to many situations, both field and
laboratory.
This paper specifically describes the molding and
casting techniques developed and presently in use in the
Laboratory of Physical Anthropology of the Cleveland
Museum of Natural History (Smith and Latimer, 1985;
1986). It is intended for those with some experience in
molding and casting and is not meant to be a review of, or
introduction to casting techniques. Several articles,
however, have detailed some of the many other molding
and casting techniques that are currently in use (Burke,
1983; Heaton, 1980; Jensen, 1961; Madsen, 1974;
O’Donnell and Hanley, 1983; Parsons, 1973; Quinn, 1940;
Reser, 1981; Rigby and Clark, 1965; Schrimper, 1973;
Siveter, 1982; Waters and Savage, 1971). Also, because
molding and casting is a "hands-on” activity we encourage
beginners to practice with less valuable objects.
As many of the materials used in casting procedures
are potentially hazardous (e.g. organic solvents, epoxies,
urethanes), instructions should be followed carefully.
Protective clothing including gloves (we recommend latex
disposable) should be worn. Adequate ventilation, such as a
fume hood, is a necessity during the mixing of the urethane
plastics, Dynacast and MasterCast, and the application of
trichloroethane, Glyptal, Krylon and silicone sprays. It is
also advisable to leave the newly unmolded plastic casts
under the hood for a 24-hour curing period.
Our presentation is divided into three sequential
sections. The first section describes specimen preparation
before molding, the selection of the molding material and
the seven basic steps of the molding process. The second
section describes casting techniques for the production of
plaster and plastic casts. The final section details finishing
methods, repair of defects and painting techniques. The
rotating casting machine used to make hollow casts is
described in Appendix 2.
Molding
Molding is the most time consuming part in the entire
procedure and should be done very carefully as mistakes are
very difficult to remedy later. The selection of an appropriate
molding material is crucial, as is the proper preparation of
the original specimen prior to molding. The following
description and illustrations will, for demonstrative
purposes, focus on a simple two-part mold. However, the
same techniques can be applied with slight modification to
create a more complex, multiple-piece mold.
Specimen Preparation
The primary objective in mold making is to
reproduce the greatest amount of detail without placing the
original specimen at risk. In order to maximize mold detail
and simultaneously protect the original specimen,
preparation techniques must be followed carefully.
A good mold will reproduce microscopic detail. It is
therefore very important that the surface of the original
specimen be cleaned so that it is free of any rubber from
previous moldings, unnecessary adhesives, clay, pieces of
matrix, plaster, etc.
At this point, because the physical properties of the
molding rubber could be affected by prior specimen
preparation, it is necessary to discuss the choice of molding
material. Our experience with various molding compounds
has shown that Dow Corning Silastic RTV E (Appendix 1)
molding rubber is superior in its ability to retain intricate
detail over the duration of the mold life. A major
drawback, however, is its extreme sensitivity to certain
compounds. Contamination by sulfonates, amines and
neoprene can severely inhibit this rubber’s vulcanization. It
is therefore necessary to review the original specimens
“history” as to how and when was it originally prepared
and what types of materials (especially adhesives) were
used. It is especially important to find out if the specimen
had been molded previously. This latter issue is pertinent
because many molding compounds are incompatible with
one another and the residue from a previous molding may
result in inhibition of the silastic rubber. In light of this, we
strongly encourage that detailed and easily accessible
records are kept regarding all preparation, molding and
casting procedures.
Finally, some fossil deposits naturally contain
elements that will inhibit molding rubber. It has been our
experience that fossils taken from a coal or shale matrix are
likely to result in some degree of molding rubber inhibition
due to the high sulfur content in these materials. In any
case the important point is to test all specimens for
chemical contamination prior to molding. This axiom holds
for modern bone specimens as well as fossils; all originals
must be checked for chemical contamination.
Molding rubber inhibition has occurred if the rubber
is gummy or uncured after the 24-hour curing period has
elapsed. Compatibility should be tested by placing a drop
of the mixed, uncured rubber on the surface to be
reproduced. An individual drop of the silastic should be
placed on each area exhibiting different surface structure or
coloration. Again, we strongly recommend testing any
object that is suspected as a potential problem; uncured,
semi-liquid silastic is very difficult to remove.
In the event of contamination, other silastics, for
example Silicone H (see Dow Coming Product Guide for a
list of Silastics), which are not as easily inhibited, may be
used in place of Silastic E. Although this publication
recommends the use of Silastic RTV E, the procedures can
be easily modified for use with other Silastic rubbers and
molding materials.
Chemically contaminated areas can often be cleaned
with a weak organic solvent such as acetone and, if
necessary, treated with a barrier coat, such as Glyptal
lacquer cement (Appendix 1 ) dissolved in acetone. After
1989
Bone and Fossil Reproductions
5
Fig lire 1. The frontal and inferior view of an adult male gibbon cranium that has been prepared for molding.
Note that the nasal aperture and all foramina and canals are blocked with clay.
coating the specimen or specific areas of the specimen with
thinned Glyptal, the surface should be washed with acetone
to reduce surface glaze. It is important to wait several hours
to allow any remaining solvent to volatilize, a step which
can be hastened by placing the specimen under a heat lamp.
Because acetone will dissolve most commonly used
Figure 2. Coronal section through skull showing reference
plane used in Figures 5,7,9,11 ,13,15 and 16.
adhesives, care must also be taken to ensure that any prior
reconstructions or repairs are not damaged. Again, it is
necessary to check that the adhesives are completely dry as
any partially dried glue or softened adhesive will result in
contamination of the molding rubber. Water when present
on the original specimens can also result in inhibition.
Having once tested the specimen to ensure its
compatibility with the molding rubber, it is then necessary
to “stabilize” fragile areas so they will not be damaged in
the subsequent procedures.
The glyptal acetone mixture described above can also
be used to stabilize or strengthen fragile surfaces that might
otherwise be damaged during the molding process.
Painting especially delicate surfaces with petroleum jelly
dissolved in 1-1-1 Trichloroethane (CH3CCI3) will prevent
adhesion of the molding rubber and in so doing will
facilitate removal of the original from the mold. This latter
step reduces mold detail, but is often necessary with
extremely fragile specimens. If the petroleum jelly solution
is used, it must be carefully removed from the original
specimen after the molding procedure is completed. After
checking for molding rubber compatibility and stabilizing
any especially fragile areas, the next step is to block small
openings to prevent the penetration of the liquid molding
rubber. This includes openings such as exposed medullary
canals, foramina, areas between teeth or deep cracks in the
surface of the specimens. These should be blocked to
within a few millimeters of the external surface with clay
or beeswax. We use Plastalina clay, a wax-type plastic clay.
6
Smith and Latimer
No. 44
Figure 3. Step 1 — Lateral view of the skull partially embedded in clay block (shown as transparent) to demonstrate position
of flash line. Note that the flash line follows natural contour lines and ridges, such as the edge of the tooth row,
the zygomatics and the nuchal ridge along the back of the skull.
but any clay which is compatible with the silastic rubber
can be used. Again, it is important to check that the clay is
compatible with the molding rubber. Many kinds of oil-
base modeling clay inhibit RTV curing.
We have found when blocking small foramina,
beeswax due to its tackiness, works better than clay. In
addition, because it is sticky and semi-transparent beeswax
can also be used to create a thin “window” of material to
block small openings. The resulting membrane of casting
material formed in the completed cast can be easily
removed with a needle or sharpened dental tool. This
technique is especially useful when molding jaws and teeth
Figure 4. Step 1 — Skull partially embedded in clay block to
establish part line Note elevated tabs which act as registration
guides. Note also that all elevations and depressions on the
face of the mold are angular with planar surfaces.
where spaces between individual teeth must be maintained.
Again, if there is any question regarding compatibility, any
clays and/or waxes should be tested prior to use.
Deep cracks or fissures in the original makes
“demolding” of the original as well as subsequent casts
difficult. If this is judged to be the case, cracks and fissures
should be blocked or filled to within a millimeter or two of
the surface with wax or clay. Reconstructions of this type
can strengthen an especially fragile original, but these must
be sharply delimited so that there is no confusion on the
finished cast as to the reconstructed and original areas.
We believe that permanent fossil reconstructions
should, when possible, be done on highly detailed plaster
casts and not on actual specimens. However, if for some
reason it is necessary to reconstruct an original specimen,
the material chosen must be dimensionally stable as well as
reversible and easily removed. Epoxy putty is a material
which, when soft, can be configured to a particular shape
and then allowed to dry away from the original specimen.
Once hardened, the putty can be sanded and shaped and
then glued in position with a reversible adhesive. Plaster or
clay should not be utilized for fossil reconstructions as these
materials are very difficult to remove from the surface of
the bone. In addition, clay-like materials shrink as they age
which can result in serious damage to the original.
Making the Mold
The molding procedure is discussed and illustrated in
seven sequential steps to make it easier to follow the
instructions. Please note that much of the important detail is
contained in the figure captions. The specimen used in this
example was a male gibbon ( Hylobates ) skull (Figure 1).
Although the methods discussed here are for a two piece
1989
Bone and Fossil Reproductions
7
mold, the same basic procedures are followed for a
multiple-piece mold. However, because of the extra part-
lines and registration problems in complex molds, they
should always be made as simple as possible. Specimens
which are substantially larger or smaller than the gibbon
skull will require some modification of the procedures.
Step 1
The gibbon skull is partially embedded in clay to
establish the mold “part” line. The positioning of the “part”
or “flash” line around the specimen is critical and
important anatomical features must not be obscured.
Furthermore, deep undercuts and especially fragile areas
must be taken into account when designing the mold to
ensure that the original and plaster casts can be removed
from the completed mold without damage. In the sample
specimen the flash line is positioned to follow the natural
contours of the skull and to produce a two-part mold, one-
part approximately two-thirds of the total volume of the
skull (Figures 3 and 4). This is an important factor when
pouring hollow casts, when the volume of casting material
used is approximately two-thirds of the total volume of the
specimen (see Casting in Plastic, this article). Molds that
are primarily for plaster casts are more easily handled
when the two parts are of equal size (see Casting in
Plaster , this article) but we have found that problems with
registration occur if one part of the mold is less than one-
third of the volume of the specimen. Thus the flash-line in
the Hylobates skull was designed to follow natural lines
and ridges on the skull, along the edge of the tooth row and
zygomatics and the nuchal ridge.
The clay is formed into a rectangular block which is
deep enough to accept approximately half of the specimen.
We find that the easiest way to do this is to roll the clay
into approximately lcm-thick slabs. We have tried melting
the clay and pouring it into shallow pans to make the clay
slabs but this procedure proved less than satisfactory for
numerous reasons and we now simply roll the clay
mechanically. Rolling can be accomplished with a baker’s
rolling pin, a process made easier by first working the clay
under a heat lamp. In addition, a mechanical wringer such
as those used on old fashioned washing machines also
works to roll the clay slabs. If possible, it is important to
keep the clay block in a square or rectangular shape with
vertical walls as this will assist mold registration in later
stages. Also, although the block must be of adequate size to
accept the original, it should not be much more than 3cm to
5cm larger on any side. This facilitates handling of the
completed mold (small molds are easier to manipulate) and
also allows the use of a dental irrigation syringe to inject
plaster into the closed mold (see below).
After making sure that the clay is deep enough, it is
placed on a base made of particle board, plywood or
plexiglass. We recommend spraying the base with silicone,
or putting wax paper or a paper towel between the clay and
the base, to prevent the clay from sticking. The outline of
the specimen is then inscribed on the clay surface and a
hole is cut large enough for the specimen to be inserted
with the predetermined part-line in the same plane as the
clay surface. Small amounts of clay are added as needed
and pushed against the specimen along the part-line,
leaving a clean, sharp contact continuous with the plane of
the clay. A small double bladed, stainless steel spatula with
the tips of both the rounded and tapered ends bent to a 45=°
angle is an excellent tool for this purpose.
In cases where the natural topography of the
specimen requires a depression or elevation of the clay, the
edges should be sharply delineated and perpendicular. As
much of the clay surface as practical should be smooth and
horizontal in order to minimize potential shear strains and
Smith and Latimer
No. 44
Figure 6. Step 2 — First thin coat of silastic molding rubber
spread over the original specimen. Care must be taken to
ensure that air bubbles are not trapped between this primary
coat and the surface of the original specimen.
consequent deformation between the two halves of the mold
when it is closed. During the formation of the part-line any
unneeded clay should be removed to ensure a precise,
continuous junction between the clay and the specimen.
A series of elevated tabs, and/or regular depressions,
are placed at regular intervals around the specimen to act as
registration guides. These tabs should be angular with flat,
beveled edges. Lastly, the edges of the clay slab should be
cut to make a square or rectangle with sharp vertical sides.
This will also assist in proper registration of the two halves
of the closed mold.
Step 2
This is the most important step during the molding
procedure. Great care must be taken to ensure that no air
bubbles are trapped at the boundary of the object’s surface
and the first coat of molding rubber. The subsequent layers
of the rubber serve only to strengthen the primary coat so
that the finished mold can only provide as much detail as
the first layer of molding rubber. The entire surface of the
original specimen must be coated in a thin layer of molding
rubber with no air bubbles at the rubber-specimen interface.
The silastic rubber and curing agent should be
thoroughly mixed immediately before using in the ratio
recommended by the manufacturer. After mixing, the
rubber should be deaerated to reduce air bubbles that were
introduced during mixing. The mixed molding rubber is
placed under vacuum (approximately 25mm pressure) until
it completely expands and recedes to its original level.
Experience will dictate the amount which can be vacuumed
at one time. Caution should be exercised to guard against
the expanding rubber overflowing its receptacle and being
drawn into the hose connecting the vacuum pump and bell
jar. We find that large disposable plastic or paper cups are
satisfactory for vacuuming the rubber. Avoid Styrofoam
cups for this purpose. After the molding rubber has been
vacuumed a disposable wooden spatula works well as a tool
for spreading the liquid silastic over the original. However,
touching the surface of the specimen should be avoided and
a stream of compressed air should be used to direct the
liquid molding rubber over the surface and especially into
small cracks, crevices or undercuts. In the case of small
specimens or difficult to reach areas a small, clean paint
brush combined with compressed air can be used.
Often under field conditions neither a vacuum pump
nor an air compressor is available. In this situation, less
vigorous stirring and mixing of the molding rubber is
advisable to reduce the introduction of air into the mixture.
The air which is inevitably introduced into the molding
CLAY
REGISTRATION
TAB
S I
LASTIC
Figure 7. Step 2 — Coronal section.
1989
Bone and Fossil Reproductions
9
Figure 8. Step 3 — Second coat of silastic molding rubber
applied and clay retaining walls added. Areas on photograph
have been shaded to highlight detail.
rubber during the mixing process will coalesce into bubbles
which gradually rise to the surface of the liquid rubber. It is
therefore necessary to check the liquid rubber which has
been placed over the specimen approximately every 20
minutes (for the first hour) in order to remove these
bubbles. The number of bubbles can also be reduced by
mixing the silastic in a paper cup and allowing the mixture
to stand for about ten minutes; a period which allows the
air bubbles to rise to the top. A small hole can then be cut
in the bottom of the cup and the silastic poured over the
specimen through this hole.
An additional aid is to use a stream of compressed air
which can be created by attaching a hose to the valve of a
spare tire which has been previously overinflated. Any
bubbles rising to the surface of the silastic can be removed
by directing the air across them. Although this may appear
to be additional work, special care must be taken to avoid
entrapment of bubbles at the part-line or the rubber-
specimen interface.
Step 3
A second thin coat of silastic is applied after the
initial coat has vulcanized (approximately 24 hours) and
retaining walls of clay are built around the mold. We do not
recommend placing retaining walls around the mold any
earlier than this stage. The reason is to permit easy access
to all areas of the specimen during the first, crucial coat.
Again, the mold and the retaining walls should be angular
to assist registration.
Step 4
Additional coats of RTV E are applied until
approximately 3mm of the material evenly covers the
specimen and surrounding clay base. It is important to wait
the recommended amount of time between coats as
improper curing might otherwise ensue. Importantly,
thicker and more layers of silastic are not necessarily better
because the thicker the silastic, the less flexible the mold
and the more difficult the unmolding process. To avoid
creating a rigid, solid block of silastic completely covering
the specimen, ground cotton or cabosil can be added to
decrease the fluidity of the silastic so that it can be applied
to strongly contoured surfaces and/or deep indentations. In
addition, surgical gauze can be placed over the external
surface of the mold to serve as a strengthening agent.
-CLAY WALL
SILASTIC
Figure 9. Step 3 — Coronal section.
10
Smith and Latimer
No. 44
Figure 10. Step 4 — Final coat of silastic molding rubber
applied to this half of the mold. Note that gauze has been
applied to the specimen at this stage to increase strength of the
rubber. Note also that the gauze and rubber follow the natural
contours of the original specimen.
Figure 12. Step 5 — Completed half of mold. Plaster jacket
partially cut away to demonstrate the shape and position of
plaster inserts. Note also the addition of silicone
rubber registration tabs.
A set of small, precut registration tabs of hardened
rubber are affixed to the surface of the final coat of uncured
silastic to firmly secure the mold to the plaster jacket.
These registration tabs should have flat, beveled edges to
ensure easy removal of the plaster jacket which is
described in the following step.
Step 5
A plaster jacket (“mother mold") minimizes
distortion in the flexible silastic rubber mold. In the
absence of a plaster jacket, any pressure on the highly
flexible rubber mold can potentially result in a distortion of
the original contours and subsequent error in the cast. The
external surface of the rubber mold must be smooth to
prevent adherence of the plaster jacket and any roughened
area should be covered with a coat of vaseline. In addition,
a liberal coating of talcum powder over the silastic will
also help the removal of the plaster jacket.
Before pouring the plaster for the external jacket, all
undercuts and insets (e.g. orbits) are filled with individual
GAUZE
SILASTIC
TAB
CLAY
WALL
Figure 11. Step 4 — Coronal section.
1989
Bone and Fossil Reproductions
11
plaster plugs (Figures 12 and 13). These plugs are liberally
covered with vaseline to prevent their bonding to the jacket.
Gypsum Hydrocal A- 11 plaster (Appendix 1) is a
good, jacketing material. A single layer of fast setting
plaster bandage (Appendix 1) applied initially, will
strengthen the jacket. The addition of sissal fiber to the
liquid plaster can also be used to strengthen and contour
the plaster jacket. Fiberglass also makes a rigid yet
lightweight material for jacketing.
Step 6
The clay walls and base are removed and the half
finished mold is inverted. Any clay adhering to the
specimen and the surrounding silastic is carefully removed.
Check especially for traces near the surface of the part line
by gently pulling the silastic away from the specimen in the
area of the part line. Having done this, it is necessary to
make sure that the silastic edge returns exactly to the
established part line and does not pucker against the edge
of the specimen.
A thin coat of release agent (petroleum jelly
dissolved in trichloroethane, CH3CCL3) is then painted
over the exposed silastic, care being taken not to wet the
original specimen. The petroleum jelly trichloroethane
mixture will create the part line by ensuring that the
silastic layers will not bond to each other. Care should be
taken to coat the entire exposed surface of the silastic or
else adherence will occur between the two sides.
Figure 14. Step 6 — The mold has been inverted and clay
removed. The release agent has been applied to
the silastic rubber surface and the specimen
is now ready for the next step.
Step 7
Steps 2-5 are repeated for the exposed half of the
skull. During the casting procedure a mold is normally
taped closed, but we have found that it is advisable to
incorporate a threaded bolt system into the design of the
larger, heavier molds. During the first step of the molding
SKULL
CLAY
BASE
CLAY
WALL
PLASTER
/ PLUS
Figure 13. Step 5 — Coronal section. Plaster jacket shown in its entirety.
12
Smith and Latimer
No. 44
procedure, four unthreaded metal rods (equal in diameter to
the bolt) should be placed vertically in the clay base, one in
each corner. The rods should stay in position throughout
Steps 1 - 7 to ensure that the holes for the bolts through the
rubber and jacket are properly aligned. The plaster jacket
can be reinforced with a piece of hollow, threaded steel
lamp pipe placed over the rods during Step 5. Four
threaded bolts closed with wing nuts will give excellent
registration for larger molds.
To unmold the original specimen, the external jacket
is first removed. The two rubber halves are then carefully
separated as far as the specimen, care being taken not to
twist or damage the original. The two halves of the mold
are then cautiously peeled away from the specimen taking
special care in the areas of foramina and undercuts. Upon
removal of the original specimen, the molding process is
nearly complete. The final step is cleaning and restoring
the original specimen.
Wear and tear of the mold surface occurs from the
moment the original specimen is removed. Occasionally,
small pieces of the molding rubber caught in cracks and
crevices may tear away during the unmolding process,
emphasizing the importance of good preparation before
molding. Torn pieces of the mold can be reattached by
gluing with Silicon RTV 700 (Appendix 1 ). These repairs are
temporary and have to be repeated each time a cast is poured.
The first cast is the most accurate and can be saved
as a "record”. Then, if and when a second mold is required,
the record cast can be used in place of the original. In
fragile and rare fossil specimens it is especially critical to
use a “record cast” for remolding in order to keep the
original as pristine as possible. We use hardened dental
stone (see below) for our record casts because of its
dimensional stability. This material, although very hard, is
brittle and easily broken. Record casts, therefore, should
always be placed in a protected area and not handled. Also,
because of the natural fragility of the dental stone casts we
usually pour two record casts. A final note concerning
record casts is our recommendation not to clean the flash
lines off the cast. By leaving the flash, if and when it
becomes necessary to remold the specimen, the technician
can easily recognize and duplicate the original part-line.
Casting
Accurate casts can be poured in a variety of
materials, including plastics, epoxy compounds and
plasters. These materials vary in terms of physical
properties, working times and cast detail. Therefore, the
casting materials should be chosen on the basis of the
number of casts needed and the ultimate use of casts.
Because these various materials differentially effect mold
life, thought should also be given as to the total number of
“pulls” needed. We recommend the use of the following
materials because they produce dimensionally stable,
highly detailed casts while minimizing mold damage.
Casting in Plaster
Coecal (Appendix 1) is a dimensionally stable, chip
resistant dental stone that gives excellent detail. We have
found that mold life using Coecal is excellent; 25-35
“pulls” as opposed to 15-20 “pulls” for plastic casts. The
trade off is, of course, the natural fragility of the dental
stone relative to the more durable plastics. Prior to the
introduction of the liquid dental stone into the mold, a
mold dressing (Appendix 1) is painted on the mold's
internal surface to reduce surface tension and prevent the
1989
Bone and Fossil. Reproductions
13
Figure 16. Coronal section of completed mold. See text for details.
entrapment of air bubbles. Coecal dental stone is mixed
with liquid Gypsum Hardener (Appendix 1) at a ratio of
100:34 by weight, in a Whip-Mix machine under vacuum.
Gypsum Hardener is used in place of water because it
reduces the pore spaces and increases the compressive
strength of the dental stone. However, one cautionary note
must be added regarding the use of Gypsum Hardener
instead of water. The liquid hardener significantly hastens
the “setting time” and as a consequence reduces the
handling time.
Prior to actually filling the mold halves, small
indentations and concavities should first be injected using a
disposable dental irrigation syringe (Appendix 1). The two
parts of the mold are then filled with liquid plaster over a
vibrating plate to dislodge trapped air bubbles adhering to
the surface of the mold. Just before closing, a little extra
plaster should be added to both halves of the mold, which
are then held side-by-side one in each palm and “slapped”
shut. When first practicing this procedure it is advisable to
wear a lab coat.
The mold should immediately be taped (using fiber
strapping tape; Appendix 1 ) or bolted closed to prevent the
loss of the liquid plaster. Importantly, if using a threaded
bolt system, the wing nuts should never be secured beyond
“finger tightness”. Any further pressure will distort the
mold and lead to shearing along the plane of the part-line.
In addition, in especially large molds the small amount of
expansion that occurs when the plaster hardens can actually
crack the mold jacket if it is tightened too securely.
To prevent a bubble from coalescing on the surface
of the cast, the mold can be rotated by hand for two to three
minutes while the plaster hardens. If deemed necessary,
extra plaster can be added after the mold is taped shut by
gently inserting the nozzle of a syringe between the mold
halves and injecting liquid plaster. This latter step is often
required if large amounts of liquid plaster are spilled during
the closing of the mold halves.
Dental stone casts can be solid or hollow. For hollow
casts the ratio of dental stone to Gypsum Hardener is
increased to 100:36 by weight to increase the handling time
and the mold is rotated in the casting machine (Appendix 2).
Casting in Plastic
Frequent use of epoxies should be avoided as epoxies
chemically contaminate silicone rubber reducing mold life.
Manufacturer specifications recommend the use of
polyesters and polyurethanes as casting materials with
RTV E. We have found that a 30:50 mixture (by weight) of
two polyurethanes (Dynacast and Master Cast 750;
Appendix 1) gives excellent detail and minimizes mold
damage (after 15 “hollow” castings most molds show no
appreciable damage). However, Master Cast and other
polyurethanes are extremely sensitive to moisture and
should be mixed in plastic containers with metal or plastic
utensils. A thin coat of silicone sprayed over the mold
surface will facilitate unmolding and can also help protect
the mold rubber from contamination.
Dynacast and Master Cast are 2-component
thermosetting urethane plastics. Precautions therefore must
be taken to prevent the formation of the small air bubbles in
14
Smith and Latimer
No. 44
the finished cast, which are the by-product of the heat
produced by the chemical reaction in the plastics.
“Shelling” the mold surface with a thin coat of liquid plastic
using a disposable brush will reduce volume and heat and
ensures an even flow of plastic over the entire mold surface.
In difficult to reach areas, (e.g. teeth, styloid processes,
pterygoid plates) plastic should be injected using a plastic
disposable dental irrigation syringe. After approximately an
hour, when the first coat of plastic has hardened, the mold
should be filled 2/3 full of the liquid plastic, closed, and
taped or bolted shut. Again, as with the plaster, extra plastic
can be injected into the closed mold using a disposable
syringe. The silastic is flexible enough to insert syringe tip
gently between the two halves of the mold.
Because this mixture of urethanes thickens fairly
rapidly (has a working life of less than ten minutes)
handling time is an important factor. The filled mold
should be wrapped to prevent spillage and fastened onto
the rotating machine (Appendix 2) while the plastic casting
material is still liquid. Although no longer liquid after 20
minutes the plastic remains malleable and the mold should
be rotated for approximately an hour.
Hollow casts can be poured in Dynacast alone.
Dynacast, however, has a pot life of only five minutes (at
25°°C) and it takes an experienced caster to work under
these conditions. Most casts can be unmolded four hours
after closing the mold. However, the urethanes often
remain somewhat malleable in small casts and should
harden for at least six hours before unmolding. We
generally leave urethane casts in a well ventilated area
(under a fume hood) for 24 hours after unmolding for the
completion of the curing process.
Finishing Techniques
When the casts are fully hardened, the flash along the
“part” line is trimmed with a scalpel. A well-registered
mold produces casts that have an all but invisible part line.
This is especially important when painting, as the part-line
takes pigment differently from the surrounding casting
material. Small defects or bubbles are filled with plaster,
urethane or epoxy putty. Positive bubbles caused by defects
in the mold surface can be flicked off with a needle point.
Painting can duplicate either the original color(s) or, in a
study cast, enhance surface detail. Record casts should
always remain unpainted as pigment will obscure surface
detail during the remolding process.
Painting Plaster Casts
Dry artist’s pigment dusted over the cast’s external
surface and sealed with an acrylic agent (Appendix 1)
highlights detail by filling cracks and fissures. For study
casts, a wash of Grumbacher raw umber pigment mixed in
acetone will intensify surface detail. Lightly rinsing with
water will ensure an even distribution of color over the
surface of the cast. A mixture of dry pigments, acrylic
paints and magic markers can also be used to duplicate the
original colors.
Painting Plastic Casts
Plastic casts are more difficult to color, owing to the
material’s natural impermeability and repellency to liquids.
However, liquid and/or dry pigments can be applied to the
mold’s internal surface prior to adding the plastic. Upon
hardening, the plastic will bond to the pigments, resulting
in a permanent finish. This method, although time
consuming, gives excellent results. Plastic casts can also be
colored by painting the external surface after unmolding. A
satisfactory paint is obtained by dissolving acrylic
pigments in a liquid matte medium to which an Acrylic
Flow Improver (Appendix 1) has been added (2 parts to
20). An acrylic flow improver slows the setting time which
is important if an even color is required. Reconstructed
areas can be delineated with acrylic or enamel paints.
Special urethane pigments are available and can be added
while mixing. Difficulties arise, however, in obtaining
identical shading if casts or portions of the cast are poured
at different times.
Conclusion
These molding and casting techniques are the result
of experimentation over the past 12 years in the Laboratory
of Physical Anthropology. The methods described are
easily learned and enable the production of economical,
dimensionally stable, high resolution casts, that are suitable
for display, research and educational purposes. A
Appendix 1
Materials
The materials listed below are those currently being
used in the casting laboratory at the Cleveland Museum of
Natural History. This list includes materials that we have
chosen through extensive trial and error.
Molding Materials and Equipment
• Silastic RTV E molding rubber: Dow Corning,
Midland, Michigan.
• Silicon RTV 700: General Electric Silicone
Products, Waterford, New York.
• Glyptal Lacquer Cement 1726: General Electric,
Schenectady, New York.
• Plastalina Leisure Clay #10422: Leisure Craft
Company, Compton, California.
• CH3CCI3 Trichloroethane: Fisher Scientific
Company, Springfield, New Jersey.
• Kling rolled surgical gauze: Johnson and Johnson,
medical supply store.
• Fast setting plaster bandage: Johnson and Johnson,
medical supply store.
1989
Bone and Fossil Reproductions
15
5" AUXILLARY DRIVE PULLEY 28 RPM
Figure 17. Schematic diagram of the rotating casting machine used in the casting laboratory at
The Cleveland Museum of Natural History. See Appendix 2 for details.
5" AUXILLARY DRIVE PULLEY
14" STATIONARY PULLEY
6" MAIN DRIVE PULLEY
Figure 18. Diagram of the rotating casting machine, shown in previous figure. See Appendix 2 for details.
16
Smith and Latimer
No. 44
• Hydrocal Plaster A- 1 1 : United States Gypsum
Company, Chicago, Illinois.
Casting Materials and Equipment
• Dental Stone (Coecal plaster): Coe Laboratories
Inc., Chicago, Illinois.
• Plaster Gypsum Hardener: Whip Mix Corporation,
Louisville, Kentucky.
• Permaflex Concentrate Mold Dressing: Permaflex
Mold Company, Columbus, Ohio.
• Whip Mix Casting Vacuum Mixer, Model B: Whip
Mix Corporation, Louisville, Kentucky.
• Dental Irrigation Syringe, Type #8881-411012
Curved Tip: Monoject; Sherwood Medical, St.
Louis, Missouri.
• Dynacast: Kindt Collins Co., Cleveland, Ohio.
• Master Cast 750: Kindt Collins Co., Cleveland,
Ohio.
• Pure Silicon Spray, Cat. #3302.7: Brookstone
Company, Peterborough, New Hampshire.
• Tape, Scotch Brand Tape No. 898.
• Dry Pigment: Grumbacher Artist Quality Dry
Color, art supply store.
• Acrylic Agent #1303 Crystal Clear Acrylic Spray
Coating: Krylon, art supply store.
• Liquid Acrylic Matte Medium: Liquitex, art supply
store.
• Acrylic Flow Improver: Windsor and Newton, art
supply store.
Appendix 2
Casting Machine
The rotating casting machine permits the production
of "hollow” plastic and plaster casts by rotating the molds
as the liquid casting material hardens. The machine rotates
the mold slowly in three dimensions, allowing the casting
material to flow equally on all internal mold surfaces. It is
important that the machine not develop sufficient
centrifugal force to overcome gravity and that the auxiliary
and main shafts not turn at identical RPM values. In our
machine we use different sized pulleys producing a speed
of 10 RPM for the main shaft and 28 RPM for the auxiliary
shaft. The platform upon which the mold is secured is
recessed and adjustable so that the center of mass can be
concentrated near the rotational axes.
Acknowledgements
We gratefully acknowledge Donald Johanson, Tim White
and William Kimbel for providing the impetus for the
casting program; Bill McIntosh for most of the original
ideas; and Rob Black and Cindy Luchetti for continual
improvements. Special considerations are due to John
Aicher who designed and built the rotating casting machine
and Bruce Frumker and Toni Hutton for their photographic
assistance. We would like to thank Luba Gudz and Kip
Dieringer for their excellent illustrations and Anthony
Mitri for his diagrammatic representation of the casting
machine. We would also like to thank Laurie Linden for
her patience for typing various versions of this manuscript.
Lastly, we would like to sincerely thank the numerous
volunteers whose dedication, hard work and creativity have
contributed greatly to the success of the casting program.
Development of these casting techniques were in part,
accomplished through support provided by the Institute of
Museum Services special project supporting grant (IP-
4001 1-84).
References
Burke, Ann C., Michael Anderson, Alison Weld, and Eugene S.
Gaffney. 1983. The Reconstruction and Casting of a Large
Extinct Turtle, Meiolania. Curator 26:5-26.
Heaton, Malcolm J. 1980. New Advances in Latex Casting
Techniques. Curator 23:95-100.
Jensen, James A. 1961. A New Casting Medium for Use in
Flexible and Rigid Molds. Curator 4:76-89.
Madsen, James H. 1974. Derakane Vinyl Ester Resins — An
Alternative to Plaster-of-Paris. Curator 17(l):64-75.
O'Donnell, Robert, and John H. Hanley. 1983. The “Gourmet”
Staining of Plaster Casts. Curator 26(4):307-3 13.
Parsons, Kenneth C. 1973. Precision Casting: A New Method in
Museum Technology. American Journal of Physical
Anthropology 38:789-802.
Quinn, James H. 1940. Rubber Molds and Plaster Casts in the
Paleontological Laboratory. Technique Series, Field Museum
of Natural History, 6.
Reser, Peter K. 1981. Precision Casting of Small Fossils: An
Update. Curator 24(3 ): 157-1 80.
Rigby. J. Keith, and David L. Clark. 1965. Casting and Molding,
pp. 389-413 in Handbook of Paleontological Techniques.
edited by Bernhard Kummel and David Raup. San Francisco
and London: Freeman.
Schrimper, George D. 1973. Hollow Casting of Fossil Skulls in
Epoxy Plastic. Curator l6(4):286-305.
Siveter, David J. 1982. Casts Illustrating Fine Ornament of a
Silurian Ostracod. The British Micropalaeontological Society.
Fossil and Recent Ostracods 6: 1 05- 1 22.
Smith, Jenny, and Bruce Latimer. 1985. The Joy of Casting. The
Explorer 27(4): 4-5, 22-24.
Smith, J. A. and B. Latimer. 1986. A Method for Making High
Quality Three-dimensional Reproductions of Vertebrate
Fossils. American Journal of Physical Anthropology
69(2):265-266.
Waters, Barbara T., and Donald E. Savage. 1971. Making
Duplicates of Small Vertebrate Fossils for Teaching and for
Research Collections. Curator 14(2): 1 23- 1 32.
KIRTLANDIA
The Cleveland Museum of Natural History
The Squaw Rockshelter (33CU34):
A Stratified Archaic Deposit
in Cuyahoga County
The Squaw Rockshelter site (33CU34), located along a tertiary tributary to Lake Erie on the edge of
the Glaciated Allegheny plateau in northeast Ohio, contained stratified sealed deposits. A small
remnant dated 9240±160 B.P., yielded fragments of a human skeleton, unifacial and bifacial scrapers,
unstemmed lanceolate points, and a corner-removed indented-base point. Additional lanceolate and
bifurcate-base points have eroded from this deposit. This lower cultural level, relatively continuous
across the rear of the shelter, represents a drip-line depression washed by water.
Within a thin zone overlying superimposed rockfall, a hearth dated to 5500±85 B.P. is associated
with a fragment of charred cordage. This level yielded expanded stemmed points, bifacial knives and
drills. The occupation surface could not be traced across disturbed areas of the shelter.
The early Archaic lithic assemblage included both Plano projectile point styles of the Great
Lakes and corner-notched types of the southern Appalachians, while the late Middle Archaic projectile
points represent a local blending of styles typical in the riverine midwest and the Mid-Atlantic states.
Skeletal remains recovered from the Early Archaic level of the Squaw Rockshelter were
identified as belonging to a young female Amerindian. Analyses of the limited fragments and dentition
suggest a balanced diet of moderate coarseness. No pathologies or trauma were noted. Two teeth from
other individuals were also present.
March 1989
Number 44:17-53
David S. Brose
Department of Archaeology
The Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Abstract
Introduction
Squaw Rockshelter (33CU34) is a small and partially
disturbed site on the southwest bank of Aurora Branch, a
second order tributary of the Chagrin River. The site is
located in the South Chagrin Reservation of the Cleveland
Metropolitan park system, Bentleyville Township,
Cuyahoga County, Ohio (Figure 1 ).
The discovery of the site occurred in 1 974. 1 Having
followed the riverbank 500m upstream from the 1884
carving “Squaw Rock," Robert M. Brose noticed a
lanceolate biface of Upper Mercer flint exposed in a silty
colluvial deposit veneering the riverbank. Within minutes
a bifurcate-base projectile point, also of Upper Mercer
18
Brose
No. 44
Figure 1. The location of northeast Ohio sites mentioned in text: 1 — Cooper Hollow: 2 — Burrill Hill Orchard; 3 — Ziegler;
4 — Regis; 5 — Hogue's Spring; 6 — House and Lukens; 7 — Squaw Rockshelter; 8 — McKiben; 9 — Holdson District
flint, was recovered from a higher gully in the same
sediments by Thomas A. Brose. By the end of that
afternoon a small lanceolate point of mottled glacial chert
and a second bifurcate-base point of Upper Mercer flint
had been collected. The condition of three of the four
artifacts suggested a primary depositional context.
Permission to investigate was obtained from Mr. Harold
Growth II, Cleveland Metroparks Director.
That autumn the colluvial fan was traced to a
presumed terrace behind several large sandstone blocks,
3.5m higher than the river. Excavation quickly revealed a
partially collapsed rockshelter, with deposits protected
from the river by massive blocks of sandstone detached
from the overhanging cliff 2m south. By the end of the first
day only a single unit, 1.0m x 1.25m, had been excavated
to a depth of 1.40m. The entire deposit was estimated to
include a zone less than 2.25m from front to back,
extending along Aurora Branch for 15m. This zone
Table 1
Archaeological Excavations at 33CU34
Excavation Unit
Surface Area
Maximum Depth
Actual Volume Excavated
1
1.5m x 1.5m
145cm
2.4m-1
2
1.25m x 1.0m
137cm
1 ,4m3
3
1 ,50m x 2.0m
128cm
2.8m1
4
1 ,50m x 1.75m
95cm
1 .8m3
5
2.0m x 1.75m
153cm
2.5m1
TX
1.50m x 0.50m
85cm
0.6m1
1989
Squaw Rockshelter
19
Figure 2a. View from old quarry on north bank, of large
roof fall blocks protecting the Squaw Rockshelter site.
contained large blocks of sandstone rising like icebergs
through the unconsolidated deposits (Figure 2a, b). The
largest area between such obstacles was only 2.5m x 1 ,5m.
The Cleveland Metroparks insisted that no excavation be
left open; that personnel be limited; and that there be no
publicity until work was completed. Due both to previous
commitments and current academic concerns, I agreed to
personally undertake the fieldwork and analyses, and to
curate recovered materials at The Cleveland Museum of
Natural History. This arrangement was accepted by Harold
Mahan, Director of the Museum in December 1974.
Through the 1975 summer 13m2 in four units and a
single exploratory trench were excavated into the
unconsolidated deposits at the Squaw Rockshelter site
(Table I, Figure 3). It is estimated that the 11.5m3
excavated represent 85% of the site preserved from
erosion. This is a small and somewhat biased sample of
what may have been 40m3 of sediment. Nevertheless, these
deposits offered evidence to support a different
interpretation of the Archaic of northeast Ohio than existed
prior to 1975.
Cultural and Historical Background
Early descriptions of Ohio antiquities concentrated
upon earthworks and the artifacts they contained (Brose
Figure 2b. View to west in Squaw Rockshelter, figures
on surface of Unit 3 before excavation.
1973). Stone tools that appeared more primitive, and badly
decayed human bone from river banks, wells, and mines
were compared to materials recovered from purported
Pleistocene deposits in other states. Most such reports were
ignored, since human occupation of the New World was
not considered to be of great antiquity (cf. Fowke 1902),
and those sites subjected to detailed scrutiny turned out to
be of dubious antiquity (Holmes 1919). The association of
artifacts with extinct fauna not only demonstrated
respectable antiquity for western Amerindians, but
revealed projectile points with analogues throughout North
America. Shetrone (1936) noted the frequency of such
points in Ohio, but lack of any stratigraphy limited
temporal interpretation. Subsequent studies (Tuck 1978)
have revealed eastern Paleolndian sites with a variety of
projectile point forms. Later Paleolndian lanceolate point
types documented from the plains were also recovered in
the Great Lakes-Ohio Valley, and geochronological
correlations suggested considerable age for these types in
the midwest (Mayer-Oakes 1955; Mason 1958).
There seemed good reason to regard the east as
having been occupied as early as the west (Mason 1962).
Highland rockshelters and deep stratified sites in
floodplains yielded fluted points as well as partially fluted.
20
Brose
No. 44
Figure 3. Excavations at Squaw Rockshelter site.
unfluted, and notched projectile points showing a transition
in lithic technology from the Paleolndian period into the
early Middle Archaic period (Coe 1964; DeJarnette et al.
1969; Griffin 1974; Chapman 1977). These sites have been
taken to indicate a wide-spread, and synchronous series of
changing projectile point styles with little or no overlap of
differing types at any component or at any given time over
nearly 1500 air miles from Maine to Mississippi (Broyles
197 1 ; Dincauze 1975; Bense et al. 1983). It was a situation
unparalled by the Early Archaic in the Midwest.
From the Dakotas into western Ohio, Paleolndian
points precede development of the prairies. With the
Hypsithermal, both prairies and gallery forest zones yield
assemblages of stemmed and unstemmed lanceolate and
partially fluted projectile points. By 9500 B.P., as the
prairie fingered eastward, the grasslands yield
assemblages in which lanceolate projectile points
predominate, while forested areas yield assemblages in
which corner-notched and side-notched projectile points
predominate. Short-duration components at ecotones from
9500 B.P. to 6500 B.P. lithic assemblages contain a variety
of stemmed and unstemmed lanceolate projectile points as
well as a variety of expanded stemmed and/or corner-
removed, and side-notched and corner-notched projectile
point types.
While the stemmed and unstemmed lanceolate
projectile points from the Great Lakes suggested
relationships to the Plano complex further west (cf. Mason
1981:114-126), they were also compared to Late Archaic
points, from 5000 B.P. to 3500 B.P. further west, or to
similar projectile points in the St. Lawrence or Atlantic
drainages of Pennsylvania and New York, dated to the end
of the Middle Archaic (cf. Dragoo 1959). Those who
argued a late date for the stemmed and lanceolate points in
the Great Lakes claimed that the accompanying corner-
removed or notched points related to Normanskill points.
1989
Squaw Rockshelter
21
Brewerton points. Otter Creek points or even Snook Kill
points, dated between 6500 B.P. and 3500 B.P. in the east.
Those who viewed the Great Lakes lanceolates as early
related the corner-removed or notched points to Thebes,
Hardin barbed, Graham Cave side-notched, Kirk corner-
notched, and St. Albans side-notched types, dated between
9500 B.P. and 6500 B.P. in river drainages to the south and
west (Brose 1975).
The absence of components that yielded only a single
clearly identified point type suggested to some that there
were no Early or Middle Archaic occupants of the region at
all. Assuming a Paleolndian focal adaptation to tundra-
edge hunting. Fitting (1968, 1970) argued for a post-
Paleolndian depopulation of the Great Lakes, suggesting
the 9000 B.P. rise from low levels created a lakeshore
unfavorable to human occupation. Those few possible large
Early and Middle Archaic sites, he believed, were now
drowned, having been located along waterways, as in the
southeast, and only smaller interior hunting camps still
existed. This ignored controversy about the reality of any
such pattern in the south. Mason however, argued that the
4500 years between Paleolndian and Archaic economy and
technology represent
...gradual transition from one dominant culture
type to another.. .in the Great Lakes, enough is known
to encourage the view that the perception of two
partly coetaneous cultural traditions is accurate.
(1981:114-115)
Indeed, Ellis and Deller ( 1986:56-57) speculated that
in southern Ontario and lower Michigan Early Archaic
Kirk and Plano types would co-occur by 8900 B.P, while
Middle Archaic side-notched forms should resemble
Godar, Brewerton-like Thebes, and Otter Creek by 4500
B.P. (cf. Lovis and Robertson 1985). Candidates for the
9500-6500 B.P. period in the Great Lakes, thus included
both lanceolate projectile points, and the corner-removed,
stemmed, bifurcated base, and notched projectile point
clusters dated between 10,000 B.P. and 6500 B.P. to the
west, and dated between 9500 B.P. and 3500 B.P. in areas
east of Ohio. Ohio remained an enigma.
In the upper Ohio Valley Mayer-Oakes (1955) had
assigned stemmed and lanceolate projectile points to an
early Archaic horizon, while Dragoo assigned a similar
lithic assemblage (which included corner-removed/side-
notched points) to a “proto-Laurentian culture” with at
least one radiocarbon date as late as 53 1 0± 1 80 B.P.
(Dragoo 1959:238-239). Nonetheless, the lowest levels of
the Rohr shelter, with Steubenville lanceolate points,
yielded a Dalton point while in the Allegheny River valley
lanceolate-free Brewerton components date to 5800 B.P.
(Caulkin and Miller 1977) and to 6090 B.P. (George and
Davis 1986). As in the upper Ohio Valley, the lanceolate
points of northern Ohio have been called Early Archaic,
Late Archaic, or both. Few components reported dated
between 9500 B.P. and 3500 B.P.
The first serious attempt to address the Paleolndian
and early Archaic occupation in northern Ohio, was
initiated by Olaf Prufer ( i 96 1 ). Based upon survey by
Prufer and his associates between 1958 and 1963, a study
of the distribution of various types across Ohio was
published (Prufer and Baby 1963). Given the nature of the
sampling, chronological typology was based upon distant
analogues. Not all of these have received radiometric
support, and the relationships between the survey sample,
sampling methods and sampling framework have been a
source of contention (Seeman and Prufer 1982; Lepper
1983). Beyond stimulating research, these studies
illustrated the variability and standardized the terminology
for early Ohio projectile points.
Although recognizing their differing temporal
positions, the Plano points in northwestern Ohio and the
fluted points of southern Ohio were attributed to
contemporary immigrations from the upper Great Lakes
and the Appalachians. After discussing specific sites Prufer
and Baby found
...no evidence permitting the linkage between
the Paleolndian assemblages of Ohio and established
local Archaic complexes.
(1963; emphasis added)
The problem was that in 1963 there were no
established local Archaic complexes other than mortuary
aspects of the Late Archaic. Although Mayer-Oakes
suggested continuity, Prufer and Baby agreed with Dragoo
that there were no Archaic predecessors in any Ohio late
Paleolndian site assemblage. They accepted Dragoo’s
chronology, and his implication that no lanceolate
projectile point in eastern Ohio was early.
Geistweit argued that Kirk and LeCroy points had
been recovered in all regions of Ohio save the northeast
(1970:1-32, figs. 9-16). She too saw no evidence for
Paleolndian Early Archaic continuity, and offered no
explanation for a proposed rapid introduction of Early
Archaic as a style horizon or as an economic pattern
(1970:162-164). Claiming dense Late Archaic occupation
of those same regions (1970:45-88) Geistweit suggested
that the Middle Archaic in Ohio was represented by a
“...continuation of Early Archaic tool types and way of
life...” (1970:44, 164) The occupational hiatus was
between 5000 B.C. and 3000 B.C. for Geistweit, who
found little local precedent for regional Late Archaic
variants, especially along the Ohio River valley.
Blank (1970) synthesized data from collections in
northern and west central Ohio with excavations in east
central Ohio. Despite the paleoecological data with which
he sought to support his model, every surface collection
was selective, and other than at quarries all sites showed a
22
Brose
No. 44
mixture of lanceolate, stemmed and notched point types.
No site yielded floral or faunal remains, and there was
confusion of single site activity with economic adaptation.
Placing the late Paleolndian and Early Archaic periods
between 8000 and 5000 B.C., Blank suggested that
lanceolate points characterized northwestern Ohio groups
hunting elk and moose in the till and lake plain swamps.
The notched points of southeastern Ohio were used by
economically diffuse hunting and gathering groups
(1970:363). While coeval. Blank saw little cultural inter-
action, suggesting seasonally exclusive occupations of sites
where both appeared. He concluded that the Kirk-like
complex of southeast Ohio was initiated by and continued
to receive direct influences from the Mid-South (cf.
1970:366). The resultant Middle Archaic cultures moved
into northwestern Ohio following its abandonment by
inflexible Plano peoples. For Blank, although Early-Late
Archaic continuity anywhere in Ohio was problematic, in
southeast Ohio there was no occupational hiatus. Rather
...this apparent gap in cultural chronology
results. ..from our inability to recognize other than a
limited number of the cultural elements.
(1970:355-6)
Accepting Blank's thesis. Fitting had argued for a
4,500 year abandonment of northern Ohio until the
appearance of a Late Archaic adaptation between 3000
B.C. and 2000 B.C. with the establishment of modern
forests. Neither archaeological nor paleoecological data
offer support for such a model. Fitting had assumed that
the settlement types to be found in northern Ohio fit his
Michigan model. Yet, as early as 1966 (cf. Brose and
Essenpreis 1973) I argued that the drop of western Lake
Erie between 12,000 and 9500 B.P. so altered the
relationships of shoreline-interior resource availability that
no Great Lakes analogy would apply.
Prufer and Long reviewing the northeastern Ohio
Archaic, now suggest that,
...it would be wiser to study.. .the local Archaic
as a fluid continuum rather than along the usual
tripartite lines of fixed Early, Middle and Late units.
This is so because, throughout there seem to have
been no variations in life-style patterns from the
beginning to the end of the Archaic, at least through
the Laurentian tradition.
(1986:50)
Certainly that approach seems required for sites such as
McKibben, House or Lukens Hill, intermittently occupied
from 7000 B.C. to 3000 B.C., with all diagnostic artifacts in
plow-disturbed soils (Prufer and Long 1986:6-10, Table 1 1 ).
In the absence of intra-site spatial analyses, functional
morphological and use-wear studies, or floral or faunal
recovery, it is difficult to see how Prufer and Long could
escape the impressions that the Archaic artifact inventory
was rather static (1986:25) and that the nature of activities
carried out at Archaic sites in differing topographic
locations could not have been very different (1986:26).
However, their tables showing relative frequencies of
chipped stone scrapers, bifaces and debitage; ground stone
tools, notched and stemmed points; indications of chert
heat pre-treatment; and breakage patterns all suggest
significant technological differences between these sites.
Of course, the extent to which differences may be
attributed to site function, as opposed to period(s) of
occupation, could not be evaluated from these mixed
components. Nonetheless, Prufer and Long saw no
discontinuity in Archaic occupation but rather a low
population density with a conservative cultural tradition.
The sealed and/or stratified sites which might have verified
their conjecture remained unpublished, and their report has
spurred me to make some known.
Environmental Setting
Regional Paleoecology
South of Lake Erie a zone of proglacial beach
formations narrows from Michigan disappearing at
Niagara. These deposits lap flat till plains in the west,
while in the east they abut the Allegheny Plateau. With the
12,500 B.P. drop of Lake Erie the islands which define the
lake’s western basin were dry, with a conjoined Thames-
Maumee-Sandusky-Cuyahoga River cutting to a shrunken
basin off Erie, Pennsylvania (Coakley and Lewis 1985).
Tributaries, below former base levels, must have been quite
dynamic. The lake effect, which currently buffers seasonal
differences between shoreline and interior, could not have
existed in the Early and early Middle Archaic. At 9500 B.P.
uplift began refilling the western basins, creating biotically
rich habitats like those of the 19th century Black Swamp.
During much of the Archaic period vegetation represented
local mosaics on uplands (now islands). Altered
groundwater and effective climate resulted in concentric
floristic zones with extensive interfluvial grassland/oak
openings with gallery forests. Through time along the
lower course of the major rivers there was succession from
beech-maple; to mixed-maple; mixed mesophytic; and
finally, mixed oak-hickory and elm-ash as lake levels rose.
South of this first zone, old lake beds extend around
the old Maumee estuary, thence northeastward with
decreasing width to merge with beach formations and
disappear west of the Cuyahoga. The zone is marked by
rolling topography with numerous kettle lakes and bogs.
Most rivers and streams are immature with stable channels
and limited biotic diversity, although there must have been
changes with the short-term lake fluctuations between 9500
and 6500 B.P.
1989
Squaw Rockshelter
23
The Glaciated Allegheny Plateau begins south of the
lake along the East fork of the Rocky River. The escarpment
gradually approaches the modern shoreline, abutting the
Lake at Dunkirk, New York. On the plateau (increasingly so
to the south) streams slow where they encounter more
resistant strata, creating variable valley morphology along
short segments, and major differences in tributary frequency
and catchments between adjacent systems. Due to altitude,
seasonal precipitation differs significantly from other zones.
Although exposure creates edaphic communities, flora
generally is beech-maple forest on the rolling interfluves,
with elm-ash or hemlock valley facies, and mixed tulip-oak-
chestnut or oak-hickory-butternut facies on ridges (Williams
1940; Gordon 1969).
Local studies (Potter 1947; Ogden 1966, 1967; Shane
1975) and regional syntheses (Webb, Cushing and Wright
1983; Davis 1983; Hollaway and Bryant 1985) suggest two
Holocene periods of major floral change in northeast Ohio.
Between 10,000 and 9500 B.P. there was a rapid shift from
the short-lived Hudsonian pine/oak forests to the mixed
Canadian hardwood/deciduous forests of the Great Lakes;
and between 6500 and 4500 B.P. a number of typically
24
Brose
No. 44
Carolinian species (such as chestnut, walnut, and
rhododendron) appeared in the region. The palynological
evidence for an early appearance of modern forests, and for
the lack of any Hypsithermal prairie development in
northeast Ohio, is fully corroborated by faunal evidence
indicating a turnover in the period between 11,300 and
9300 B.P. to communities which are essentially modern,
with little significant change due to Hypsithermal wanning
(Semken 1983:193ff).
Holocene Geomorphology
Squaw Rockshelter lies at the southern edge of the
Defiance moraine (Goldthwait, White and Forsyth 1967). It
sits on the west bank of Aurora Branch, a tributary which
joins the Chagrin River 2.5km below the Falls, and 3km
above the junction with Griswold Creek (Figure 4). The
upper Chagrin above the Falls, and the ower Chagrin below
Griswold Creek, flow through large valleys carved into
Paleozoic rock, and now filled with fluvio-glacial and
lacustrine deposits (Winslow, White, and Webber 1953:41-
43). Through these two segments which originally drained
south (but not in the intervening section) the river now
flows southwest, then north. Earlier Pleistocene drainage
(Rau 1969:10-14) was probably controlled by the Ancient
Chagrin River whose buried valley, at an elevation between
400' and 500' above mean sea level (AMSL), flowed
southward up Aurora Branch (Winslow, White and Webber
1953:41-43, PI. 3).
During the Wisconsinan stadial, ice overrode the
region at least four times. The Woodfordian advance was
marked by the Defiance moraine, about 19,000 B.P. This
moraine crosses the lower Chagrin valley 2km west of the
Falls, just north of the junction with Aurora Branch. The
Defiance moraine was subsequently overlain by a veneer of
Hiram outwash and till (cf. White 1982). With the melting
of Hiram ice, after 14,000 B.P. (White and Totten 1982:48;
1988:66-68), the upper Chagrin River valley drained
southwest to the upper Cuyahoga (Wittine 1970). Its
channel lay 1500m to the east of present Aurora Branch.
Aurora Branch, above the riffles at Solon Road, now flows
north-northwest through a buried valley which, during the
Altonian interstadial, formed a portion of a stream system
which included Griswold Creek and flowed southward into
the Upper Cuyahoga, thence south across the present
Akron divide to the Tuscarawas River. That portion of
Aurora Branch below Solon Road, and the section of the
Chagrin River between the Falls and Griswold Creek, were
abandoned during much of the late Pleistocene.
Reoccupation of the present Aurora Branch valley,
with reversed (north flowing) stream entrenchment and
erosion of the Hiram Till which filled it, began when
headward erosion of the north flowing lower Chagrin River
downcut through the Defiance moraine north of the site
and captured the Upper Chagrin River (Feldmann, Coogan
and Heimlich 1977:192-193). The hydrostatic event which
triggered this erosion was the drop to lowest Lake Erie
levels, around 12,500 B.P. (Lewis 1969; Forsyth 1973;
Coakley and Lewis 1985). Aurora Branch appears to have
downcut to a resistant high of Sharon Conglomerate, and
then shifted 1.5km west to reoccupy its present channel
during the subsequent low-water period (Feldmann,
Coogan and Heimlich 1977:107-142; John Hall, personal
communication, 1983-1985).
With a rapid rise in the level of Lake Erie to 495'
AMSL, fluvial erosion decelerated by 10,000 B.P., and
ended between 7500 and 6000 B.P, with Lake Erie at a
stillstand between 538' and 554' AMSL. There appears to
have been virtually no downcutting between 5000 and
4300 B.P. when lake levels rose rapidly to ca. 590' AMSL
(Coakley and Lewis 1985:198-200, 208-210). After 3900
B.P. there was again a drop to a level controlled by the
Niagara Sill at 555' AMSL. With crustal upwarping after
3000 B.P, there was a slow rise to the present level of
572.5' AMSL (Calkin 1970:Table 2). At present the bed of
Aurora Branch, where it joins the Chagrin River 2km
downstream from the Squaw Rockshelter, is composed of
nearly 200' of alluvial deposits with surface elevation of
830' AMSL. Some 20m upstream from Squaw Rockshelter
the floor of Aurora Branch is exposed on the upper surface
of Bedford Shale (Prosser 1912:529) at an elevation of 854'
Figure 5. Berea Sandstone roof at Squaw Rockshelter
1989
Squaw Rockshelter
25
Figure 6. Stratigraphic profile across Aurora Branch at the Squaw Rockshelter site.
AMSL. Thus, retrenchment of the 3km section of Aurora
Branch in which the site is located, was rather rapid. It
began after ca. 12,000 B.P. when Lake Erie stood below
440' AMSL. The surface of the old buried southflowing
stream in upper Aurora Branch crosses Orange and Solon
townships at an elevation just under 500' AMSL (Winslow,
White and Webber 1953:42) and thus could not have been
tributary to a stream whose surface stood at 500' AMSL,
the lowest floor for sediments within the valley of the
present Aurora Branch-Chagrin River junction (Rau
1969:figure 2, II; 1, 12, 31-32). That suggests that
headward downcutting in the relatively recently captured
northflowing Aurora Branch-Chagrin drainage ended
before 10,000 B.R when a relatively rapid rise brought lake
Erie to a stand at 523' AMSL (cf. Lewis 1969:276; Coakley
and Lewis 1985). While lateral erosion of this segment of
Aurora Branch undoubtedly occurred after 10,000 B.P. it
must have been at a slower rate than during previous
millenia. Indeed, the lower Aurora Branch valley may have
been partially re-filled between 4300 B.P. and 3900 B.P. as
base level seems to have been higher than present (ibid.).
Bedrock Lithology
Squaw Rockshelter formed beneath an overhang of
Berea Sandstone, a thick, variably bedded Mississippian
sandstone, lying unconformably upon thinly bedded late
Devonian shale of the Bedford Formation (Newberry 1870;
Bownocker 1915; Prosser 1912). The Berea Sandstone is
overlain by thin siltstones and shales of the Sunbury and
Aurora submembers of the Orangeville Member of the
Mississippian Cuyahoga Formation (Coogan et al. 1986).
At the site the section forms one of the classic exposures
which led to considering Berea Sandstone as a massive
deltaic channel fill in the underlying Bedford mudflat
shales (Prosser 1912:530-534; Pepper et al. 1954:204, pi 5;
cf. Coogan et al. 1986; however, see Lewis 1988 for an
alternative interpretation) (Figure 5).
About 200m southwest of the shelter. Deer Lick Run
flows north from a small kettle lake at 943' AMSL,
cascading into Aurora Branch 40m downstream of the
Shelter. The Orangeville Member here is an 8' to 10' (2.5-
3m) thick stratum of thinly bedded, soft argillaceous and
grayish-black bituminous shales, interbedded with the
layers of blue fine-grained sandstone and siltstones of the
Aurora submember. These lie upon Berea Sandstone.
The Berea Sandstone at the site consists of three
layers with a total thickness of 42' to 64' ( 12.7 to 19m). The
upper 18' (5.5m) are flat, thinly bedded, ripple marked
sandstone composed of medium to medium fine sand
grains. Below these lie two or three distinguishable strata
of cross-bedded sandstones, with a total thickness of 22'
(6.7m). These are composed of medium to coarse sand
grains. Below these is a massive stratum of festoon cross-
bedded sandstone (Coogan et al. 1986:7) composed of
26
Brose
No. 44
medium to medium-fine sand grains, and having a
thickness of 23' (7m). This stratum occupies a “channel”
deeply cut into the bluish sandy shale of the Bedford
Formation (Figure 6). Across Aurora Branch, Berea
Sandstone was quarried by the Independent Stone
Company beginning somewhere about 1900 (Bownocker
1915). Prior to that the valley was about 35m wide at the
top (cf. Newberry 1870).
Stratigraphy and Sedimentology
Analyses
Sedimentological studies were carried out during and
after the excavation of Squaw Rockshelter (Appendix).
Because Holocene processes operated upon formations
with facies displaying considerable variation in short
distances, results are less than ideal for understanding
deposition during the occupation of the site (cf. Farrand
1985). Under similar circumstances (whatever the actual
antiquity of the site) the quartz sand grains at Meadowcroft
Rockshelter reflect the primary Paleozoic depositional
processes, not the secondary Holocene processes of
archaeological interest. Only with great caution ( contra
Donohue 1976) could shelter evolution be reconstructed
from studies of sediment derived with little transport from
variable sandstones such as the Berea Sandstone at Squaw
Rockshelter or the Connellsville Sandstone at
Meadowcroft. At Squaw Rockshelter profile inspection of
in situ sediments and subsequent analyses distinguished
five strata (Appendix).
Stratum 1
This lowest unconsolidated post-incision fill of the
shelter represents water-sorted Hiram till, deposited along
the undercut wall of Aurora Branch. Samples A1 and B1
both consisted of poorly sorted and poorly consolidated
lenses of dark yellowish brown (10YR 4/6) to greenish
gray (5GY 5/1) fine sandy loam and silt loam with sub-
angular flat lying cobbles of greenish-gray to dark bluish-
gray shale, and small angular to sub-angular cobbles of
reddish-yellow to dark grayish-brown stratified sandstone.
These samples lay conformably upon and graded into the
underlying sediment (Ao), at depths below datum of
140cm and 135cm, respectively. There is a relative
decrease in very fine silt to clay-sized particles to the rear
of the deposit, suggesting some low energy groundwater
movement along the rear wall. No bedding differences
could be noted across what seems to be this same level in
those excavation units where it could be identified (XU 1 ,
5, 3, 5, and the excavation at the north end of the deposit),
although there were fewer cobbles and finer sediments in
higher levels.
Stratum 2
These deposits, encountered at 110cm and 125cm
below datum consisted of a very compact lens of dark
grayish brown (10YR4/2) to dark strong brown (4.5YR
4/6) medium to coarse sandy loam flecked with granules of
charcoal, and lying conformably upon the underlying silt
loam and inclusive cobbles. This lens, which ranged from
10cm in thickness at the front of excavation Unit 3, to
20cm in thickness at the back of Excavation Unit 5, was
virtually free of inclusive cobbles. As identified
lithologically it occurred in all areas of the site except for
the eastern halves of Excavation Units 2 and 4 and the
excavation where recent erosion appears to have removed
it. Within Excavation Units 1, 3, and 5, where continuous
stratigraphic exposures could be observed, the upper 5 to
10cm of the deposit contained small areas with thin
laminae of yellowish brown (10YR 5/4) or light grayish
brown ( 10YR 2/4) fine silts and clays. These laminae were
more common to the south. Horizontally lying flint tools
and chipping debris were encountered throughout this lens.
Within the lowest and thickest portion of this deposit
running about 160cm along the southern wall of
Excavation Units 1 and 5, 115cm below the datum, human
skeletal remains were encountered within a thin zone
(Feature 2) 30cm wide. This stratum appears to represent at
least the basal portions of an anthropic epipedon. The
relative removal of silts and clays, the horizontal placement
of larger (cultural) particles, the few finer laminar facies
within this deposit, which parallel the overall westward dip
toward the rear wall and strike to the north, all suggest that
this level represents some portion of a surface in part
deposited by gravity, minimally altered by prehistoric
occupation, and then significantly reworked by low energy
water transport. In Excavation Unit 4, at a depth below site
datum of 120cm, a concentrated lens of oak charcoal
associated with the midsection and tip of a lanceolate point,
yielded a radiocarbon date of 9480±160 B.R (DIC-586).
This stratum is called Cultural Level II.
Stratum 3
This is an unconsolidated zone of numerous large
angular cobbles and boulders of cross-bedded Berea
Sandstone within a matrix of fine to very fine sands.
Encountered in all excavation units between 95cm and
110cm below datum, this stratum was about 45cm thick at
the front of the shelter and about 65cm deep along the rear
wall and lay conformably upon Stratum 2. Stratum 3
appears to represent in place chemical weathering of
material derived by mechanical processes from the massive
Berea Sandstone “channel fill” which forms the present
back shelter wall and which must have formed the shelter
overhang during the deposition of Stratum 2. The massive
blocks of cross-bedded Berea Sandstone which presently
lie between these unconsolidated deposits and the Aurora
Branch, seem to lie directly upon Stratum 2. Where recent
rodent disturbance has not confused subsequent
depositional sequences (in XU 1 and 3), Stratum 3 appears
to have been deposited between these blocks and the rear
wall of the shelter.
1989
Squaw Rockshelter
27
Stratum 4
This deposit lies conformably on the surface of
Stratum 3. It is a 15cm to 25cm deep zone of yellowish
brown (10YR 5/4) medium to coarse loamy sands
including relatively few pebbles or small cobbles. Large
cobbles of massive unbedded Berea Sandstone occurred
throughout the stratum, however, and while portions of
stratum 4 were encountered at depths between 55cm and
70cm below datum in every excavation unit except the
northern test trench, it was not possible to trace this
deposit continuously for distances greater than 80cm in
any direction. Indeed the correlation of Stratum 4 in
excavation units 1, 5, and 3 and in Excavation Units 2 and
4 is a conjecture largely based upon similar granulometric
analysis and relative stratigraphy. Although no vertical
sorting or pedogenic structures of any sort could be
identified in Stratum 4, there was some clear front-to-
back sorting with coarser sands at the front. Several very
shallow cultural features were encountered within this
stratum, and their origins varied in surface elevation by as
much as 20cm. The most unambiguous of these. Feature
3, encountered at a depth of 70cm below datum in XU 2,
consisted of a shallow oblate depressed area of fire
reddened (2.5YR 3/5) silty clays about 25cm by 20cm in
surface area and about 5cm thick. This ‘hearth’, located
about 45cm from the rear shelter wall, was about 10cm
deeper in the center than at the periphery. The base of the
center rested on a block of sandstone buried in Stratum 3.
Within this depressed central area a matrix of dark
yellowish brown (10YR 5/4) to dark grayish brown (10
YR 4/2) silty loam contained small flakes of chert,
charcoal identified as maple, and a small charred
fragment of walnut hull. The wood charcoal fragments
yielded a radiocarbon date of 5500±85 B.R (DIC-321).
Between this feature and the rear wall, at slightly
shallower depths, were two discontinuous horizontal
areas of more compact and darker ( 10YR 3/2) silty sands,
each about 10cm thick. These zones yielded several lithic
tools and flakes and a single fragment of charred cordage.
A second possible hearth. Feature 1, was encountered at
the northern edge of Excavation Unit 1 between 75cm and
85cm below datum. No fire-reddened subsoil was noted,
but a lens of sandy ash (pH 6.0) filled the central portion
of this circular clayey lens, some 30cm across. Between
the northern periphery of Feature 1 and the block of fallen
sandstone fronting the river, were loose sandy silts
riddled with recent rodent burrows. The feature contained
no organic material, although an expanding stemmed
projectile point was recovered from the compact silts at
the southeastern edge of the feature. Several other formal
lithic tools and as well as lithic debitage were recovered
from this stratum in excavation units 2, 3. and 5, but no
other cultural features and no other areas of apparent
concentration were found. Stratum 4 is called Cultural
Level 1.
Stratum 5
This stratum rests unconformably upon the surface of
Stratum 4 with a clear contact. Many of the upper surfaces
of the large cobbles and blocks of Stratum 4 (and several of
those which obtrude from Stratum 3 through Stratum 4) are
incorporated into the lower portions of this 40cm to 60cm
thick zone of dark grayish brown (10YR 4/2) to dark
yellowish brown (10YR 3/2) sandy loam and sandy clay
loam. Large cobbles of Berea Sandstone and pebbles and
small cobbles of siltstone from the overlying Orangeville
member are entirely incorporated in this matrix. At the
time of excavation the surface of this stratum in Excavation
Unit 2, 3, and 5 was capped by a 20cm to 35cm thick zone
of decomposing organic material (leaf litter, dead rodents,
cigarette filters) and some amount of very recent inorganic
trash. Humic acid staining from this Ao horizon had
penetrated to variable depths within the upper 20cm of
Stratum 5, but no soil development had occurred (cf.
U.S.D.A. 1977).
During the accumulation of at least the lower portion
of Stratum 5 sediments apparently weathered from some
portions of the Orangeville Member and perhaps even
some weathered from Hiram Till were being introduced to
the front of the shelter by a higher energy transport
medium. This situation seems to have ceased during the
period in which the upper portions of the stratum were
deposited. While unconfirmed by stratigraphic excavation,
it is possible that the large foreset-bedded Berea Sandstone
block presently forming the southeast wall of Excavation
Unit 4 fell onto lower Stratum 5 during this depositional
episode. Unfortunately, the area along that face of the
Excavation Unit 4 block and the entire stratigraphic
sequence above Stratum 3 were represented by rodent
disturbed deposits (Sample X). All cultural materials
encountered within Stratum 5, whether prehistoric or
recent, were recovered from such rodent-disturbed areas in
Excavation Units 1, 3, and 4. The presence of mid-20th
century material in these areas suggests that their absence
on the surface of Stratum 5 is due to Metropark cleaning
efforts since the 1960’s.
Summary
Most non-cultural sediment within the site is
attributed to the in-place deposition of sands and silts
chemically and mechanically weathered from bedrock.
There was sheet runoff in portions of the shelter, and with
the exception of a short period around 5500 B.R the net
effect of water transport after 10,000 B.R appears to have
been erosive. Sedimentological analysis and the vertical
and horizontal stratigraphy within and downstream from
the shelter suggest three relatively distinct episodes of
human occupation of variable intensity and importance.
Squaw Rockshelter appears to have been relatively deep
when occupied about 9200 B.R The shelter was more
exposed when occupied around 5500 B.R Finally, the
r r
28
Brose
No. 44
v ' N .kH *.
Figure 7. Excavation of Unit 5 showing limits of space.
barely protected surface of shelter deposits has been used
intermittently during past decades. Evidence for anthropic
episol formation appears only in discontinuous portions of
Cultural Level II, and at the top of Stratum 5 postdating
aboriginal use.
Archaeological Recovery
Methods
Rockfall made shovelling impossible within the
confines of excavation units (Figure 7). All sediments were
removed by hand and trowel. When a distinct stratum
surface was encountered the contact was cleaned across the
floor of the excavation unit. Large fallen blocks often made
this procedure difficult to follow. After their non-cultural
origins had been confirmed, there was no screening of
strata 5, 3, or 1. Although all feature matrix and fill was
screened through 3/32" mesh, fine screening of cultural
“floors” was not systematic. No flotation samples per se
were collected at the site, but flotation was undertaken on
each soil sample collected. These methods yielded
fragmentary and complete mollusk shells and minute
fragments of wood charcoal. No pollen was preserved in
any of the sediments inspected, and no attempt to identify
phytoliths was made. Overall, fine screening and flotation
yielded little independent data on the proximal
environment of Squaw Rockshelter when Cultural Level II
(9480± 1 60 B.P.) and Cultural Level I (5500 + 85 B.P.)
sediments accumulated.
Mollusks
Analysis of recovered mollusks was performed by
Brose in 1984. A total of 36 identifiable gastropods and 61
Table 2
Identifiable Molluscs Recovered from Squaw Rockshelter
Cultural Level I
Cultural Level II
Aquatic Gastropods
N (% class)
N (% class)
Valvata tricarinata (Say)
2 (15.4)
•
Stagnicola sp.
6(46)
1 (50)
S. reflexa (Binney)
3(23)
•
Helisoma sp.
2(15.4)
1 (50)
Subtotal
13 (99.8)
2(100)
Terrestrial Gastropods
Stenotrema fraternum
4 (17.4)
3 (33.3)
Triodopsis alholahris
8 (34.8)
1 (H.l)
Zonitoides arboreus
6 (26.1)
1 (ll.D
Discus cronkhitie
5 (21.7)
4 (44.4)
Subtotal
23(100)
9(99.9)
TOTAL
36
11
1989
Squaw Rockshelter
29
Table 3
Charcoal Fragments Recovered from Squaw Rockshelter
Cultural Level I
Cultural Level SI
Genus
N (% ID)
N (% ID)
Pinus
5 (9.6)
10(23.3)
Tsuga
16(30.8)
8 (18.6)
Acer
10(19.2)
8 (18.6)
Quergus
5 (9.6)
17 (39.5)
Fagus
8 (15.4)
0(*)
Ulnuis
5 (9.6)
0(0
Jugulans
3* (5.8)
0(0
Total Identified
52(100)
43 (99)
Unidentifiable
21
31
TOTAL
73
74
*
2 fragments of charcoal and 1 charred nut shell fragment
unidentifiable fragments was recovered from Cultural Level
I, while 1 1 identifiable gastropods and 27 unidentifiable
fragments were recovered from Cultural Level II sediments
(Table 2). No bivalves were recovered, suggesting that all
mollusks present were due to non-cultural factors. Even
standardized for the relative volumes of sediment analyzed
from Cultural Level II (.94m3) and Cultural Level I
(1.75m3) it is clear that there was a somewhat greater
frequency of gastropods of all types in Cultural Level I than
in Level II, and there was a significantly greater frequency
of aquatic gastropods in Level II.
There are few environmental differences among the
recovered aquatic gastropods. All are extant and have
inhabited the region since the mid-Pleistocene, living on
emergent vegetation in a wide range of energy
environments (LaRoque 1968:367-506). The relative
paucity of Cultural Level I aquatic gastropods is consistent
with sediment analyses suggesting little flooding of the
shelter during its early history.
Differences in overall frequencies of the terrestrial
gastropods from the two cultural levels reflect both the
relative loss of shelter overhang during the millennia
between depositional events, and the greater duration of
Level I sediment accumulation. Differences suggested by
the habitat preferences of the 47 terrestrial gastropods must
be tempered by small sample sizes, the unsystematic
recovery strategy, and by the assumption that for over
4,000 years the shelter represented a varying taphonomic
catchment. Ignoring these caveats, the shift from an
assemblage of Stenotrema fra termini and Discus
cronkhitei to an assemblage dominated by Triodopsis
albolabris and Zontoides arboreus reflects the change from
cool damp hardwood/conifer Canadian forests to forests in
which beech-maple and mixed Carolinian mesophytic
hardwoods predominated. However, all of these species are
extant locally and have been present since pre-
Wisconsinan time (LaRoque 1968:570-680).
Charcoal
Fragments of charcoal and charred nut hull (Table 3)
recovered from Squaw Rockshelter were identified in 1976
by Dr. F. DiMarinis, Cleveland State University Biology
Department. None of the uncharred seeds or nut hulls
recovered from potentially rodent disturbed soils (x) were
included in these analyses. Assuming these specimens have
been humanly introduced to the shelter, each assemblage
represents a mixture of local availability and cultural
selectivity. The absence of walnut and beech in level II is
predictable, since neither were present in northeast Ohio at
9500 B.P. (Webb et al. 1983:154-157; Davis 1983:173).
Elm, however, should have been even more common than
at present and its absence from Level II samples may be
due to the fact that deadfall of prefered species was more
easily obtained. Hemlock is presently a common species in
the Chagrin River valley ravines (Williams 1940:19-23)
and apparently has been for millennia. Other significant
differences between the assemblages, from one dominated
by oak and pine at 9500 B.P. to one of beech and maple at
5500 B.P. may reflect general environmental availability
rather than shifts in cultural preference.
The fragment of charred walnut hull from Level I is a
poor seasonal indicator, walnuts in the hull being storable
for several seasons. Certainly, plants yielding fruits widely
utilized throughout the prehistoric record existed in the
proximity of the shelter, so that the lack of other charred
seeds could be considered at least seasonally indicative had
the recovery methods been more thorough.
30
Brose
No. 44
i METRIC li
, , : 2| , ,1,3;
1 4i
5
6
! o ' | j m ' : j ; | j n s j !
ill! ;|]j
jm|
1 1 [ j i !
Figure 8. Fragment of charred cordage from Cultural level I.
Charred Cordage
A single piece of knotted cordage was recovered
during the excavation of Excavation Unit 4. Within the
sandy Stratum 4, at a depth of 70cm, Feature 3 consisted of
a shallow fire reddened silty clay hearth 25cm in diameter
and 5cm thick. Between this hearth which contained flakes
of chert and charcoal dated to 5500±85 B.P., and the rear
shelter wall 45cm south, slightly shallower compact dark
silty sands 10cm thick yielded several lithic tools and
flakes and a single fragment of charred cordage. The
cordage, analyzed by Drs. Andrews and Adavasio (this
volume), consists of a single fragment of spun and twisted
plant fibers knotted at both ends. It may have been a
knotted clothing fringe, a bundle of construction material,
or a part of a snare or trap (Figure 8).
Human Skeletal Material
From the lowest portion of Stratum 2, running for
160cm along the southern wall of Excavation Units 1 and 5
at a depth of 115cm, fragmentary remains representing at
least two individuals were recovered within a thin zone
(Feature 2) 30cm wide. The stratum is a dark brown sandy
loam flecked with granules of charcoal dating to 9200+150
B.P. Feature 2 within this stratum was from 10 to 20cm
thick at the back of Excavation Units 1, 3, and 5, and
contained small areas with thin laminae of yellowish or
light grayish-brown fine silts and clays. These laminae
parallel the rear wall of the shelter and represent a surface
reworked by low energy water transport, perhaps
seasonally. Based upon the analyses (Prior, this volume)
the majority of the material is from a young adult female,
additional individuals being represented by three isolated
teeth not in association. Unfortunately, while adding to a
limited data base (Protsch 1977), the skeletal material
does not allow much comparision of Early Archaic
regional demography or morphology possible at other
Table 4
Combined Squaw Rockshelter Debitage Categories
Cultural Level II
Used
Unused
Total
Block Cores
Upper Mercer
•
2(1)
2(1)
Plum Run
•
2(2)
2(2)
Glacial
•
•
•
Pebble Cores
Glacial
e
1
1
Decortication Flakes
Upper Mercer
•
•
•
Plum Run
9
•
•
Glacial
4(2)
9(3)
13(5)
Block Shatter
Upper Mercer
9
4(3)
4(3)
Plum Run
1 (1)
5 (4)
6(5)
Glacial
2
3(2)
5(2)
Thinning/Retouch Flakes
Upper Mercer
5(3)
3 (3)
8(6)
Plum Run
1 (1)
8(8)
9(9)
Glacial
2(1)
4(1)
6(2)
Subtotals
Upper Mercer
5 (3)
9(7)
14(10)
Plum Run
2(2)
15 (14)
17(16)
Glacial
8(3)
17(6)
25(9)
TOTAL
15 (8)
41 (27)
56 (35)
1989
Squaw Rockshelter
31
sites ( e.g . Charles and Buikstra 1983; Redder 1985; Dickel
and Doran 1988).
Chipped Stone
During the excavation of the shelter 168 pieces of
debitage were recovered. Application of several
distributional analyses (cf. Brose and Scarry 1977), adjusted
for volume of sediment from each excavation unit, does not
indicate statistically significant clustering. No doubt, more
rigorous analyses (Vance 1987:59) would suggest
otherwise, but the waterworked surfaces of Cultural Level
II, the longer but unknown erosional processes during the
formation of Cultural Level I, and the large size of most
debitage recovered, argue that concentrations would be lag
deposits, geological rather than cultural in origin. Those
processes urge caution in any attempt to reconstruct not
only spatial but technological aspects of lithic reduction
during either period of occupation. The distribution for each
cultural level (Tables 4 and 5) is presented in terms of
debitage categories as stages in the manufacturing trajectory
(Brose 1970:97-106, 1978; Prufer and Long 1986:42-44).
Analyses of debitage weight and surface area reveal that
debris from all stages of biface reduction and use are
present in both occupations (Stahle and Dunn 1984:4-9). In
that sense, the Early Archaic lithic procurement and
reduction is similar to the Kirk occupation at the Calloway
Island site (Chapman 1979:36-62).
Identification of lithic sources was primarily visual,
although petrographic analyses confirm some attributions.
The nearest outcrops of the black Upper Mercer flint lie
55km east, along the Portage/Trumbull County line (Prufer
and Long 1986:18). The mottled orange to greyish-blue
Plum Run flint is from a small quarry 70km south in Stark
County (Murphy and Blank 1970: 198). Glacial flints are
mostly brownish-grey, dark grey, and mottled blue-grey
Palaeozoic cherts of the Niagaran formations in Ontario
(Parkins 1977), present in local stream gravels throughout
glaciated portions of Ohio (Brose et al. 1981). Heat
treatment was identified by the presence of texture and
color changes, as well as by “pot-lid” fractures underlying
primary flaking scars.
Utilization of debitage edges for expedient tools was
determined by low powered opaque microscopic
observation, following methods described by Frison
(1968), Brose (1970, 1975, 1978), Tringham et al. (1974)
and Keeley (1980). Although no significant changes in
knapping chipped stone can be demonstrated within Squaw
Rockshelter, procurement and reduction differed between
levels. There is a small, but statistically significant increase
Table 5
Combined Squaw Rockshelter Debitage Categories
Cultural Level I
(Number showing heat-treatment)
Used
Unused
Total
Block Cores
Upper Mercer
•
3(1)
3(1)
Plum Run
2(1)
3(2)
5(3)
Glacial
•
•
•
Pebble Cores
Glacial
•
4
4
Decortication Flakes
Glacial
6(2)
8(4)
14(6)
Block Shatter
Upper Mercer
1 (1)
7(5)
8(6)
Plum Run
5(2)
12(8)
17(10)
Glacial
8(2)
11 (6)
19(8)
Thinning/Retouch Flakes
Upper Mercer
4(4)
7(5)
11 (9)
Plum Run
1 (1)
8(7)
9(8)
Glacial
7(1)
15(5)
22(6)
Subtotals
Upper Mercer
5 (5)
17(11)
22(16)
Plum Run
8(4)
23(17)
31 (21)
Glacial
21 (5)
38 (15)
59 (20)
TOTAL
34(14)
78 (43)
1 12 (57)
32
Brose
No. 44
Figure 9. Projectile points from Cultural Level II. See Table 6 for sizes.
in the frequency of the more local of the three sources of
raw material used from the Early Archaic (9300 B.P.) to
the late Middle Archaic (3300 B.P.). The Plum Run
quarry was initially favored over the closer Upper Mercer
outcrops, possibly because the former source displays
thicker bedding. Through time, however, there was a
greater utilization of the local glacial cherts for the
production of expedient tools. This localization of
procurement is related to the highly significant decrease
in heat treatment in utilized and unutilized debitage. In
the earlier period almost all Plum Run flint was heated,
while in the later period only one third was. A rough
index, comparing initial reduction stage decortication
flakes and block shatter, to final reduction stage thinning
flakes, suggests little change over four millennia.
However, it is unlikely that all areas of a rockshelter
would have been equally attractive for performing
different aspects of flint knapping. Expedient tool use was
more likely to have taken place in the well-lighted shelter
opening, while curated tool loss was more likely to have
occured in the darker rear areas of the shelter. Most
frontal portions of Level II had been lost to erosion before
Level I formed so that if similar portions of the shelter
were used for rather similar activities differing segments
of the production sequence could be recovered from
sequent levels in a single unit, even if overall reduction
strategies were similar. Gramley (1980) discussed site
function in terms of a curatorial index, comparing ratios
of bifaces and formal tools to the ratios of expedient tools
and utilized flakes. Fitting (1967) had used a similar
index to infer group sexual composition. Clearly, the
interaction of site function and demography is responsible
for the fact that at Squaw Rockshelter 40% of lithic
artifacts in the Early Archaic Level II were curated, while
only 20% were in late Middle Archaic Level I.
Beyond the utilized debitage, 28 formally produced
tools were recovered from the shelter. Not every
provenience from the site is unimpeachable, and even
though there is stratigraphic separation between cultural
levels, neither demonstrate the internal integrity needed for
satisfactory social interpretations. Nonetheless, these
examples of local style and technology can be compared to
coeval assemblages in other regions to illuminate the Early
and Middle Archaic occupations in northeast Ohio. Indeed,
the bifacial projectile points from the Early Archaic level
are evidence for the interaction of cultural patterns hitherto
isolated in Ohio, with exceptions so infrequent their very
existence had been questioned.
Table 6
Projectile points from Squaw Rockshelter (33CTJ34) Cuyahoga Co., Ohio, Level II
(in millimeters)
1989
Squaw Rockshelter
33
c
Ol
E
E
o
U
3 H
>
cS
Cu
Oh
o
o
cd
X)
OX)
C
o
-CD JD
S 'Oh
o 2
X 04
)— i >
xr x
- o
04
> X
04 o
X o
©X)
C
2
£
0
04 X
X On
73 04
1 3
O .2?
db ^
£ £
> S
cd
04 c
X 3
C/D Cd
8> E
X 3
73
C
cd
11
o —
Otj 3
3 E
-O ^
•g 3
a )
cp
S
ON
cd
Cu
C/D
Cd
cd
cp
Xi
UM
04
X
OD
rn
C
ip)
Of)
04
ON
'S
3
O
S3
c
Q
’E
D
5-h
c
cd
04
04
S_
D
-x
X
Id
D
04
c
Z3
, .
Lh
g
s
£
£
-2
c
X
04
X
E
C/D
cd
04
C
o
X
c
cd
C/D
& on
04 X
> o
^ 3
X Of)
O E
'% 6
E c
O 03
CQ
O CQ
3 o
c/d cd
04 x
■3 E
« -r
2;
60 J3
3 ^ _
04 4-h XJ
— (1) On
2 S3 2!
22 c 3
3 g .2?
pu
c/)
-a 3
03 c 3
H 03 r3
U « *"
« oo c
E § p£
cd
04
-1
E
53 |
CQ
. X
04 0)
-j cd
« 2
=j 2
o c
CL) O
x a>
c s
3 »-
2 13
“•■3
P> CD
•? E
03 60
3 C
— o
£ 3
00 t/3
T3 £
LU 5
. ©£)
H _Q4
"04
> 2
04 (L)
X X
04 04
X J-
3 O
r~| ©X)
©Xj X
C
'S 2
C r-
04
2 S
, — 1
04
04
n
c
Of)
Qh
4_J
c
E
4_>
o
a
O-
'S
04
X
X
'S
U
o
D
’>
c
D
C/D
C/D
C/D
04
4— *
OXJ
04
04
C/D
O
c
o
i—
Of)
C/D
E-
c
cd
CJ
cd
04
cd
ee
C
04
Os
_<
PI
GO
X
X
X
CP
IP)
X
c
X
WD
oo
p
o
e
04
#
ON
»
pi
X
X
X
P;
P
CP
CP
X
pi
ON
pi
pi
CP
£
OX)
-o
d>
a
X
03
CO
X3 .g
d>
£ -X
C / 3
C >
oij o
>A c
> o
d>
>
d>
X
*5
Dh
-a
d>
OX)
E
c
c
o
>
C
> o
l-H C
d> d>
> be
X
OJ CO
.2* «
X) 3
O bO
OX) rZ
.E
D- c
n. ■ !=
03 C
C
3
Q£
E
3
aj ex
3 2 Q.
_G 3
-J D.
~ a.
3
E-
c
. o
c/3 -r*
S..2
T3
d) C/3
X
X C
X 03 ,
td ^
j: d>
x x
.
3 D
O O 04)
X jr1
c/3 OX) Uh
C 03
O O
C/3
03 d) X
03 OX)
T» X
— 1 OX)
d> c
OX) O
s— -—h
o3 03
^ X
• o
^ 3
■s s
> d)
d) *~
X a.
. a>
^
>
d)
X
*c
d>
Of)
5
X
c
D
D
g
" •—
C/3
D
z
D
O
X
C/3
d>
OX)
C/3
&
X
O
i—
C/3
O
z
.12
£ “O
Q W
<
Z
o
x
o
o
00
aj 0/
I
in a
o
\D
‘t: OX)
“ e
SS 5
rj
<
z
<
z
<
z
3
sc
3|-
CN
<
z
<
z
3 ~
i r. "O
3 •—
so £
o
[V
<
z
c
X
ON
00
o
04
— or
Os
O'’
04
X
Of)
c
QJ
X
x
CO
or
04
04
of
oT
Os
ir~t
sc ©
Z z
g ^5
U CJ
<
<
CQ
<
CQ
X
O'
0
0
O
Tf
^r
•o
•o
0
O
0
0
QO
00
00
00
00
(bifacial) convex scraper. Minimal retouch
on distal ventral surface. Transverse
striations on all ventral edges. Heavy
platform crushing. Glacial chert. Figure 13e.
42
Brose
No. 44
that each areally extensive sealed horizon shows a waxing
and waning of differing projectile point hafting categories,
but that every unmixed sealed Koster component yielded
from three to seven named corner-notched, side-notched,
stemmed and lanceolate point types. The variable haft
styles showed definite clusters so that any one of the
activity foci at any occupation level might yield a
predominance of one haft category (cf. Brown and Vierra
1983). Similar situations seem to have existed in Missouri,
(O'Brien and Warren 1983:82-93, fig. 5.13), and in central
Illinois (Lewis 1983:102-105), and also at Modoc Shelter
in the Mississippi Valley (Styles, Ahler, and Fowler 1983).
Montet-White (1973) documented the statistical
significance of this phenomenon, but failed to explicate
potential causes. The faunal assemblages indicate similar
diffuse hunting/gathering economies with small game
predominant. From Late Paleo through Middle Archaic
there is unambiguous evidence for increasing reliance on
seeds and nuts (Brown 1983).
In the east, a sub-continental synchrony of change in
Early Archaic projectile points types seems anchored by a
suite of radiometric determinations. It has been argued
(Brose and Lee 1980) that hafting morphology within any
eastern point “type" shows as much variability as do the
sets of different point types from coeval sites to the west.
Even allowing Coe (1964:123) to order the undated site
profiles as best he can, his classic excavations do not
exactly show one point type for one level.2 This is clearly
seen in his summary figure (1964:121). From 8400-7500
B.P. Palmer and Kirk corner-notched, Kirk side-notched
and stemmed, and Stanly stemmed co-occur. From 7500-
6500 B.P. Kirk stemmed and corner-notched, Stanly
stemmed. Morrow Mountain I, and Guilford Lanceolate
points exist at virtually every level.
Excavation of the St. Albans site along the Kanawha
River in West Virginia exposed two areas of Early Archaic
and stratified early Middle Archaic occupation (Broyles
1971:19- 20, 47-48). Type descriptions and accompanying
illustrations (1971:53-76) suggest morphological and
metric continua from large Kirk corner-notched, to Amos
corner-notched, to Charleston corner-notched, to small
Kirk corner-notched, to large Kirk corner-notched; and
from Kanawha stemmed, to Kirk stemmed, to MacCorkle
stemmed, to St. Albans side-notched variety B, to St.
Albans side-notched variety A, to LeCroy Bifurcate base,
to Kanawha stemmed. Indeed, Broyles herself (1971:71)
noted this intergradation. Brose and Lee (1980) suggested
that using the type descriptions, only Broyles’ Kessel side-
notched projectile points (1971:60-61) might be resorted
into the same type. The statement that, “Each type of
projectile point was confined to one or two zones...”
(Broyles 1971:24), even ignoring presumably displaced
Kirk and St. Albans points at the site (ibid.), reflects a
situation where each “type” is arbitrarily segregated.3
The Little Tennessee river terraces revealed deeply
buried sites, stratified from late Paleolndian through
Middle Archaic. At Icehouse Bottom (Chapman 1973,
1976); Rose Island (J. Chapman 1975, 1977); Patrick
(Chapman 1977) Harrison Branch, Calloway Island, and
Thirty Acre island (Chapman 1977), Bacon Farm
(Chapman 1978), and at Howard Farm, with the largest
Early Archaic exposures (Chapman 1978) there are
consistent series of what Chapman (1977:52-55) called
projectile point/knife Type Clusters. There is a Kirk corner-
notched cluster dated between 9500 B.P. and 8500 B.P.
(although Chapman prefers 9400 B.P. to 9100 B.P.) a
Bifurcate Point Cluster between 9400 B.P. and 7800 B.P.; a
Stanly (Kanawha?) Cluster between 8000 B.P. and 7000
B.P. and a Morrow Mountain phase from perhaps as early
as 7000 B.P. (Chapman 1977; 16 1 - 167). Within each
discrete stratigraphic level, at each site where the “type
cluster” exists, a variety of hafting morphologies co-occur.4
Thus, even in the prototypic sealed stratified
floodplain sites of the Appalachian southeast. Late
Paleolndian stemmed and unstemmed lanceolate points,
and Early Archaic side and corner-notched projectile points
co-occur. Further, in the levels running into the Middle
Archaic, the named “type clusters” represent as much
variability in hafting as would define distinct projectile
point types in the midwest. It seems clear that, not only is
there some importance to the excavated area of relevant
levels (Koster = 625m2; Napoleon Hollow = 435m2; St.
Albans = 240m2; Icehouse Bottom = 170m2; Neville =
54ni2; Hardaway = 42m2; Gaston = 28m2) as I had
suggested (Brose and Lee 1980), but that differing schools
of typology might be responsible for the apparent
difference between Archaic horizons in the east and the
midwest. If Early Archaic assemblages from the riverine
midwest and the southeast indicate the potential variability
in single components between 9400 B.P. and 4500 B.P.,
then stylistic or functional lithic variability is also expected
in the ecologically complex region of northeast Ohio.
Local Lithic Assemblages
While diagnostic projectile points have been
recovered from loci throughout northeast Ohio, which
appear to be relatively “pure” Plano, “pure” Appalachian
Archaic, or “pure” Laurentian, given the environmental
complexity, there is little reason to expect that all Archaic
sites would display similar variability. The McConnell
workshop at the Nellie Chert quarry in the Glaciated
Plateau, with a preponderance of lanceolate points is one
example of a site of limited function, although even there
other types were manufactured from specific preforms
(Prufer 1963:13, 31 fn.2). The recovery of only Matanzas/
Brewerton eared-notched points from the small Merkle 2
site, on a buried Cuyahoga River terrace (Brose 1975c: 14,
20) related to a channel shift long after 8540±70 B.P.
(Miller 1983), appears to be a site with similarly limited
lithic style.
1989
Squaw Rockshelter
43
Based upon recent paleoecological and
archaeological data, Brose and Lee (1980) argued that
while some Archaic sites in northern Ohio may have been
drowned, points recovered from the intermoraine region
suggest that rather large sites now lie buried along major
rivers, or were located in the adjacent uplands (Brose
1975c, 1976a, Brose et a! . 1981). Those sites, not functional
equivalents of smaller occupations, should yield a
considerable range in projectile point styles often
considered representative of discrete Late Paleolndian
through early Middle Archaic occupation. 1 do not intend
to suggest that multi-component sites with millennial
lacunae do not exist: they surely do, and the Welti site
(Fitting 1963), the Hospital and Academy sites (Brose
1 975b:24-38), the Norman ‘P’ site (Brose 1976b) and
perhaps the McKibben site (Prufer and Sofsky 1965)
appear to be of that type. But, just as small “pure” sites
have been reported, larger sites occupied between 9000
B.P. and 4500 B.P. exist and, like Squaw Rockshelter,
display considerable lithic variability. Rather than an
arbitrary segregation into Paleolndian and Archaic
occupations the assemblage from these sites may represent
a single component.
Among reported sites of this type are the Hogue’s
Spring site in Cuyahoga County, Ohio (Brose 1975c: 15-
lb), and the Holdson District sites 2,3, and 4 in Ashtabula
County, Ohio (Brose 1977:83- 91). The recovery of Kirk,
Dalton-like, and St. Albans side-notched ‘A’/LeCroy
Bifurcate Base projectile points, along with ovate bifacial
knives in salvage of the Regis site, on a plateau above a
tributary of the Cuyahoga (Brose 1975c: 17,48) in Summit
County, Ohio, may also be noted. Several undescribed sites
known through test excavations also illustrate the lithic
variability of this period.
The Cooper Hollow site on a bluff of the deeply
incised Vermilion River in northwest Lorain County was
identified by a band of unprincipled amateurs. I tested the
area in the fall of 1972 and extensive excavations were
conducted by Lee through the summer of 1974. The
excavations revealed a sub-plowzone paleosol containing
discrete areas of firecracked rock. Within a 180m2 zone
there was a large focal cleared zone and a large focal
roasting pit, as well as several shallow pits. The overall site
configuration is rather like that at Holcomb Beach (Fitting,
DeVisscher, and Wahla 1966). The site yielded twelve
complete finely made lanceolate bifaces which in hafting
morphology grade from narrow based Cumberland to
stemmed Scottsbluff lanceolate bifaces. They duplicate
points from the Fisher site in Ontario (inspection of
materials at the R.O.M. courtesy of Dr. Peter Storck, 1988).
Twenty tip and midsection fragments also reflect this
continuum. Lateral edge grinding is common but not
ubiquitous. From sealed features and paleosol Lee
recovered a corner-notched/stemmed point, and the lower
lateral portion of the base of a Stringtown Spured-Stem or
Hardin Barbed point. Excavation also yielded “Micro” drill
tips, rough flake tools, cobble tools and choppers, and over
1100 fragments of debitage (15% used). Flotation of the
roasting pit produced fragments of nut shell and
fragmented calcined bone representing a large cervid. It
also contained charcoal yielding an unacceptable
12,100±250 B. P. date for the Paleolndian-Early Archaic
occupation.
The Zigler Farm site, on the lake plain in
southeastern Lorain county, sits on a knoll at the edge of a
bog draining the West Branch of the Black River. Testing
by N’omi Greber in the summer of 1973 recovered a single
Cumberland point, two edge ground and four unground
Scottsbluff points and five McCorkle-like points from
undisturbed paleosols in ground undulations.
The Burrill Orchard site, sits on the point where
French Creek joins the Black River in Lorain County.
Excavation by Brose in the summer of 1971 revealed
discontinuous areas of sealed paleosol and two pits.
Among minute fragments of charred bone, the pits yielded
five points grading from Hardin Barbed to Scottsbluff, and
three points intermediate between Kanawha Side-notched
and St. Albans Type B. Edge grinding occurs on most
points. The presence of nearly lm of overburden, and
wishes of the sometime owners to preserve a peach
orchard, prevented exposure of continuous site areas.
Subsequent excavations 40m south along the same plateau
(Brose 1978b) revealed an arc of post holes associated with
a charcoal and ash-filled pit dated 7 1 20± 1 50 B.P. (DIC-
734) containing one Hardin Barbed point, and one point
which Chapman (c.f. 1 977:30-35, 163-168) calls
Stanley/Morrow Mountain II.
With the clarity of hindsight, we can argue that not
all of the “sites” documented by Prufer and Baby were
multi-component. The data used to construct their Paleo
Complex were abstracted, post hoc, from assemblages
which contained bifurcate-based, expanding stemmed, and
notched points in numerous varieties and frequencies. At
the Sawmill site in Erie County, whose late Paleolndian
complex had been separated from a “later” component on
the basis of 1963 professional typologies. Smith (1967)
illustrated unambiguous Hardin Barbed, Charleston
Corner-notched (some approaching Palmer in morphology)
McCorkle Stemmed, St. Albans Side-Notched ‘A’ (or Kirk)
and Thebes points as well as the types defined as late
Paleolndian. Like Squaw Rockshelter, these sites show no
Archaic occupational hiatus across northeast Ohio. They
reveal local assemblages which show stylistic and
technological continuities despite the frequent
morphological assignment of specific artifacts to
discontinuous distant prototypes.
Speculations
To what cultural processes may we attribute the
distinct Early Archaic lithic variability and the Middle
44
Brose
No. 44
Archaic stylistic homogenization at Squaw Rockshelter?
Answers must rest on assumptions concerning the number
and composition of occupants; the season or seasons during
which they used the shelter; and the economic activities
that drew them to this portion of Aurora Branch.
Despite evidence from Michigan to Maine for a
Clovis settlement system exploiting migrating caribou,
Payne (1987:2) admitted that no clear evidence for any
Paleolndian settlement pattern has yet emerged in
northwest Ohio. The same must be admitted for all of the
state, and for the entire Archaic as well. No stratigraphic
evidence recovered from either the Early or late Middle
Archaic levels permits a decision as to whether Squaw
Rockshelter was occupied at a single time or was
intermittently reoccupied. Although artifacts in both levels
were distributed through 10 to 30cm, in sediments of
similar texture microbiotic processes and soil plasticity can
vertically displace artifacts by 40cm or more (Hoffman
1986:167).
In eastern Pennsylvania, at Shawnee-Minisink
(McMillan 1986:264), and at Sheep Rock Shelter (Michels
1968; Michels and Smith 1967) the Early Archaic is taken
to represent transitory, limited purpose occupation by small
band segments at the limit of their large range. At Sheep
Rock the Middle Archaic occupation was seen as multi-
purpose reoccupations for longer segments of the year by
several families occupying a more demographically
circumscribed territory. The latter interpretation appears to
have been derived principally from analysis of regional
projectile point relationships, not from any spatial data
recovered at the site. So too, Michlovic (1976:14-15)
offered ethnographic analogs suggesting that the observed
blending of Archaic projectile point stylistic modes
reported in western Pennsylvania may have been the result
of ephemeral interaction by short term, limited work
parties of mixed ethnic or social composition. Anderson
and Hansen (1988) have proposed a social and economic
geography model for the Early Archaic of the southeast
coastal plain. Their model involves up to 10 bands of 50-
150 people occupying the territory of a single river valley,
congregating where resources are annually abundant to
exchange information and genes (cf. Brose 1979).
Matching the ethnoarchaeological data of Wiessner ( 1983),
Anderson and Hansen see assemblages of hafted lithic
bifaces displaying the greatest variability at the sites which
represent the aggregation of several small bands. Curated
tools and possibly the range of raw materials may also have
higher frequencies at such sites. Somewhat in opposition to
these conjectures, Behm (1985) has suggested an absence
of distinct social territories in the Early Archaic of the
Upper Mississippi Valley based upon analyses of spatial
patterning of stylistic variation in projectile points.
While there would seem little potential for Squaw
Rockshelter to have been inhabited by even one small
band, several similar (yet untested) shelters in the
environmentally diverse kame and kettle topography for
2km downstream along Aurora Branch could have been
occupied contemporaneously. It actually remains unclear
whether the Squaw Rockshelter ever housed a family.
There is every reason to believe that most, if not all, of the
stone tools recovered in either level could have been used
(and many, if not most, of the tools could have been
produced) by women (Gero 1988). Perhaps the increasing
use of locally available and unheated chert to produce more
expedient tools represents a shift to more female occupants
from Early to late Middle Archaic period use of the site.
If the demography of the Squaw Rockshelter
occupants is uncertain, the season in which occupation was
probable may be approximated. Deller (1976:5-6)
suggested that Ontario Paleolndian and Early Archaic sites
had southern exposures to protect from the north wind and
to provide light. The latter would only be true in the winter,
however, and early Amerindians must have lived
somewhere during other times of the year. Hall and Klipple
(1988) argue that most occupied rockshelters on the
Cumberland plateau had a southern exposure, suggesting
occupation primarily in the winter and early spring for
protection from northern storms. Size of protected area,
proximity to varied resource catchments and avoidance of
flooding seem the factors which conditioned choices
among shelters. Similar patterns have been observed in
southern (Bush 1987:121-129) and northern Kentucky (C.
Wesley Cowan, personal communication 1988), and to a
lesser extent in eastern Ohio (Olaf Prufer, personal
communication 1988), but most of those occupations
postdate the Archaic.
If the northern exposure of Squaw Rockshelter
argues against a winter occupation, the total absence of
faunal remains, bivalves or any edible floral remains save a
single burned fragment of walnut hull in Level I, may
argue against occupation in the summer or fall. Storck
(1982:23-25) suggests that fishing had more significance
for late Paleolndian site location than is usually considered,
but it is unlikely many fish could have been procured this
far upstream. In the late spring, with most resources in
short supply, subsistence efforts may have concentrated at
lakeshore marshes or upland fens and bogs where emergent
vegetation and tubers could be obtained with little effort.
While such resources were available nearby as early as the
occupation of Level II, they could never have been capable
of supporting social aggregates of any size, and the limited
area available within Squaw Rockshelter may have been
more than adequate for any population the catchment could
sustain.
Like most such ethnoarchaeological models, the
relevant environmental and cultural factors can be
demonstrated only for late prehistoric or for subsequent
periods. Their applicablity to any portion of the Archaic is
conjectural at best, and present data are far too thin to
support arguments concerning whatever cultural changes
1989
Squaw Rockshelter
45
Figure 14. Location of soil samples from the Squaw Rockshelter site.
may have occured over the first two-thirds of that period.
Nonetheless, I believe that the technical and stylistic data
to be derived from such Archaic lithic assemblages as were
present at Squaw Rockshelter can no longer be seen as
mere indicators of gross chronological position and general
inter-group relationships. For a vast stretch of time they
constitute the only evidence we are likely to have from
which to recover details of how critical functional and
economic vectors may be tempered by variable social
structure. To be sure, the quantity and the quality of the
data are less than we might wish. But our potential for
discovering more Archaic sites remains obscure, and our
potential for less archaic interpretations is clear. A
Appendix
Sample Lithology
During the excavation at Squaw Rock Shelter a series
of 1.50 liter soil samples were collected from the eastern
corner of Unit 3. A series of column-like samples were also
taken from the western edge of floors of Unit 5 (Figure 14).
The frequency of fragments of sandstone embedded within
the finer matrix of sands and silts precluded strict column
sampling (cf. Farrand 1985). Sampling was also biased
because irregular cobbles of sandstone roof fall, which far
exceeded 64cm in every dimension, could hardly be
included. Nor were the very fine silt and clay sized
particles separated by hydrometry. The remaining samples,
from pebble (-5 Phi) to fine silt (+7 Phi) were air dried,
quartered and mechanically shaken through a nested
column of Wentworth grade sieves. The relative frequency
of each grade class by weight is presented as granulometric
histograms (Figure 15) in which column A represents the
excavation Unit 3 sample while Column B represents the
sample from Excavation Unit 5. Small samples were also
studied for grain lithology and wear (cf. Brose 1970:27-9),
with the assistance of the late Dr. John Hall, Case Western
Reserve University Department of Geology.
Sample AO
A moderately well sorted deposit of yellowish brown
( 10YR4/2) to strong brown (7.5YR5/6) fine sandy silt loam
lying unconformably upon the eroded surface of the
Bedford Shale at a depth of 155cm below survey datum,
itself about 10cm higher than the surface in XU1. Included
within the coarse fraction of this sample were several small
rounded pebbles of black shale and several of quartzite.
Coarse sands included about 60% quartz, and about 40%
46
Brose
No. 44
Granulometric Histograms
Squaw Rockshelter
33CU34
Figure 15. Granulometric histogram of soil samples from the Squaw Rockshelter site.
1989
Squaw Rocks h e l t e r
47
feldspars along with a few carbonate grains. Virtually no
heavy minerals were identified in the sample. Hiram Till is
always characterized as a silty clayey till with very few
pebbles, predominantly of black shale, and rarely
containing cobbles or boulders. The grain size distribution
for local deposits is variously reported as sands: 15-20%;
silts: 45-48%; clays: 33-40% (Shepps 1953; Winslow and
White 1954; White 1982:25; White 1984:17; Szabo and
Ryan 1981: 242).
Samples A I , B1
The small coarse sand sample analyzed revealed
about 60% rounded grains, about 50% of which were
feldspars, although a few grains of chert and tourmaline
were present. The presence of the numerous large cobbles
has shifted the relative histograms (Figure 15) toward the
negative Phi units, but the matrix in which they are
embedded appears to represent detritus chemically
weathered from both the Bedford Shale and Berea
Sandstone. The larger fragments are presumably detached
by mechanical weathering processes such as freezing.
Samples A2, B2
Qualitative geochemical analyses indicate a rather
high organic content (5% loss on ignition) and a neutral
pH. The sedimentary sources for this deposit appear to
have been chemically weathered predominantly from the
Berea Sandstone. There were, within the medium to coarse
sand sample, only a few grains of chlorite, garnet and
zircon and most of the quartz grains were sub-angular (cf.
Pepper et al. 1954:91-95). The A2 and B2 samples, when
compared to the relative granulometric distribution of the
local Berea “channel fill,” show strong positive Phi
skewness (Krumbein and Pettijohn 1938:229-254),
especially laminae from within portions of Sample B2
along the rear shelter wall.
Samples A3, B3
Relatively similar predominantly dark yellowish
brown (10 YR 4/4 to 10YR 3/6) and light yellowish brown
(10YR 6/4) fine sands although small indistinct mottles of
dark reddish brown (2.5YR 3.5/5) and strong brown
(7.5YR 4/6) medium sands occur throughout the stratum.
Quartz grains within these mottled areas tend to show
considerably more rounding than do those in lighter
colored areas of the deposit. No geochemical differences
between mottled and unmottled areas was noted. Nor were
there any significant differences in the low frequencies of
heavy minerals (zircon, tourmaline, chlorite, and apatite),
or in the rare presence of well-rounded pebbles of dark
bluish gray (5B4/1) shale. Although a few coarse sand-
sized grains of dolomite or ankerite were present in Sample
A3, the entire deposit appears relatively acidic (pH 4.5).
Neither organic nor cultural material was encountered.
Samples A4, B4
Analyses suggest some degree of water transport of
sediments fallen from the roof and back wall with lower
energy levels (or less frequent episodes) in the front shelter
area. Those sedimentological processes resulting in the
formation of Stratum 4 appear to have persisted throughout
the period (or periods) of human occupation. It seems
likely that persistent seepage across the fore-sloping zone
was responsible for the removal of relatively finer
sediments from the rear of the shelter. The lack of
contiguous floor areas may be in part due to such a
phenomenon. While the exact depositional nature of the
stratum remains somewhat ambiguous, introduction of
coarser sediments by overbank flooding or by wash in from
areas upstream or upslope would seem unlikely. The rocks
of the Orangeville Member and Bedford Formation
exposed in those areas are both consistently finer than the
Bedford Sandstone or the Stratum 4 sediments, and they
display a far greater incidence of rounded grains and a very
different suite of heavy minerals (Prosser 1912:519-30;
Pepper et al. 1954:42-45,91-95).
Samples A5, B5 and A6, B6
Within Stratum 5 there was a consistent difference in
the grain size distribution skew from lower levels (Samples
A5, B5) to upper levels (Samples A6, B6). Lower levels
had a far greater frequency of silts, pebbles and small
cobbles than did the upper levels. Rear wall areas of this
deposit also exhibited a higher relative frequency of fine
silts than riverward areas of the same depth. Although no
bedding or lamellar structures could be identified within
Stratum 5, there were significant differences between the
front and the rear in the small pebbles included within the
samples analyzed. Numerous sub-rounded shale and
quartzite pebbles were encountered in Samples A5 and A6
while only a few well-rounded shale pebbles occurred in
Samples B5 or B6. Within the coarse sand fractions, A5
and A6 yielded dolomite, quartzite and zircon while B5
and B6 yielded tourmaline and chert only. Although all of
these minerals occur in the Berea Sandstone (Pepper et al.
1954:92-94) it seems clear from their spatial distributions
and from inclusive clasts, that the depositional processes
responsible for Stratum 5 were rather more complex than
those responsible for the grossly similar Stratum 3.
Historical Reconstruction
The following scenario for the evolution of the Squaw
Rockshelter is based upon regional geomorphology, analyses
of deposits in the shelter itself, and a tentative sequence for
collapse of large portions of the roof overhang5’.
1 . Following the 12,500 B.P. drop of Lake Erie and
the subsequent Chagrin River capture of upper
Aurora Branch, lower Aurora Branch began
retrenchment of late Pleistocene deposits in a
pre-Wisconsinan valley.
2. Bedrock floor of Aurora Branch was exposed by
fluvial erosion and Stratum I, resorted Hiram
48
Brose
No. 44
Till, deposited on Bedford Shale floor by 10,000
B.P.
3. Upper Stratum 1 minimally reworked by ground
water flow. Large block of Berea Sandstone east
of the Test Trench Excavation is detached from
the edge of the shelter roof.
4. Prehistoric occupation in Squaw Rockshelter
9200 B.P Stratum 2 deposits formed.
5. Reworking and sheet erosion of Stratum 2
surface. Fluvial removal of northern edge of
deposit.
6. Large blocks of Berea Sandstone detached from
overhang fall between Aurora Branch and
Excavation Units 1, 5, 3, and 2.
7. Accumulation of Stratum 3 primarily by in-place
weathering of overhang. Much roof fall, little
chemical weathering or water sorting.
8. Large block of Berea Sandstone from overhang
falls into position south of Excavation Unit 3,
west of Excavation Units 2 and 4.
9. Prehistoric occupation in shelter while
chemically and mechanically weathered Stratum
4 sediments accumulate through gravity and low
energy ground water. 5500 B.P
10. Deposition of lower Stratum 5 from roof fall and
reworked upstream deposits. Large fractured
block of Berea Sandstone roof falls between test
trench and Excavation Unit 1. Large block of
Berea Sandstone falls south of Excavation Unit 4.
11. Deposition of upper Stratum 5, sediments from
chemically weathered shelter wall.
12. Modern use of surface of Stratum 5, A.D. 1870-
1970.
Acknowledgements
Few sections of this report could have been written without
the advice and enthusiasm of the late Dr. John Hall, former
research associate at The Cleveland Museum of Natural
History, and professor of geology. Case Western Reserve
University. Beyond John’s advice, thanks are due to many
collegues who shared with me their knowledge of the
Archaic Period in the mid-continent. Among those
deserving signal recognition are James Brown, Jefferson
Chapman, William Fox, Ian Kenyon, William Lovis, Olaf
Prufer and Peter Storck. If I have not always accepted their
interpretations, I have never forgotten their generosity.
Notes
1 It is possible that these disturbed and redeposited sediments had
previously been noted as an archaeological site. A 1953 catalogue
card of the Western Reserve Historical Society (53.471) lists an
earlier donation by Mr. George E. Tow[N]er of the "Seventeen
Indian relics picked up near Squaw Rock at Chagrin Falls.”
Though shown as being in box 305 on shelf 15, those "relics”
could not be located with certainty nearly 25 years later. One
artifact box with no provenience designation, found on what may
have been shelf 15 within that storage room, contained two
complete and three broken bifacial blades and a single St. Albans
side-notched point (Variety A [Broyles 1971]). All were made of
Upper Mercer flint.
2 Coe’s excavations at the Hardaway site (basal level IV) yielded
only Hardaway points. Level III yielded Hardaway blades and
Hardaway side-notched. Palmer Corner-notched, Kirk Corner-
notched and stemmed, and even a few Stanly and Morrow
Mountain stemmed points. Level II contained everything from
Stanly through protohistoric triangular Caraway points (Coe
1974:56-83). At Doerschuk zone XI yielded Stanly stemmed
points with hafting morphology ranging nearly from Bifurcate
base to side-notched (Coe 1964:36). Zone X yielded Morrow
Mountain I and II stemmed points only, while Zone V, VI, and
VII each yielded a variety of stemmed and unstemmed lanceolate
points (Coe 1964:14-50). At the Gaston site each level from 24"
to 68" below surface showed mixed assemblages of unstemmed
lanceolate blades and stemmed and side-notched points (Coe
1964:84-91, 107-112).
3 Every sealed level at St. Albans site shows variation in the
projectile point/kmfe hafting: The earliest levels yielded both
Dalton/Hardaway and Palmer corner-notched. All Charleston
components (Levels 32-36 at 9800 + 500 B.P.) included
unstemmed lanceolate: (Palmer and Charleston) corner-notched;
and (Kessell) side-notched points and knives. All Kirk
components (Levels 14-20 dated between 8930±160 B.P. and
8800±32Q B.P.) included (Kirk and/or MacCorkle) corner-
notched; (Kirk) stemmed; and (St. Albans) side-notched points as
well as unstemmed lanceolate knives. St. Albans components
(levels 11 and 12, dated from 8830+700 B.P. to 8820±500 B.P.)
included Kirk and MacCorkle corner-notched; (St. Albans) side-
notched; and (Kirk or tentative MacCorkle) stemmed. LeCroy
components (Levels 6 and 8 dated at 8250±100 B.P.) included
(LeCroy Bifurcated based) corner-removed or corner-notched;
(St. Albans B) side-notched; and (Kanawha) stemmed projectile
points. And, finally, all of the Kanawha components (Level 4 and
6 dated to 8250+100 B.P. and 8160+100 B.P.) contained (LeCroy
Bifurcated base) corner-removed or corner-notched; and
(Kanawha) stemmed projectile points and unstemmed lanceolate
and side-notched knives and blades.
4 For example, the stratum X Kirk Complex at Howard Farm
included (Kirk) corner-notched; (category 6, 13/14, and 17)
stemmed; and (category 7) expanded stemmed/side-notched
point/knives (Chapman 1978:16-21-32) while at Icehouse Bottom
the 170m2 exposures of the equivalent “Upper Kirk" component
(Strata I-L) included (large and small Kirk) corner-notched;
Hardaway-Dalton; (St. Albans and Pseudo) side-notched; and a
variety of (LeCroy Bifurcated) stemmed point/knives (Chapman
1977:26-57). The same variability is seen in late Middle Archaic
components. Morrow Mountain levels VII and VIII at Howard
Farm yielded (Kirk, Morrow Mt. Indeterminate, and Categories 6
and 17) stemmed; (category 7) expanding stemmed/side-notched;
and (Kirk and Eva II) corner-notched projectile point/knives
(Chapman 1978:14-16, 21-30).
5 Just below Bridge Street in Chagrin Falls, the walls of the
Chagrin River gorge are composed of laminar-bedded Berea
Sandstone. The existing property surveys and historical records
from 1834 (Johnson 1879) indicate that the topography of this
area has remained relatively unchanged for over a century and a
half. The Berea outcrop, where unaffected by recent protective
backcutting, is a light buff to steel blue grey in color.
1989
Squaw Rockshelter
49
characteristic of what Bownocker (1915:114) called “recent
exposure”.
In the South Chagrin Metropark, the bas-relief Squaw-and-
Snake figure was carved into a fallen block of Berea Sandstone in
1884. The foreset beds of the block are now vertical, showing
discontinuous and moderate degrees and depths of banded
ferruginous staining on only a few of the higher relief areas. A
comparison with the Chagrin Falls gorge suggests that the Squaw
Rock block had been in place for over 150 years when the carving
was made. Some estimate of the relative sequence, if not (he
period, at which large blocks of roof fall took place has been
obtained by a comparison of the degree of iron staining seen in
their cross-bedding and, the staining seen the cross-bedding
exposed in still intact sections of the shelter wall.
References
Adovasio, James, and Ronald Carlisle. 1982. Meadowcroft :
Collected Papers on the Archaeology of Meadowcroft
Rockshelter and the Cross Creek Drainage. Pittsburgh:
Department of Anthropology, University of Pittsburgh.
Ahler, Stanley. 1971. Projectile Point Form and Function at
Rodgers Shelter, Missouri. Missouri Archaeological Society
Research Series 8.
Ahler, Stanley A., and Bruce McMillan. 1976. Material Culture at
Rodgers Shelter: a reflection of past human activities, pp. 163-
199 in Prehistoric man and his environments, edited by R.
Bruce McMillan. New York: Academic Press.
Anderson, Duane C. 1980. Stone Tool Assemblages from the
Cherokee Site. pp. 197-238 in The Cherokee Excavations:
Holocene Ecology and Human Adaptations in Northwestern
Iowa, edited by D. C. Anderson and H. A. Semken, Jr. New
York: Academic Press.
Anderson, David G., and Glen Flansen. 1988. Early Archaic
Settlement in the Southeastern United States: A Case Study
from the Savannah River Valley. American Antiquity 53(2):
262-286.
Anderson, David G., and Joseph Shuldenrein. 1985. Prehistoric
Human Ecology Along the Upper Savannah River:
Excavations at the Rucker's Bottom, Abbeville and Bullard
Site Groups. Russell Papers. Atlanta: Archaeological Services,
National Park Service.
Behm, Jeffery. 1985. Stylistic Analysis of Hardin Barbed Points:
Cultural Interaction Within the Early Archaic. Paper Presented
at the 47th Annual Meeting of the Midwest Archaeological
Conference.
Bense, Judith A. (editor). 1987. Final Report: The Midden Mound
Project. University of West Florida, Office of Cultural and
Archaeological Research Report of Investigations 6.
Pensacola.
Blank, John E. 1970. The Ohio Archaic: A Study in Culture
History. Unpublished Ph.D. dissertation. University of
Massachusetts, Amherst.
Bownocker, J. A. 1915. Building Stones of Ohio. Ohio
Department of Natural Resources, Geological Survey of
Ohio, Fourth Series, Bulletin 18. Columbus.
Brose, David S. 1970. The Archaeology of Summer Island:
Changing Settlement Systems in Northeastern Lake Michigan.
Museum of Anthropology, University of Michigan. Papers 41.
1972. The Northeast, pp. 11 6- 139 in Development of North
American Archaeology, edited by J.E. Fitting. New York:
Doubleday.
1975a. Functional Analysis of Stone Tools: A Cautionary Note
on the Role of Animal Fats. American Antiquity 40:86-94.
1975b. The Hospital and Academy Sites, Archaic Occupations
in Monroe Co., Michigan. Aboriginal Research Bulletin
5(3):24-38.
1975c. Preliminary Results of a Recconnaissance of the Area of
Lower Tinker's Creek Valley. Peninsula: Cuyahoga Valley
National Park Board.
1976a. Locational Analysis in the Prehistory of Northeast Ohio,
pp. 3-18 in Cultural Change and Continuity: Papers in
Honor of James B Griffin, edited by C. Cleland. New York:
Academic Press.
1976b. Botanical Archaeology at the Norman P Site: A Method
for the Identification of Recently Disturbed Stratigraphy at an
Early Alluvial Floodplain Occupation. Journal of Field
Archaeology 2(4):293-307.
1977. Archaeological Reconnaissance and Sub-surface
Investigations of the Proposed U.S. Steel Co. Greenfield Steel
Plant. Ashtabula Co. Ohio, and Erie Co., Pennsylvania .
Cambridge: Arthur D. Little, Inc.
1978a. Lithic Technology and Cultural Ecology in Northeastern
Ohio. pp. 87-115 in Lithic Technology and Prehistoric
Subsistence, edited by D. Davis. Vanderbilt University
Publications in Anthropology 20. Nashville.
1978b. Archaeological Investigations at the Burr ill Farm,
Lorain County, Ohio, Prehistoric and Historic Evidence. The
Cleveland Museum of Natural History. Report to the Lorain
County Metropolitan Park District, Lorain, Ohio.
1979. A Speculative Model of Prehistoric Exchange in the
Eastern United States, pp.3-8 in Hopewell Archaeology: The
Chillicothe Conference, edited by D. Brose and N. Greber.
Kent: Kent State University Press.
Brose, David S., and Stephanie J. Belovich. 1984. Archaeological
Investigations at the Greenwood Village Site, 33Su92
Cuyahoga Valley N.R.A. Museum of Natural History,
Archaeological Research Report 47( 1 ): 1-194. Cleveland.
Brose, David S., S. Belovich, M. Brooslin. R. Bums, J. Hall, H.
Haller, C. Pierce, and C. Ubbeohde. 1981. Prehistoric and
Historic Archaeological Investigations of the Cuyahoga Valley
National Recreation Area, Ohio. Museum of Natural History,
Archaeological Research Report 30:1-599. Cleveland.
Brose, David S., and Patricia Essenpreis. 1973. A Report on the
Preliminary Archaeological Survey of Monroe Co., Michigan.
The Michigan Archaeologist 19( 1 &2): 1-182.
Brose, David S., and Alfred M. Lee. 1980. A Reinterpretation of
the Late Paleoindian to Early Middle Archaic Period in
Northern Ohio. Presented at the Ontario Archaeological
Society Symposium on the Prehistory of Lake Erie. London,
Ontario.
Brose, David S., and John F. Scarry. 1977. Boston Ledges: Spatial
Analyses of a Late Woodland Rockshelter in Summit Co.,
Ohio. Mid-Continental Journal of Archaeology 1(2): 179-228.
Brown, James A. 1983. Summary, pp. 5-11 in Archaic Hunters
and Gatherers in the American Midwest, edited by James H.
Phillips and James A. Brown. New York: Academic Press.
Brown, James A. 1986. Food for Thought: Where has Subsistence
Analyses Gotten Us? pp. 315-330 in Foraging, Collecting
and Harvesting: Archaic Period Subsistence and Settlement
in the Eastern Woodlands, edited by Sarah W. Neusius. Center
For Archaeological Investigations. Occ. Paper 6. Carbondale:
SIU.
50
Brose
No. 44
Brown, James, and Charles Cleland. 1969. The Late Glacial-Early
Post-Glacial Faunal Resources in Midwest Biomes Newly
Opened to Human Adaptation. University of Illinois, College
of Agriculture, Special Publications 14.
Brown. James A. and Robert K. Vierra 1983. What Happened to
the Middle Archaic: Introduction to an Ecological Approach to
Roster Site Archaeology, pp. 165-322 in Archaic Hunters and
Gatherers in the American Midwest, edited by James H.
Phillips and James A. Brown. New York: Academic Press.
Broyles, B.J. 1971. Second Preliminary Report: the St. Albans
Site, Kanawha County, West Virginia. Report of
Archaeological Investigations No. 3. West Virginia Geological
and Economic Survey, Morgantown.
Bush, David R. 1987. Prehistoric Lifeways in the Cumberland
Plateau. Unpublished Ph.D. Dissertation in Anthropology,
Case Western Reserve University, Cleveland.
Calkin, Parker E. 1970. Strandlines of Glacial Great Lakes in
Northwestern New York. Ohio Journal of Science 70(2)78-96.
Calkin, P. E. and K. E. Miller. 1977. Late Quaternary
Environment and Man in Western New York. Annals of the
New York Academy of Science 288:296-315.
Chapman, Carl H. 1975. The Archaeology of Missouri, Volume 1 .
Columbia: University of Missouri Press.
Chapman, Jefferson. 1975. The Rose Island Site and Bifurcate
Point Tradition. Report of Investigations 14, Dept, of
Anthropology, University of Tennessee, Knoxville.
1976. The Archaic Period in the Lower Little Tennessee River
Valley: The Radiocarbon Dates. Tennessee Anthropologist
( 1 ): 1 - 1 2.
1977. Archaic Period Research in the Lower Little Tennessee
River Valley. Report of Investigations No. 18, Dept, of
Anthropology, University of Tennessee, Knoxville, Tennessee.
1980. The Early and Middle Archaic Periods: A Perspective
from Eastern Tennessee, pp. 123- 132 in Proceedings of the
Conference on Northeastern Archaeology, edited by J. A.
Moore. University of Massachusetts, Department of
Anthropology, Research Report 1 9. Amherst.
Charles, Douglas K., and Jane E. Buikstra. 1983. Archaic
Mortuary Sites in the Central Mississippi Drainage:
Distribution, Structure, and Behavioral Implications, pp. 117-
146 in Archaic Hunters and Gatherers in the American
Midwest, edited by James H. Phillips and James A. Brown.
New York: Academic Press.
Coakley, J. R, and C. F. M. Lewis. 1985. Postglacial Lake Levels
in the Erie Basin, pp. 195-212 in Quaternary Evolution of the
Great Lakes, edited by P. F. Karrow and P. E. Calkin.
Geological Association of Canada Special Paper 30.
Coe, Joffre C. 1964. The Formative Cultures of the Carolina
Piedmont. Transactions of the American Philosophical
Society. New Series 54(5).
Collins, Michael B., Editor. 1979. Excavations at Four Archaic
Sites in the Lower Ohio Valley, Jefferson County, Kentucky.
Occasional Papers in Anthropology Number 1 ( I & 1 1 ) .
Lexington: Department of Anthropology, University of
Kentucky.
Coogan, Alan H., Loren E. Babcock, Joseph T. Hannibal, Doyle
W. Martin, Karen S. Taylor, and David C. Wehn, 1986, Late
Devonian and Early Mississippian Strata at Stebbins Gulch [,]
Geauga County, and Quarry Rock, Cuyahoga County, Ohio.
Field Trip Guide Book Field Trip Number One. Geological
Society of America, Northcentral Section Meeting, edited by
Alan Coogan. Kent: Kent State University Department of
Geology, and Geological Society of America.
Cook, Thomas G. 1976. Koster: An Artifact Analysis of Two
Archaic Phases in Westcentral Illinois. Prehistoric Records 1 ,
Koster Research Report 3. Evanston: Northwestern University.
1980. Artifactual [sic] Studies, pp. 222-455 in Archaeological
Salvage Excavations At Patoka Lake Indiana, edited by
Cheryl A. Munson. Bloomington: Glenn Black Laboratory of
Archaeology, Indiana University.
Cunningham, Roger. 1973. Paleo Hunters Along the Ohio River.
Archaeology of Eastern North America 1(1 ): 120-126.
Davis, Margaret Bryan. 1983. Holocene Vegetation History of the
Eastern Unites States, pp. 166-181 in Late Quaternary
Environments of the United States: Volume 2: The Holocene.
edited by H. E. Wright Jr. Minneapolis: The University of
Minnesota Press.
Deller, D. Brian. 1976. Paleolndian Locations on Late Pleistocene
Shorelines, Middlesex County, Ontario. Ontario Archaeology
26:3-19.
Dincauze, Dena F. 1976. The Neville Site: 8000 Years at
Amoskeag Manchester, New Hampshire. Cambridge:
Peabody Museum Monographs 4.
Donahue, Jack. 1976. Geological Framework of Meadowcroft
Rockshelter. Paper prepared for Meadowcroft Rockshelter
Symposium, 41st Annual Meeting of the Society for American
Archaeology. St. Louis.
Doran, Glen H., David N. Dickel, William E. Ballinger Jr., O.
Frank Agee, Philip J. Laipis, and William W. Hauswirth. 1986.
Anatomical, Cellular and Molecular Analysis of 8000 year old
Human Brain Tissue from the Windover Archeological Site,
Nature 323(6091 ):803-806.
Dragoo, Don W. 1959. Archaic Hunters of the Upper Ohio Valley.
Carnegie Museum Annals, 35:139-246.
1976 Some Aspects of Eastern North American Prehistory: A
Review 1975. American Antiquity 4( 1 ):3-27.
Ellis, Chris. 1984. Paleolndian Lithic Technological Structure and
Organization in the Lower Great Lakes: A First
Approximation. Unpublished Ph.D. Thesis, Department of
Archaeology, Simon Fraser University, Burnaby, B.C.
1987. Stanley/Neville Points. KEWA: Newsletter of the London
Chapter, Ontario Archaeological Society 87-9:21.
Ellis, Chris J., and D. Brian Deller. 1982. Hi-Lo Materials for
Southwestern Ontario. Ontario Archaeology 38:3-22.
1986. Post-Glacial Lake Nippising Waterworn Assemblages
from the Southeastern Huron Basin Area. Ontario
Archaeology 45:27-38.
1988. Some Distinctive Paleolndian tool types from the Lower
Great Lakes. Mid-Continental Journal of Archaeology 13(2):
111-158.
Evans, June 1988. Paleolndian to Early Archaic Transition at the
Shawnee-Minisink Site. pp. 221-259 in Shawnee-Minisink: A
Stratified Paleolndian-Archaic Site in the Upper Delaware
Valley of Pennsylvania, edited by C. W. McNett Jr. New York:
Academic Press.
Farrand, William R. 1985. Rockshelter and Cave Sediments.
pp.21-40 in Archaeological Sediments in Context, edited by
Julie K. Stein and William R. Farrand, Orono: Center for the
Study of Early Man.
Feldmann, Rodney, Alan Coogan, and Richard Heimlich. 1977.
Southern Great Lakes: Field Guide. K/H Geology Field Guide
Series, edited by J. Harbaugh. Dubuque: Kendale Hunt.
Fitting, James E. 1963a. The Welti Site: A Multi-component site
in southeastern Michigan. The Michigan Archaeologist
9(3 ):204-2 14.
1989
Squaw Rockshelter
51
1963b. The Hi-Lo Site: A Late Paleolndian Site in Western
Michigan. Wisconsin Archaeologist 44( 2 ): 87-96.
1964. Bifurcate-stemmed Projectile Points in the Eastern United
States. American Antiquity , 30( 1 ): 92-94.
1967. The camp of the careful Indianian Upper Great Lakes
chipping station. Papers of the Michigan Academy of Science ,
Arts and Letters 52:237-242.
1968. Environmental Potential and the Post Glacial
Readaptation in Eastern North America. American Antiquity
33(4)441-445.
1970. The Archaeology of Michigan. Natural History Press.
New York.
Ford, John P. 1987. Glacial and Surficial Geology of Cuyahoga
County, Ohio. State of Ohio, Department of Natural
Resources, Division of Geological Survey, Report of
Investigations 1 34.
Fowke, Gerard. 1902. Archaeological History of Ohio. Columbus:
Ohio Archaeological and Historical Society.
Fox, William A. 1980. The Kirk Cluster in Southwestern Ontario.
Presented at the Ontario Archaeological Society Annual
Symposium on the Prehistory of Lake Erie. London, Ontario.
Frison, George. 1968. A Functional Analysis of Certain Chipped
Stone Tools. American Antiquity 33: 149-1 55.
Funk, Robert 1978. Post-Pleistocene adaptations, pp. 16-27 in
Handbook of North American Indians. Volume 15. Northeast.
edited by Bruce G. Trigger. Washington: Smithsonian
Institution.
George, Richard L. and Harry Bassinger. 1975. The Wadding
Rockshelter, 36AR2 1 . Pennsylvania Archaeologist 45(4): 1-21.
George, Richard L., and Christine E. Davis. 1986. A Dated
Brewerton Component in Armstrong County, Pennsylvania.
Pennsylvania Archaeologist 56( 1-2): 1 2-20.
Gero. Joan. 1988. Women's Roles in Stone Tool Production.
Paper presented at the Women and Production in Prehistory
Conference. Columbia, S.C.
Geisthweit, Barbara. 1970. Ohio Archaic. Unpublished Masters
thesis in Anthropology. The Ohio State University, Columbus.
Goldthwait, Richard, G.W. White, and J.L. Forsyth. 1961. Glacial
Map of Ohio. Miscellaneous Geological Investigations. Map
1-316. U.S. Geological Survey, Washington, D.C.
Gordon, Robert B. 1966. Map — Natural Vegetation of Ohio at
the Time of the Earliest Land Surveys. Ohio Biological Survey,
Columbus.
1969. Vegetation of Ohio at the Time of Original Survey. Ohio
Biological Survey, Columbus.
Gramly, Richard M. 1980. Raw Material Source Areas and
“Curated" Tool Assemblages. American Antiquity > 45:823-833.
Griffin, John W. 1974. Investigations in Russell Cave. U.S.
Department of the Interior, National Park Service Publications
in Archaeology 13. Washington, D.C.
Hall, Charles L., and Walter E. Klipple. 1988. A Polythetic-
Satisficer Approach to Prehistoric Natural Shelter Selection in
Middle Tennessee. Mid-Continental Journal of Archaeology
13(2)159-186.
Hassen, Harold, and Kenneth Farnsworth. 1987. The Bullseye
Site, ll-GE-127: A Floodplain Archaeology Mortuary Site in
the Co. Illinois River Valley. Illinois State Museum Report 42.
Hoffman, Jack L. 1986. Vertical Movement of Artifacts in
Alluvial and Stratified Deposits. Current Anthropology
27(2): 163-171.
Holloway, Vaughn M., and Richard G. Bryant. 1985. Late
Quaternary Pollen Records and Vegetational History of the
Great Lakes Region: United States and Canada, pp. 205-245 in
Pollen Records of Late Quaternary North American
Sediments, edited by R. G. Bryant and V. M. Holloway.
Dallas: The American Association of Stratigraphic
Palynologists Foundation.
Holmes, William H. 1919. Handbook of Aboriginal American
Antiquities: Part I. Introductory. The Lithic Industries.
Smithsonian Institution, Bureau of American Ethnology,
Bulletin 60.
Jeffries, Richard W., and B. Mark Lynch. 1983. Dimensions of
Middle Archaic Cultural Adaptation at the Black Earth Site,
Saline County, Illinois, pp. 299-322 in Archaic Hunters and
Gatherers in the American Midwest, edited by James H.
Phillips and James A. Brown. New York: Academic Press.
Justice, Noel D. 1987. Stone Age Spear and Arrow Points of the
Midcontinental and Eastern United States: A Modern Survey
and Reference. Bloomington: Indiana University Press.
Keeley, Laurence H. 1980. Experimental Determination of Stone
Tool Uses, A Microwear Analysis. Chicago: The University of
Chicago Press.
Kenyon, Ian. 1980. The Satchell Complex in Ontario: A View
from the AuSauble Valley. Ontario Archaeology 34: 28-41.
Krinsley, D. H. and J. Donahue. 1976. Application of Sand Grain
Sufacetextural Analysis to Archaeology. Paper prepared for
Meadowcroft Rockshelter: Interim Statement Symposium.
41st Annual Meeting of the Society for American
Archaeology. St. Louis.
LaRoque, Auriel. 1967. Pleistocene Mollusca of Ohio. State of
Ohio, Department of Natural Resources, Division of
Geological Survey Bulletin 62(2): 1 1 3-356.
Lewis, C.F.M. 1970. Late Quaternary History of Lake Levels in
the Huron and Erie Basins. Proceedings of the 12th
Conference on Great Lakes Research , 1969:250-270. Ann
Arbor: International Association of Great Lakes Research.
Lewis, T. L. 1988. Late Devonian and Early Mississippian Distal
Basin-Margin Sedimentation of Northern Ohio. The Ohio
Journal of Science 88( 1 ):23-39.
Lewis R. Barry. 1983. Archaic Adaptations to the Illinois Prairie:
The Salt Creek Region, pp. 99-116 in Archaic Hunters and
Gatherers in the American Midwest, edited by James H.
Phillips and James A. Brown. New York: Academic Press.
Logan, Wilfred D. 1952. Graham Cave: An Archaic Site in
Montgomery County, Missouri. Missouri Archaeological
Society Memoir 2.
Lovis, William A., and James A. Robertson. 1985. Rethinking the
Archaic Chronology of the Saginaw Valley Michigan. Paper
presented at the 12th Annual Symposium of the Ontario
Archaeological Society, London.
Luchterhand, Kubet. 1970. Early Archaic Projectile Points and
Hunting Patterns in the Lower Illinois Valley. Illinois State
Museum, Report of Investigations 19.
Marshall, Sydne. 1985. Paleolndian Artifact Form and Function
at Shawnee-Minisink. pp. 165-209 in Shawnee-Minisink: A
Stratified Paleolndian- Archaic Site in the Upper Delaware
Valley of Pennsylvania, edited by C. W. McNett Jr. New York:
Academic Press.
Mason. Ronald J. 1963. Two Late Paleolndian Complexes in
Wisconsin. Wisconsin Archaeologist 44(4): 199-21 1 .
52
Brose
No. 44
1958. Late Pleistocene Geochronology and the Paleo-Indian
Penetration into the Lower Michigan Peninsula. University of
Michigan, Museum of Anthropology, Anthropological Papers
No. 11.
1962. The Paleolndian Tradition in Eastern North America.
Current Anthropology 3(3): 227-78.
1981. Great Lakes Archaeology. New York: Academic Press.
Mayer-Oakes, William J. 1955a. Prehistory of the Upper Ohio
Valley: An Introductory Archaeological Study. Annals of
Carnegie Museum , 34.
1955b. The Globe Hill Shell Heap (Site 46Hk34-l), Hancock
County, West Virginia. West Virginia Archaeological Society,
Publications Series 3.
McKenzie, Douglas H. 1967. The Archaic of the Lower Scioto
Valley, Ohio. Pennsylvania Archaeologist 37( 1 -2): 33-5 1 .
McMillan. Barbara. 1985. A Technological Analysis of the Early
Archaic, pp.26 1-320 in Shawnee-Minisink: A Stratified
P aleolndian- Archaic Site in the Upper Delaware Valley of
Pennsylvania, edited by C. W. McNett Jr. New York:
Academic Press.
McNett, Charles W. Jr. 1985. Artifact Morphology and
Chronology at the Shawnee-Minisink Site. pp. 83-122 in
Shawnee-Minisink: A Stratified Paleolndian- Archaic Site in
the Upper Delaware Valley of Pennsylvania, edited by C. W.
McNett Jr. New York: Academic Press.
Michels, Joseph W. 1968. Settlement Pattern and Demography at
Sheep Rock Shelter: Their Role in Culture Contact.
Southwest Journal of Anthropology 24:66-82.
Michels, Joseph W., and Ira F. Smith (editors). 1967.
Archaeological Investigations of Sheep Rock Shelter (2 vols.).
Department of Anthropology. Pennsylvania State University.
University Park.
Micholovic, Michael G. 1976. Social Interaction and Point Types
m the Eastern U.S. Pennsylvania Archaeologist 46( 1-2): 13- 16.
Miller, Barry B. 1983. Late Quaternary Fluvial History of the
Lower Cuyahoga River in Northeastern Ohio, Journal of
Great Lakes Research 9 ( 1 ):97- 1 05.
Montet-White, Anita. 1973. Significance of Variability in Archaic
Point Assemblages. Plains Anthropologist 19(3): 14-24.
Munson, Cheryl A., and Thomas G. Cook. 1980. The French Lick
Phase: A Dimensional Description, pp. 721-741 in
Archaeological Salvage Excavations At Patoka Lake Indiana.
edited by Cheryl A. Munson. Bloomington: Glenn Black
Laboratory of Archaeology, Indiana University.
Murphy, James, and John E. Blank. 1970. The Plum Run
Quarries, The Ohio Archaeologist 20: 198-2 1 7.
Newberry, J. S. 1873. Report on the Geology of Cuyahoga
County. Report on the Geological Survey of Ohio 1 ( 1 ): 1 70-
200. Columbus: Nevins and Myers.
O'Brien, Michael J., and Robert E. Warren. 1983. An Archaic
Projectile Point Sequence from the Southern Prairie Peninsula:
The Pigeon Roost Creek Site. pp. 71-98 in Archaic Hunters
and Gatherers in the American Midwest, edited by James H.
Phillips and James A. Brown. New York: Academic Press.
Ogden, J.G., III. 1966. Forest history of Ohio. I. Radiocarbon
dates and pollen stratigraphy of Silver Lake, Logan County,
Ohio. Ohio Journal of Science 66(4 ):387-40 1 .
Parkins, W. 1977. Onondaga Chert: Geological and Palynological
Studies as Applied to Archaeology. Unpublished MSc thesis.
Department of Geological Sciences, Brock University, St.
Catharines, Ontario.
Payne, James H. 1982. The Western Basin Paleolndian and Early
Archaic Sequences. Unpublished honors thesis in
Anthropology. University of Toledo.
1987. Some Notes on Paleolndian Cultures in Northwestern
Ohio. Paper presented at the December Meeting of the Ohio
Archaeological Council, Cincinnati.
Pepper. James F., Wallace DeWitt, Jr., and David F. Demarest.
1954. Geology of the Bedford Shale and Berea Sandstone in
the Appalachian Basin. U.S. Geological Survey Professional
Paper 259:1-1 14. U.S. G.P.O. Washington, D.C.
Potter, Loren D. 1946. Post-Glacial Forest Sequence of North-
Central Ohio. Unpublished M.A. thesis in Botany, Oberlin
College, Oberlin.
Prosser, Charles S. 1912. The Devonian and Mississippian
Formations of Northeastern Ohio. Ohio Geological Survey.
Fourth Series, Bulletin 15:323-534.
Protsch, Reiner R. R. 1978. Catalog of Fossil Hominids of North
America. New York: Gustav Fischer.
Prufer, Olaf H. 1961. Survey of Ohio Fluted Points (Cleveland,
The Cleveland Museum of Natural History), No. 4; No. 5; No.
6.
1963. The McConnell Site: A Late Paleolndian Workshop in
Coshocton County, Ohio. Scientific Publications of The
Cleveland Museum of Natural History , (New Series) 2(1).
Cleveland.
Prufer, Olaf H., and Raymond S. Baby. 1963. Paleolndians of
Ohio. Columbus: Ohio Historical Society.
Prufer, Olaf H., and Dana A. Long. 1986. The Archaic of
Northeastern Ohio. Kent State Research Papers in
Archaeology 6:1-89. Kent: Kent State University Press.
Prufer, Olaf H., and Charles Sofsky. 1965. The McKibben Site
(33TR-57), Trumbull Co., Ohio: A Contribution to the Late
Paleolndian and Archaic Phases of Ohio. Michigan
Archaeologist 11(1 ):9-40.
Quimby, George I. 1960. Indian Life in the Upper Great Lakes.
Chicago: University of Chicago Press.
Redder, Albert J. 1985. Horn Shelter Number 2: The South End:
A Preliminary Report. Centred Texas Archaeologist 10:37-65.
Rippeteau, Bruce E. 1973. The Principles and Theory of
Radiocarbon Chronology and a Radiocarbon Chronology for
Lithic Variation in the Northeastern U.S. ...3000 to 500 B.C.
Unpublished Ph.D. dissertation in Anthropology. Case
Western Reserve University, Cleveland.
Ritchie, William A. 1965. The Archaic in New York. New York
State Archaeological Association Bulletin 52:2-12.
1971. A Typology and Nomenclature for New York Projectile
Points. Bulletin Number 384. New York State Museum and
Science Service , Albany.
Ritzenthaler, Robert. 1975. A Guide to Wisconsin Indian
Projectile Point Types. Popular Science Series II . Milwaukee
Public Museum.
Rule, Pamala, and June Evans. 1985. The Relationship of
Morphological Variations to Halting Techniques among
Paleolndian Endscrapers at the Shawnee-Minisink Site.
pp.211-220 in Shawnee-Minisink: A Stratified Paleolndian-
Archaic Site in the Upper Delaware Valley of Pennsylvania.
edited by C. W. McNett Jr. New York: Academic Press.
Sears, Paul. 1931. Pollen Analysis of Mud Lake Bog in Ohio.
Ecology 1 2 (4):75 1 -76 1 .
Seeman, Mark F., and Olaf H. Prufer. 1982. An Updated
Distribution of Ohio Fluted Points. Mid-Continental Journal
of A n haeology 7(2): 1 55-1 67.
1989
Squaw Rockshelter
53
Semenov, S. A. 1964. Prehistoric Technology. Translated by M.
W. Thompson. London: Cory, Adams and MacKay.
Semken, Holmes A. Jr. 1983. Holocene Mammalian
Biogeography and Climate Change in the Eastern and Central
United States, pp. 182-207 in Late Quaternary Environments
of the United States: Volume 2; The Holocene, edited by H. E.
Wright Jr. Minneapolis: The University of Minnesota Press.
Shane, Linda. 1975. Palynology and Radiocarbon Chronology of
Battaglia Bog, Portage County, Ohio. Ohio Journal of
Science 75 (2):96- 1 02.
Shepps, Victor C. 1953. Correlation of the Tills of Northeastern
Ohio by Size Analysis. Journal of Sedimentary Petrography
23:34-48.
Shetrone, Henry C. 1936. The Folsom Phenomena as Seen from
Ohio. Ohio State Archaeological and Historical Quarterly
55:240-256.
Smith, Arthur G. 1967. The Sawmill Site, Erie County, Ohio.
Ohio Archaeologist 10: 84-97.
Snow, Dean. 1977. Archaeology of the Lake George Region, pp.
431-438 in Amerindians and Their Paleoenvironments in N.E.
North America, edited by Walter S. Newman and Bert Salwen.
Annals of the New York Academy of Science. 288.
Stafford, C. Russell. 1985. The Campbell Hollow Archaic
Occupations: A Study of Intrasite Spatial Structure in the
Lower Illinois Valley. Kampsville Archaeological Center.
Research Series 4. Kampsville: Center for American
Archaeology.
Stahle, David W., and James E. Dunn. 1984. An Experimental
Analysis of the Size Distribution of Waste Flakes from Biface
Reduction. Arkansas Archaeological Survey, Technical Paper 2.
Stewart, Andrew. 1984. The Zander Site: Paleolndian occupation
of the southern Holland marsh region of Ontario. Ontario
Archaeology 41 :45-79.
Storck. Peter L. 1972. An unusual late Paleolndian projectile
point from Grey County, southern Ontario. Ontario
Archaeology 18:37-45.
1978. The Coates Creek Site: A possible late Paleo-Indian-Early
Archaic site in Simcoe County, Ontario. Ontario Archaeology
30:25-46.
1979. A Report on the Banting and Hussey Sites: Two
Paleolndian campsites in Simcoe County, southern Ontario.
Archaeological Survey of Canada Survey Paper 93 Ottawa:
National Museums of Canada.
1982. Palaeo-Indian Settlement Patterns Associated with the
Strandline of Glacial Lake Algonquin in Southcentral Ontario.
Canadian Journal of Archaeology 6: 1-31.
1984. Research into the Paleolndian occupations of Ontario: A
review. Ontario Archaeology 41:3-28.
Styles, Bonnie W., Steven R. Ahler, and Melvin L. Fowler. 1983.
Modoc rock Shelter Revisited, pp. 261-298 in Archaic
Hunters and Gatherers in the American Midwest, edited by
James H. Phillips and James A. Brown. New York: Academic
Press.
Szabo, John P, and Dale E. Ryan. 1981. Quaternary Stratigraphy
of the Lower Mud Brook Basin, Northhampton Township,
Summit County, Ohio. The Ohio J ow ned of Science 81:239-
246.
Tippitt, V. Ann, and William H. Marquardt. 1984. The Gregg
Shoals and Clyde Gulley Sites: Archaeological and
Geological Investigations at Two Piedmont Sites on the
Savannah River. Russell Papers, Archaeological Services,
National Park Service. Atlanta.
Tringham, R., G. Cooper, G. Odell, B. Voytek, and A. Whitman.
1974. Experimentation in the Formation of Edge Damage: An
Approach to Lithic Analysis. Journal of Field Archaeology
1:171-196.
Tuck, James. A. 1974. Early Archaic Horizons in Eastern North
America. Archaeology of Eastern North America 2( 1 ):72-80.
Vance, Elizabeth D. 1987. Microdebitage and Archaeological
Activity Analysis. Archaeology 40 (4):58-60.
Webb, T. Ill, E. J. Cushing, and H. E. Wright, Jr. 1983. Holocene
Changes in the Vegetation of the Midwest, pp. 142-165 in
Late-quaternary Environments of the United States: Volume 2;
The Holocene, edited by H. E. Wright, Jr. Minneapolis: The
University of Minnesota Press.
White, George, and Stanley Totten. 1982. Glacial Geology of
Northeastern Ohio. State of Ohio, Department of Natural
Resources, Division of Geological Survey Bulletin 68.
1984. Glacial Geology of Summit County, Ohio. State of Ohio,
Department of Natural Resources, Division of Geological
Survey, Report of Investigations 123.
White, Nancy M., Chun Ho Lee, and Judith A. Bense. 1983.
Archaeological Investigations in the Upper Tombigbee Valley,
Mississippi: Phase II. University of West Florida, Office of
Cultural and Archaeology Research Report of Investigations
4. Pensacola.
Wiant, Michael D., Edwin R. Hajic, and Thomas R. Styles. 1983.
Napoleon Hollow and Koster Site Stratigraphy: Implications
for Holocene Landscape Evolution and Studies of Archaic
Period Settlement Patterns in the Lower Illinois River Valley,
pp. 147-164 in Archaic Hunters and Gatherers in the
American Midwest, edited by James H. Phillips and James A.
Brown. New York: Academic Press.
Wiessner, Polly. 1983. Style and Social Information in Kalahari
San Projectile Points. American Antiquity 48:253-276.
Williams, Arthur B. 1940. The Native Forests of Cuyahoga
County, Ohio. Scientific Publications of The Cleveland
Museum of Natural History 2.
Winslow, John, George W. White, and Earl E. Webber. 1953. The
Water Resources of Cuyahoga County, Ohio. Ohio
Department of Natural Resources, Division of Water Bulletin
26:1-123.
Witline, Arthur H. 1970. A Study of a Late Pleistocene lake in the
Cuyahoga River Valley, Summit Co., Ohio. Unpublished M.S.
thesis in Geology, Kent State University, Kent.
Wright, J. V. 1978. The Implications of Probable Early and
Middle Archaic Projectile Points from Southern Ontario.
Canadian Journal of Archaeology 2:59-77.
KIRTLANDIA
The Cleveland Museum of Natural History
March 1989 hi umber 44:55-58
Skeletal Remains from
Squaw Rockshelter
Fred Prior
Research Engineering Division
Information Systems, Siemens International
Wheaton, Illinois 60191
Abstract
Skeletal remains recovered from the Early Archaic level
of the Squaw Rockshelter were identified as belonging to
a young female Amerindian. Analyses of the limited frag-
ments and dentition suggest a balanced diet of moderate
coarseness. No pathologies or trauma were noted. Teeth
from at least two other individuals were present.
Introduction
This report on the skeletal material from Squaw
Rockshelter includes an inventory and a brief anatomical
description of pathologies and anomalies where evident. It
also provides estimates of the sex and age at death of the
single most complete specimen. The materials are curated in
the Physical Anthropology collections of the Cleveland
Museum of Natural History as CMNH 8080. As noted by
Brose (this volume), the circumstances of recovery suggest
surface abandonment of the body in the rear of the shelter,
with natural internment occuring only after some period of
loss due to erosion. With the exception of a relatively com-
plete mandible and five small cranial fragments all remains
are post-cranial. Where it can be determined, the majority of
this material is from the right side of a single individual.
Inventory
Post-cranial Remains
A right proximal humerus with the head well
preserved was present. The shaft is broken slightly inferior
to the deltoid tuberosity. The distal portion was not
recovered. All epiphyses of the head are fused, and there is
no observable pathology.
Diameter of head:
vertical — 38.4mm
horizontal — 36.0mm
A complete left clavicle with slight damage to the
acromion is present. Both epiphyses are fused.
Maximum length — 138.4mm
Circumference — 32.4mm
Robustness index — 23.4
Fragments of one cervical and three thoracic
vertebrae are present. All are marked by a lack of
robustness, and exhibit no gross abnormalities. The
cervical (possibly the fourth) vertebrae is complete with
the exception of the anterior and posterior tubercles on
both sides and the lateral borders of both transverse
foramina. The three thoracic vertebrae are extremely
56
Prior
No. 44
fragmentary. The most complete (possibly the second or
third) consists of the posterior half of the centrum and the
left transverse process with the spinous and right transverse
processes missing. The remaining two thoracic vertebrae
are represented by a spinous process and limina with both
superior articular processes, and a small body fragment
with left superior articular process respectively.
A fragment of the right scapula containing the intact
glenoid fossa is present. The scapula has been fractured
(post mortem) along a line from a point in the scapular
notch just medial to the base of the coracoid to a point on
the axillary boundary about 5cm inferior to infraglenoid.
The bases of both the coracoid process and the spine are
present, but neither the coracoid nor the acromion processes
remain. There is no observable lipping of the glenoid
margin. There is also a second, smaller fragment of the right
scapula composed of a section of the spine medial to that of
the first fragment. The two sections do not articulate.
Length of glenoid — 33.6mm
Width of glenoid — 24.3mm
Width of axial border (3cm below infraglenoid) — 6.5
A badly preserved right distal extremity of the femur
with no articulated shaft, was recovered. The patellar and
tibial surfaces are largely intact as is the lateral condyle.
Only a small section of the medial condyle remains.
The right tibia is represented by two fragments: a
92mm section of shaft of the proximal tibia beginning
immediately distal to the popliteal line, and a section of the
proximal extremity composed of the articular surface of the
lateral condyle and the fibular articular facet.
A small section (44mm in length) of the right distal
ulna (shaft) is present. A complete right first metacarpal
and a right second with the proximal end damaged. Four
phalanges are present. These include a proximal, middle
and terminal from the right hand and a terminal from the
fifth digit of the foot.
Five small rib fragments lacking any distinguishing
landmarks are the only remains of the rib cage.
Miscellaneous fragments include eighteen small,
indeterminant, irregular pieces of compact tissue and four
irregular chunks of predominantly cancelous tissue are
present. All are presumed to be long bone fragments.
Cranial Remains
The cranial remains consist of five small fragments
and a relatively complete mandible. The body of the
mandible was broken (post mortem) at the socket of the
right canine, but has subsequently been restored. The left
ramus is missing, the body having been fractured through
the midline of the socket of M3. The right ramus is present
but broken superior to the mandibular foramen. Both the
coronoid and condylar processes are thus absent. The first
and second molars on both sides were lost ante-mortem
and the sockets had been resorbed, thus giving the
mandible a low profile. There is some evidence of
periodontal infection of the alveolar process in the region
of the anterior dentition. The chin is relatively square, but
the mandible as a whole lacks robustness.
Symphysial height - 30.5mm
The parietal is represented by a large rectangular
section of the right parietal. No sutures are evident. The
meningial grooves are not pronounced.
The maxilla is represented by a small section of the
right alveolar process containing the socket for M3 and the
distal wall of the socket of M2 with the channels for the
lingual and disto-buccal roots. Sections of both the right and
left greater sphenoid wing are present. The fragment of the left
greater wing is roughly triangular in shape having three
surfaces: the posterior, the superior or cerebral, and the orbital.
The inferior section is not preserved, thus the sphenoidal spine
and the sphenoidal foramina (ovale, rotundum, spinosum) are
absent. The fragment of the right greater wing is smaller and is
predominantly the cerebral surface. Only a small triangular
section of the posterior surface and a small section of the
spheno-squamosal suture are present.
Dental Remains
All mandibular dentition can be unambiguously
associated with the remains of the single individual
discussed above. In addition there are 14 isolated teeth,
seven of which were found in association with these
remains. The remaining seven were recovered from other
units. Due to an accident in the CWRU offices, all
dentition was placed together making it impossible to
determine which were associated. Triplication of
mandibular second and third molars (three left M2 and
three right M3) indicates the minimum number of
individuals is three. Table 1 lists the materials present and
the state of attrition or loss.
With the exception of several isolated teeth, the
general state of wear is relatively low. In the mandibular
dentition M, and M2 are missing on both sides and the
alveolar processes have been resorbed. There is a small
cavity on the mesial surface of the right P2 and the tooth is
rotated slightly distally and lingually. No other caries are
evident. Slight shoveling of the three incisors is evident. A
protostylid appears on the right M3. The isolated teeth are
in a very poor state of preservation. All are cracked or split,
some to the extent of having the entire pulp cavity eroded.
Age and Sex
Due to the paucity of material in general, the absence
of complete longbones, and the lack of reliable indicators
(e.g. pelvic or cranial material), determinations of age at
death and sex are uncertain. All available indicators were
utilized. Table 2 lists the metric analyses employed for the
determination of the sex of the specimen.
All listed measures indicated that the individual was
1989
Skeletal Remains
57
Table 1
Dental Attrition and Loss
Mandible Misc. Dentition
Maxillary central I
Right
•
Left
•
Right
0
Left
0
Maxillary lateral I
•
•
3*
•
Maxillary canine
•
•
6
0
Maxillary premolar 1
•
•
•
9
Maxillary premolar 2
•
•
0
0
Maxillary molar 1
•
•
6
*
Maxillary molar 2
•
•
6
5
Maxillary molar 3
•
•
0
5
Mandibular central I
4
1
0
0
Mandibular lateral I
4
4
*
0
Mandibular canine
1
4
0
0
Mandibular premolar 1
5
4
0
0
Mandibular premolar 2
5
5
0
5*
Mandibular molar 1
2
2
0
0
Mandibular molar 2
2
2
0
6* (6*)#
Mandibular molar 3
3
1
6 (6 *)#
5
I — postmortem loss, socket present; 2 — antemortem loss, socket resorbed; 3 — little; 4 — dentine visible;
5 — cusps gone; 6 — pulp exposed; • — not available; * — poorly preserved; # — two specimens
Table 2
Determination of Sex
Characteristic
Measurement
Indication
Reference
Humeral Head Diameter
Female
Krogman 1962:144
vertical
38.4mm
horizontal
36.0mm
Clavicle Length
138.4mm
Female
Krogman 1962:148
Scapular Dimensions
length of glenoid
33.6mm
width of glenoid
24.3mm
width of axillary border
3cm below infraglenoid
6.5mm
Mandibular Incisor and
Canine
Z=7.2423
Female
Ditch and Rose 1974
female. This finding is consistent with the general lack of
robustness which characterizes the remains. The only
contradictory evidence is the squareness of the chin which
appears somewhat masculine. It must be emphasized,
however, that all available indicators of sex are not only of
relatively low reliability, but also are all measures of sexual
dimorphism.
A tentative age at death of 25 years for this
individual has been based on the following criteria. Both
maxillary and mandibular third molars had erupted. The
right M3 was in occlusion, but is only slightly worn
(enamel polished) implying an age of 17-25 years
(Brothwell 1965:69). The epiphysis of the head of the
humerus was completely fused indicating an age in excess
of 24 years (Bass 1971:1 16); as were both epiphyses of the
clavicle indicating 25-28 years (Krogman 1962:32). Small
sections of the spheno-frontal and spheno-squamosal
sutures are present, but disarticulated, perhaps indicating a
relatively low degree of suture closure. No lipping of the
glenoid of the scapula is observed (lipping normally
commencing at age 30-35, Krogman 1963:55) nor are
arthritic changes of the vertebrae evident.
58
Prior
No. 44
Discussion
The fragmentary nature of this material makes
comparison with material from sites of similar antiquity
practically impossible. The few measurements possible are
of relatively little comparative interest. The relatively
moderate tooth wear stands in distinction to the finding of
extreme wear reported by Angel (1966) for the Tranquility
Site skeletal material. It is unknown whether this
discrepancy is due to a dietary difference or the relatively
young age of the Squaw Rockshelter individual. Perhaps
the ante-mortem loss of M[ and M2 is indicative of the
effects of a similarly coarse diet.
The skeleton is in general lightly built and rather
petite. There are no observable pathologies except those
associated with the mandible and mandibular dentition.
There is indication of neither trauma nor cause of death.
The predominance of remains from the right side would
indicate that the individual was buried lying on her right
side. A
Acknowledgements
The author thanks Professor Joseph Katich, CWRU School
of Dentistry, for independent classification of dental
remains.
References
Angel, Lawrence. 1966. Early Skeletons from Tranquility,
California. Smithsonian Contributions to Anthropology 2(1).
Bainbridge, D., and S. Genovese. 1956. A study of the sex
differences in the scapula. Journal of the Royal
Anthropological Society 86(2): 1 09- 1 34.
Bass, William. 1971. Human Osteology: A Laboratory and
Field Manual. Columbia: Missouri Archaeological Society.
Brothwell, D. 1965. Digging Up Bones. London: The British
Museum.
Ditch, L., and J. Rose. 1974. A Multivariate Dental Sexing
Technique. American Journal of Physical Anthropology
37:61-64.
Krogman, Wilton. 1962. The Human Skeleton in Forensic
Medicine. Springfield: C. C. Thomas.
KIRTLANDIA
The Cleveland Museum of Natural History
March 1989 Number 44:59-62
Knotted Cordage from
Squaw Rockshelter (33CU34),
Aurora Run, Cuyahoga County, Ohio
R.L. Andrews and J.M. Adovasio
Department of Anthropology
U niversity of Pittsburgh
Pittsburgh, Pennsylvania 15260
Abstract
A charred fragment of cordage from the late Middle
Archaic level of Squaw Rockshelter represents a Z-
spun, S-twist, 2-ply 4-yarn construction of bast
fiber. A complex series of overhand knots cannot be
attributed to specific function. With other
specimens, this fragment documents an aboriginal
textile industry of great antiquity in the Eastern
Woodlands.
Introduction
A single piece of knotted cordage was recovered
during the excavations at Squaw Rockshelter (33CU34). In
the present context, cordage refers to a class of elongate
fiber constructions which herein includes only knotted
cordage.
Methods
The single artifact thus distinguished was classified
according to the ply (1, 2, or more); the direction of initial
spin (S or Z); and the direction of final twist (S or Z). Also,
the specimen was scrutinized for the number and types of
knots present. Knots were identified and described
according to the terminology contained in Shaw (1972).
All pertinent measurements were made with a Helios
needlenose sliding dial caliper, and angles were measured
with a protractor. Measurements taken follow the
specifications of Emery (1966). The single piece of
cordage recovered from the site is ascribable to one basic
construction type which is described below.
Description Type 1:
Four-ply, Z-spun, S-twist Cordage
Technique and comments
This construction essentially consists of four Z-spun
single elements with a final S-twist (Figure la, b). The
specimen has been knotted at both ends in the following
fashion. Initially, two unspun, single-ply elements were
60
Andrews and Adovasio
No. 44
Figure 1. Reconstruction of Squaw Rockshelter Cordage:
a — twisted yarns and piys; b — knotted element ; c — reconstruction groups and knots
knotted near their passive ends with an overhand knot.
These elements were laid next to another pair of unspun
elements, and the free ends were then manipulated into
another overhand knot. Of the eight emergent ends, six
were further manipulated while two remained inactive.
These elements extend from the knot in groups of two and
four. At this point, a slight spin was imparted to those six
elements (the term spin denotes only initial twisting and
not spinning as that term is used in describing yarn
preparation). The bundle of four elements was immediately
divided into two groups of two each. The resultant pattern
of free ends is 2-2-2, and these units are labeled exterior
group, interior subgroup, and exterior subgroup,
respectively.
The interior subgroup was looped around the exterior
group and subgroup as shown (Figure 1 c) and was finally
combined with the exterior group of fibers. The exterior
subgroup of elements was looped around and behind the
other four elements, and its two plys were knotted into an
overhand knot, then ended by wrapping one ply around the
other and tucking under its loose end. The remaining four
actively spun elements were again subdivided into two
groups and S twisted. Each group functioned essentially as
one element for ca. 2cm after which the subgroups were
united again. The entire construction was then terminated
by whipping. The final whipping knot is the functional
equivalent of an overhand noose and is the fourth in a
series of overhand knots.
1989
Knotted Cordage
61
The specimen is thoroughly charred and highly
friable. The function of this construction is unknown.
Raw material
All elements of this specimen are constructed with an
unknown genus and species of plant material. Fred H.
Utech (1980, pers. com.) suggested that this raw material
consisted of grass or sedge leaves. Richard I. Ford (1987,
pers. com.) however, suggested to Brose that the material
was some sort of bast fiber. More specifically, James K.
Bissell and David S. Brose (1987, pers. com.) suggest the
specimen was made of cedar {Juniperus spp.) or hemlock
( Tsuga spp.) root. Pollen diagrams in the region of the site
indicate that although cedar and spruce were moderately
abundant at ca. 10,000 B.P., they had vanished by 6500
B.P. whereas hemlock appeared ca. 7500 B.P. and is still a
ravine dominant (Brose, this volume). The method of
preparation of the raw material is unknown.
Length — 6.40cm
Range in diameter of individual Z-spun elements —
1.75-2. 70mm
Mean diameter of individual Z-spun elements —
2.23mm
Range in diameter of two-ply units — 2.70-4. 65mm
Mean diameter of two-ply units — 3.68mm
Range in diameter of finished four-ply cordage —
6.95-7. 05mm
Mean diameter of finished four-ply cordage —
7.00mm
Number of twists per cm — 1
Angle of twist: 28°
Provenience and Chronology
This single specimen of cordage is ascribable to a
level in excavation unit 2 of the shelter which has been
radiocarbon dated at ca. 5300 B.P.
Internal Correlations
Despite the paucity of the data, a number of
observations can be made on the cordage construction from
Squaw Rockshelter. First, the single specimen is de facto
proof that a relatively sophisticated cordage industry
existed in this part of the Midwest in the mid-fourth
millennium B.C. Second, the specimen is technically well-
made and exhibits all of the “earmarks” of a complex
perishables industry. Specifically, the diameter of the
individual elements is remarkably consistent. The knots,
though simple, are combined in a complicated fashion, and
the overall impression of the piece suggests that its makers
were more than casually familiar with the manipulation of
string and rope.
The function of this particular specimen is unknown,
but it could represent one of a variety of items, such as
knotted fringe from a skirt or loin covering, a bundle of
twined basketry construction material, a part of a snare or
trap, or, perhaps, a “doodle” fashioned for no specific
purpose. Whatever its function, we stress that this item was
manufactured by an individual who was no stranger to the
production of high-quality cordage.
External Correlations
The specimen under discussion represents at once the
oldest well-dated piece of cordage recovered in the state of
Ohio as well as one of the older pieces of cordage ever
recovered east of the Mississippi River. Far older cordage
materials are known from western North America where
they are ascribable to at least the 1 1th millennium B.C. (see
Adovasio 1974, 1977). Other early specimens are known
from portions of Mesoamerica (Taylor 1966; MacNeish et
al. 1967) and South America (Adovasio and Lynch 1973).
Actual specimens of cordage or, indeed, any other kind of
perishables from the eastern United States are relatively
uncommon. Though it has been widely assumed that the
perishable industries of Archaic populations east of the
Mississippi were as old and as complex as those in the arid
West, actual proof is somewhat difficult to acquire because
of factors of preservation. This is the principal value of the
piece from Squaw Rockshelter as well as the scattered
perishable remains from localities with Paleolndian and/or
Archaic period components, such as Meadowcroft
Rockshelter, Pennsylvania (Stile 1982); Petit Anse Island,
Louisiana (Wilson 1888; Figure C VI 1 ); Graham Cave,
Missouri (Logan, 1952); Icehouse Bottom, Tennessee
(Chapman and Adovasio 1977); the Harts Falls site, Maine
(James L. Petersen 1984, pers. com.); Russell Cave,
Alabama (Griffin 1974); the Long Branch site, Alabama
(Webb and DeJarnette 1942); Salts Cave, Kentucky (King
1974); the Ozark Bluff Shelters, Arkansas (Schotz 1975);
the Picton site, Ontario, Canada (Ritchie 1949); the
Riverside site, Michigan (King 1968); and the Windover
site, Florida.
Cordage remains per se are limited in distribution to
Icehouse Bottom (Chapman and Adovasio 1977; 622),
Salts Cave (King 1974), the Picton site (Ritchie 1949), and
the Ozark Bluff Shelters (Scholtz 1975). However,
multiple-ply warp and/or weft elements produced via
cordage- making techniques are identifiable in twined
basketry and/or fabric from a variety of Archaic period
sites including Graham Cave (Logan 1952: Plate XXlc),
Icehouse Bottom (Chapman and Adovasio 1977), Salts
Cave (King 1974), the Riverside site (King 1968), and
Windover.
While both final S- and/or Z-twist cordage and
cordage warps/wefts are represented in the aforementioned
assemblages, the diversity of both geographic and
chronological distributions precludes any statistically
significant correlation of twist types. Taken as a unit, these
widely distributed remains indicate that populations east of
62
Andrews and Adovasio
No. 44
the Mississippi did possess a complex and technologically
sophisticated cordage and basketry industry of an antiquity
comparable to those in the Desert West. It is also
interesting to note that although the production of cordage
and basketry from grass stems is ethnographically well
documented in the Great Lakes region (Jones 1936;
Douglas 1939), the use of “soft,” bast fibers is widespread
among both prehistoric and ethnographic populations
(Driver and Massey 1957; Spencer and Jennings et al.
1977; Holmes 1896; Mason 1904; Swanton 1946).
Unfortunately, the condition of the Squaw Shelter
specimen precludes both an exact identification of raw
material and also the “method(s)” of its preparation.
Presumably, the recovery of more materials such as the
piece described here will greatly augment the very dim
picture we presently possess of the production of
perishables in portions of eastern North America. A
Acknowledgements
Special thanks are extended to Dr. Fred H. Utech, Botany
Department, Carnegie Museum of Natural History,
Pittsburgh; Dr. Richard I. Ford, Ethnobotanical Laboratory,
Museum of Anthropology, University of Michigan; and
James K. Bissell, Department of Botany, Cleveland
Museum of Natural History, for suggestions concerning the
identification of plant materials reported in this study.
References
Adovasio, J. M. 1974. Prehistoric North American Basketry.
Nevada Slate Museum Anthropological Papers 16:100-145.
Carson City.
1977. Basketry Technology. Chicago: Aldine Publishing
Company.
Adovasio, J. M., and T. F. Lynch. 1973. Prehistoric Textiles and
Cordage from Guitarrero Cave, Peru. American Antiquity
38 ( l):84-90.
Chapman. J., and J. M. Adovasio. 1977. Textile and Basketry
Impressions from Icehouse Bottom, Tennessee. American
Antiquity 42(4):620-625.
Douglas, F. H. 1939. Indian Basketry East of the Rockies.
Denver Art Museum Leaflet 87.
Driver, H. E., and W. C. Massey. 1957. Comparative Studies of
North American Indians. Transactions of the American
Philosophical Society. New Series 47(2).
Emery, I. 1966. The Primary Structure of Fabrics, an Illustrated
Classification. Washington: The Textile Museum.
Griffin, J. W. 1974. Investigations in Russell Cave. U.S.
Department of the Interior, National Park Service Publications
in Archaeology 13. Washington D.C.
Holmes, William H. 1896. Prehistoric Textile Art of Eastern
United States. Thirteenth Annual Report of the Bureau of
American Ethnology, 1891-1892. Washington, D. C.
Jones, V, H. 1936. Some Chippewa and Ottawa Uses of Sweet
Grass. Papers of the Michigan Academy of Science, Arts, and
Letters 2 1 .
King, M. E. 1968. Textile Fragments from the Riverside Site,
Menomine, Michigan. Verhandlungen des XXXVIII
internationalen Amerikanistendongresses I: 1 17-123.
1974. The Salts Cave Textiles: A Preliminary Account, pp. 31-
40 in Archaeology of the Mammoth Cave Area, edited by P. J.
Watson. New York: Academic Press.
Logan, W. D. 1952. Graham Cave: An Archaic Site in
Montgomery County, Missouri. Missouri Archaeological
Society Memoir 2.
MacNeish, R. S., A. Nelkin-Terner, and 1. W. Johnson. 1967. The
Prehistory of the Tehuacan Valley. Volume 11. Austin:
University of Texas Press.
Mason, O. T. 1904. Aboriginal American Basketry: Studies in a
Textile Art without Machinery. Report of the U.S. National
Museum for 1902: 171-548.
Ritchie, W. A. 1949. An Archaeological Survey of the Trent
Waterway in Ontario, Canada and Its Significance for New
York State Prehistory. Researches and Transactions of New
York State Archaeological Association 12(1). Albany.
Scholtz, S. C. 1975. Prehistoric Plies: A Structural and
Comparative Analysis of Cordage, Netting, Basketry, and
Fabric from Ozark Bluff Shelters. Arkansas Archaeological
Survey Research Series 9.
Shaw, G. R. 1972. Knots. New York: Collier Books.
Spencer, R. F., and J. D. Jennings, et al. 1977. The Native
Americans. New York: Harper and Row.
Stile, T. E. 1982. Perishable Artifacts from Meadowcroft
Rockshelter, Washington County, Southwestern Pennsylvania,
pp. 130-141 in Meadowcroft: Collected Papers on the
Archaeology of Meadowcroft Rockshelter and the Cross
Creek Drainage, edited by R. C. Carlisle and J. M. Adovasio.
Department of Anthropology, University of Pittsburgh.
Swanton, J. R. 1946. The Indians of the Southeastern United
States. Bureau of American Ethnology Bulletin 137.
Washington, D.C.
Taylor, W. W. 1966. Archaic Cultures Adjacent to the
Northeastern Frontiers of Mesoamerica. pp. 59-94 in
Handbook of Middle American Indians, Volume 4:
Archaeological Frontiers and External Connections, edited by
G. F. Ekholm and G. R. Willey, pp. 59-94. University of Texas
Press, Austion and London.
Webb, W. S., and D. L. DeJarnette. 1942. An Archaeological
Survey of Pickwick Basin in the Adjacent Portions of the
States of Alabama, Mississippi, and Tennessee. Bulletin of the
Bureau of American Ethnology 1 29. Washington, D.C.
Wilson, T. 1888. Ancient Indian Matting - From Petit Anse
Island, Louisiana. Report of the U.S. National Museum for
1888: 673-676.
M0(
IkT o
■NATURAL HISTORY-
Published by The Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Design and Production by Melody Oakes
KIRTLANDIA
Invertebrate Paleontology
Stratigraphic Distribution of Brachiopods and Bivalves
in the Upper Devonian (Famennian) Chagrin Shale
in the Cuyahoga River Valley , Northeast Ohio 7
Barbara A. Schwimmer and Rodney M. Feldmann
Zoology
Estimation of Numbers for a Riverine Necturus Population
Before and After TFM Lampricide Exposure 33
Timothy O. Matson
•NATURAL HISTORY*
JUNE, 1990
KIRTLANDIA
The Scientific Publication of The Cleveland Museum of Natural History
David S. Brose and
Joseph T. Hannibal. Editors
Brief History and Purpose:
Kirtlandia, a publication of The Cleveland Museum of Natural History, is named
m honor of Jared Potter Kirtland, a noted nineteenth-century naturalist who lived
in the Cleveland, Ohio area. It began publication in 1967 and is a continuation of
the earlier series Scientific Publications volumes 1 to 10 (1928-1950), and new
series volumes 1 to 4 (1962-1965). Supported by the Kirtlandia Society of The
Cleveland Museum of Natural History, Kirtlandia is devoted to the publication of
scientific papers in the various fields of inquiry within the Museum’s sphere of
interest: Cultural and Physical Anthropology; Archaeology; Botany; Geology;
Paleobotany; Invertebrate and Vertebrate Paleontology; Systematic Ecology; and
Invertebrate and Vertebrate Zoology. Issues will vary from single monographs to
collections of short papers, review articles, and brief research notes.
Kirtlandia is abstracted in Biological Abstracts and indexed in Bibliography and
Inde x of Geology and Zoological Record.
Kirtlandia No. 45
ISSN 0075-6245
© 1990 by The Cleveland Museum of Natural History
Cleveland, Ohio
KIRTLANDIA
The Cleveland Museum of Natural History
June 1990 Number 45
Invertebrate Paleontology
Stratigraphic Distribution of Brachiopods and Bivalves
in the Upper Devonian (Famennian) Chagrin Shale
in the Cuyahoga River Valley, Northeast Ohio 7
Barbara A. Schwimmer and Rodney M. Feldmann
Zoology
Estimation of Numbers for a Riverine Necturus Population
Before and After TFM Lampricide Exposure 33
Timothy O. Matson
KIRTLANDIA
Editors
David S. Brose
Joseph T. Hannibal
The Cleveland Museum of Natural History
Associate Editors
James K. Bissell, Curator of Botany
The Cleveland Museum of Natural History
Bruce Latimer, Curator of
Physical Anthropology
The Cleveland Museum of Natural History
Martin Rosenberg, Editorial Assistant
Case Western Reserve University
Sonja Teraguchi, Manager of Collections
The Cleveland Museum of Natural History
Editorial Advisory Board
Rodney Feldmann, Professor of Geology
Kent State University
Michael C. Hansen, Geologist
Ohio Geological Survey
Richard Meindl, Associate Professor
of Anthropology
Kent State University
G. Michael Pratt, Associate Professor
of Anthropology
Heidelberg University
David H. Stansbery, Director,
Museum of Zoology
Ohio State University
Frederick H. Utech, Curator of Botany
Carnegie Museum of Natural History
Ed Voss, Curator of the Herbarium
University of Michigan
Andrew M. White, Professor of Biology
John Carroll University
KIRTLANDIA
The Cleveland Museum of Natural History
June 1990 Number 45 :7 -31
Stratigraphic Distribution
of Brachiopods and Bivalves
in the Upper Devonian (Famennian) Chagrin Shale
in the Cuyahoga River Valley,
Northeast Ohio
Barbara A. Schwimmer *
and Rodney M. Feldmans
Department of Geology
Kent State University
Kent, Ohio 44242
Abstract
The Upper Devonian Chagrin Shale, exposed in the Cuyahoga River Valley in northeast
Ohio, has a relatively rich, brachiopod-dominated fauna. Two stratigraphic sections were
sampled to determine the distribution of brachiopods in these shale and siltstone beds.
Nineteen brachiopod and four bivalve taxa have been described from the unit in this study.
Conodont faunas, and to a lesser extent, brachiopod ranges, provide evidence of a late
Famennian age for the Chagrin Shale
The Chagrin sediments were deposited on a relatively shallow shelf, below normal wave
base. Sedimentation rates were slow, energy was low, and the substrate was fluid. Turbidity
was at least moderate, as evidenced by dominant colonization by brachiopods instead of
bivalves. Episodic storms carried sediments westward from the prograding Catskill Delta
complex, leading to the periodic burial of brachiopod communities. The low degree of
fragmentation and lack of abrasion of specimens suggest rapid burial rates. Preservation was
influenced by the processes of dissolution, pyritization, and phosphatization.
* Present address : The Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
8
SCHWIMMER AND F ELD MANN
No. 45
K M
i j
FIGURE 1. Upper Devonian shale outcrop pattern in northeastern
Ohio, showing the locations of the two study sites in the Cuyahoga
River Valley.
Introduction
The purpose of this study is to describe brachiopod
and bivalve faunas and their stratigraphic distribution in the
Chagrin Shale from Chippewa Creek at Brecksville
Reservation on the western side of the Cuyahoga River
Valley, and from Brandywine Creek at Brandywine Falls,
on the eastern bank (Figure 1). The brachiopod fauna
supports a late Famennian age for the Chagrin Shale.
The Chagrin Shale is exposed across 150 kilometers of
northeast Ohio, along the southern shore of Lake Erie, and
along many of the streams and their tributaries which flow
into the lake (Figure 1). The Chagrin attains a maximum
thickness of 400 m at the Ohio-Pennsylvania border, and
thins to an estimated 35 meters in the Cleveland area
according to Szmuc et al. (1976). Pepper et al. (1954)
described the Chagrin as a thick wedge of interbedded
siltstones and shales, bounded above by the Cleveland
Shale, and below by the Fluron Shale, and pinching out
between these two black shales in a westward direction.
The Cuyahoga River flows through the western part of the
Upper Devonian shale outcrop area, where its valley is
floored by the Chagrin Shale. From Cuyahoga Falls,
Summit County, the river travels northwest across Summit
and Cuyahoga Counties in a pre-glacial valley,
characterized by steep hills and cliffs for many miles. The
Cuyahoga River tributaries often cut valleys or gorges
through the hills, in which many of the best Chagrin Shale
exposures are seen.
Various aspects of the Chagrin Shale have been
studied in recent years: the petrography (Broadhead and
Potter, 1980); facies relationships (Potter et al., 1980);
sedimentation (Potter et al., 1980); trace fossils (Feldmann
et al., 1978; Hannibal and Feldmann, 1983; Stukel, 1987);
and arthropods (Weidner and Feldmann, 1983). Previous
reports (Prosser, 1912; Cushing et al., 1931; Hannibal and
Feldmann, 1983; and Weidner and Feldmann, 1983) cited
the sparsely fossiliferous nature of the Chagrin Shale.
While faunal elements are restricted, recent studies, listed
below, concluded that the Chagrin fauna is more diverse
and abundant than previously supposed. Stukel (1987)
described fifteen ichnotaxa from the unit; Weidner and
Feldmann (1983) reported several species of arthropods
found in the Chagrin; and Feldmann et al. (1986) identified
fossil worms from the unit. The most abundant megafaunal
elements appear to be brachiopods, with bivalves as a
minor constituent of the total fauna.
Localities
Locations of the two sections studied are shown in
Figure 1. Chippewa Creek, located on the western side of
the Cuyahoga River Valley, flows through Brecksville
Reservation, one of the Cleveland Metroparks. The
Brecksville section is located in the west-central 1/9 of the
Northfield, Ohio 7.5 min. topographic map (1963,
photorevised 1979). The section begins on the southern
side of the creek, at the base of the Bridle Path, and extends
upstream, approximately 900 m, to a steep cliff on the
north side of the creek. The upper part of the measured
section is approximately 170 m north of the Trailside
Interpretive Center on Chippewa Creek Road.
Brandywine Creek, part of the Cuyahoga Valley
National Recreation Area, lies on the eastern side of the
Cuyahoga River Valley. It is located in the southeast 1/9 of
the Northfield, Ohio 7.5 min. topographic map (1963,
photorevised 1979). From the intersection of Brandywine
Road and 1-271 on the map, the measured section is 580 m
west on Stanford Road, and 50 m north into the park area.
The Chagrin crops out on the banks of Brandywine Creek
for almost a kilometer, ending west of a waterfall,
approximately one-half km upstream.
Lithologies
The Chagrin Shale consists of interbedded greenish-
gray to bluish-gray shales, and discontinuous siltstones. The
siltstones range in thickness from one to tens of centimeters,
generally increasing in number and thickness eastward,
toward the Ohio-Pennsylvania border (Stukel, 1987).
At Brecksville Reservation, the Chagrin Shale
exposure measures approximately 6 m from stream level,
at the base of the Bridle Path, to the contact with the
overlying Cleveland Shale. Here the Chagrin consists of
1990
Brachiopods and Bivalves in the Chagrin Shale
9
generally fissile, greenish-gray shales, current-rippled,
cross-laminated siltstones, and plane-parallel laminated
siltstones (Stukel, 1987). Near the contact with the
Cleveland Shale, the Chagrin at Brecksville exhibits
several thin, horizontal, discontinuous pyrite beds. Pyrite is
abundant in other forms near the contact, and is manifested
in concretions, as nodules, as coatings on fossil molds, and
as vertical burrow fillings. Brachiopods at Brecksville
occur throughout the shales and at the bases and tops of
many of the siltstone beds. Fossils are generally not found
within the siltstones themselves.
In contrast, the Chagrin Shale at Brandywine Falls
measures about 10 m in thickness, and consists of bluish-
gray shales with interbedded, discontinuous siltstones.
Siltstones exhibit current-rippled, cross-laminated bedding,
plane-parallel bedding, and very gentle hummocky cross-
stratification. At Brandywine the types of pyrite found at
Brecksville are typically absent. Pyrite occasionally occurs
as coatings on brachiopod molds. Brachiopods at
Brandywine are distributed throughout the shales, and at
the bases and tops of some siltstone beds. Again, fossils are
rare within the siltstones.
In general, the units within the Chagrin at the
Brecksville Reservation and Brandywine Falls sites cannot
be correlated on the basis of lithology. The discontinuous
nature of the siltstone beds, and the absence of horizontal
pyrite layers at Brandywine Falls preclude using these
parameters for correlation. There is no distinctive bedding
feature of the shales or siltstones that permits a bed to bed
correlation between sections. Some patterns of faunal
distribution only provide an approximate correlation
between the two sections. Therefore the two sections must
be correlated using the Chagrin Shale-Cleveland Shale
contact as a datum, even though it is erosional in nature.
Any differences in elevation between the two sites, which
are approximately 9 km apart, might be explained in terms
of regional dip and/or differential rates of erosion at the
sites. The effect of regional dip is minimized because the
two sections are in close proximity.
Methods
Two sections in the Cuyahoga River Valley were
sampled to determine the stratigraphic distribution of
brachiopods and bivalves (Figures 2 and 3). Each section
was trenched and sampled on a bed by bed basis using a
pocket knife to remove individual shale and siltstone units.
The trench at Brecksville Reservation exposed
approximately 6 m of section, while the trench at
Brandywine Falls exposed approximately 10 m of section.
The trenched surface area averaged 20 cm by 30 cm. At
each level where brachiopods and bivalves were
discovered, collections were made, fossil orientation noted,
and distance from the datum was measured. No attempt
was made to determine numerical abundance at each level.
Rather, generic identification and stratigraphic position
were the primary goals.
Paleontology
Preservation
The Chagrin Shale megafauna in the Cuyahoga Valley
consists almost exclusively of brachiopods, but includes a
number of bivalves as well. Original shell material and
internal structures of brachiopods are often preserved,
notably of the spiriferids, orthids, and inarticulates.
Densities of fossils in the Chagrin vary; often a
bedding surface will contain clusters of one or two species,
or randomly spaced taxa, and other surfaces may be barren.
The degree of valve articulation is moderate, and the
degree of fragmentation is low. Many specimens occur in
life position, while others occur in more hydrodynamically
stable positions.
Some fossils have original shell material preserved so
that delicate ornamentation is visible. Others are well-
preserved molds of the interior, exhibiting muscle scars and
other structures. However, most of the brachiopods are
characterized by poorly preserved molds of the interior.
Juvenile and adult stages were observed in most species.
Evidence of predation is lacking, and epibionts are rare.
In some forms, shell material is absent, and molds of
the exterior are rare. Sometimes critical characters, such as
cardinalia and interareas, are not available for
identification. While some of this may be due to selective
positioning after death, or to some disarticulation and
fragmentation in the normal course of events between
storm activities, diagenesis should also be considered.
Rapid burial may be responsible for a variety of chemical
reactions, including skeletal dissolution, pyritization, and
phosphatization of some faunal elements.
In the Chagrin brachiopods, skeletal material is often
dissolved, leaving only molds or casts. Burial in organic
rich muds often accelerates dissolution (Brett and Baird,
1986), probably due to the acidic nature of the sediments.
Pyrite assumes a variety of forms in the Chagrin Shale,
including concretions, nodules, burrow infillings, and thin
coatings on fossil molds. Since most bottom waters contain
some dissolved oxygen, Berner (1984) indicated that pyrite
usually precipitates in anoxic environments below the
sediment-water interface. During periods of rapid burial,
organic material is exposed to interstitial waters rich in
reactive iron compounds. Together, these are subject to
sulfate reduction by anaerobic bacteria, and pyrite is
formed. In some of the Cuyahoga Valley brachiopods,
pyrite occurs as thin coatings or as crystals encrusting
fossil molds. From the sporadic occurrence of pyrite
associated with fossils, it can be concluded that anaerobic
microenvironments were locally present in the sediments.
FIGURES 2 (BreCKSVILLE) AND 3 (Brandywine Falls). Distribution patterns of taxa at Chippewa Creek,
Brecksville Reservation, and at Brandywine Creek. Brandywine Falls, respectively. Black intervals represent siltstones, and
white intervals represent shales. Twenty— cm sample intervals are indicated directly to the right of the stratigraphic column.
10
SCHWIMMER AND FELDMANN
No. 45
interval
Acanthatia ?
Ambocoelia
Athyris - Composita
Aulacella
Cent ror hynchus
Cyrtospirifer
"Leiorhynchus"
Lingula
Orbiculoidea
productellid gen indet
Retichonetes
Schellwienella
Sphenospira
Spinospiriter
Toryniferella
Trigonoglossa ?
Leptodesma
interval
Acanthatia ?
Ambocoelia
Athyris- Composita
Aulacella
Centrorhynchus
Cyrtospirifer
"Leiorhynchus"
Lingula
Orbiculoidea
productellid gen. indet
Retichonetes
Schellwienella
Sphenospir a
Spinospiriter
Toryniferella
Trigonoglossa ?
Leptodesma
Pterinopecten ?
BRECKSVILLE BRANDYWINE FALLS
1990
Brachiopods and Bivalves in the Chagrin Shale
11
It must be noted that several thin (<1 cm), horizontal pyrite
layers were found near the erosional Chagrin Shale-
Cleveland Shale contact at Brecksville Reservation. The
topmost of these is the Skinner’s Run pyrite bed, also found
elsewhere in the Chagrin outcrop area, at the contact with
the overlying black Cleveland Shale. Pyritized material was
exhumed and reworked during periods of erosion on
normally anaerobic sea floors (Zagger and Banks, 1989).
The Skinner’s Run pyrite bed may, in turn, be analogous to
the Leicester Pyrite Member, which is associated with a
regional unconformity {Baird and Brett, 1986).
Phosphatic concretions are sometimes found in the
Cuyahoga Valley, but can be found in much greater
numbers at numerous sites in the central and eastern
sections of the Chagrin outcrop area (Schwimmer et al.,
1987). Manheim et al. (1975) listed conditions necessary
for formation of modern marine phosphorite. These
included: sediments with high TOC (total organic carbon)
associated with dysaerobic waters, low rates of
sedimentation and a low volume of calcium carbonate in
sediments. Slansky (1986) noted that certain modern
phosphates result from the reworking and concentration of
existing phosphatic bioclasts, such as inarticulate
brachiopod shells, arthropod remains, or bone fragments.
Apatite precipitation may be triggered at one or more
“nucleation sites,” which initially become coated with
humic acids. Phosphate precipitation is catalyzed by the
acids, and not directly by the phosphatic surface of the
organism. Once the process is initiated, precipitation
continues, due to the lateral diffusion of ions from
interstitial waters. Wetzel (1983, p. 264) stated that, if
sediments are anoxic, phosphates may be released from
sediments, diffused into surface waters and, over time,
would be lost. However, if an oxidized zone is present near
the sediment-water interface, then phosphate could be
reduced by ferrous iron and precipitated as ferric phosphate
and by absorption onto ferric hydroxide and calcium
carbonate (Wetzel, 1983, p. 261-3).
Phosphate formation is favored by rapid burial of
organic material, followed by long periods of little or no
sedimentation (Brett and Baird, 1986). Wet chemical tests
and analysis by energy-dispersed x-rays (EDX) have
determined that concretions found in the Chagrin are
phosphatic. The surrounding silts and shales tested
negative for phosphorus. Generally, phosphatic material
such as lingulid brachiopod shells, arthropod remains, and
fish fragments acted as nucleation sites for phosphate
formation in Chagrin concretions. EDX analyses also
showed that non-phosphatic material in the concretions,
such as pelmatozoan fragments, and the matrix itself, are
enriched in phosphate (Joseph Hannibal, pers. comm.).
In summary, the Chagrin brachiopods show varying
qualities of preservation. Some display shell material and
delicate internal structures. Others are less well preserved,
and are oriented so that vital identifying characters are
either absent or obscured. Often molds are the only
evidence of ancient life, and these cannot always be
removed from the surrounding matrix to expose
undersurfaces. The unbroken, articulated appearance of
many of the specimens, as well as their orientation,
distribution, and density, is evidence for quiet water
conditions, with periodic burial of life assemblages by
storm sediments. Chemical alteration includes shell
dissolution, pyrite formation in local anaerobic
microenvironments, and phosphatization during times of
low sedimentation, after episodes of rapid burial.
Summary oftaxa
The Cuyahoga Valley Chagrin fauna consists
predominantly of brachiopods, with fewer bivalves. The
description of taxa is designed not as a formal systematic
description of the fauna, but as a guide for identification.
The most important diagnostic features are presented along
with remarks about preservation, distribution, and
comparisons with similar taxa. The vertical distribution of
brachiopods and bivalves at Brecksville Reservation and at
Brandywine Falls are shown in Figures 2 and 3,
respectively. The black intervals in each stratigraphic
section represent silts, while the white intervals represent
shales. For ease in charting distributions, each stratigraphic
section was arbitrarily divided into consecutively
numbered 20-cm intervals, shown to the right of each
column. Distribution patterns discussed for each taxon will
refer to these figures.
All figured specimens are deposited in The Cleveland
Museum of Natural History (CMNH). The remainder of
the material from this study is deposited in the
paleontology collection in the Department of Geology at
Kent State University (KSU).
Six orders of brachiopods have been identified in the
Upper Devonian Chagrin Shale of the Cuyahoga Valley.
The Inarticulata are represented by Lingulida and
Acrotretida, while the Articulata are represented by
Orthida, Strophomenida, Rhynchonellida, and Spiriferida.
Four taxa of bivalves, representing three orders, are also
presented.
Systematic paleontology
Phylum Brachiopoda Dumeril, 1806
Class Inarticulata Huxley, 1869
Order Lingulida Waagen, 1885
Superfamily Lingulacea Menke, 1828
Family Lingulidae Menke, 1828
Genus Lingula Bruguiere, 1797
Lingula eriensis Girty, 1939
Figures 4.5a, b, 4.6
12
SCHWIMMER AND FELDMANN
No. 45
FIGURE 4. Inarticulata. 1-2, Orbiculoidea sp. 1, pedicle interior, Mill Creek, KSU 4899; 2, mold of brachial valve interior, Brecksville
Reservation, interval 6, CMNH 8401 ; 3a, b. Lingula arcta, partially exfoliated pedicle valve in concretion, Brandywine Falls, interval 59,
CMNH 8402; 4, Trigonoglossa sp., partial pedicle valve, Brecksville Reservation, interval 17, CMNH 8403; 5-6, Lingula eriensis. 5a,b,
partially exfoliated specimen in concretion, Brandywine Falls, interval 67, CMNH 8404; 6, specimen found in living position, valves
splayed, Brecksville Reservation, interval 20, CMNH 8405. Bar scale = 1 cm.
Description of material
Shell medium size, ovate, anterior end rounded,
posterior end slightly pointed. Length approximately 1.2- 1.5
times width. Valves generally flat, slightly biconvex in
umbonal regions. Lateral margins rounded. Surface
ornamentation consists of irregularly spaced growth lines,
and very faint radial striations. Exfoliated specimens show
alternating chitinous and phosphatic shell layers. Chitinous
shell layers marked by closely spaced, regular, concentric
lines of a different character from the external growth lines.
Pedicle valve interior with cordate visceral area highlighted
by curved, transverse striations. Central muscle scars ovate.
In some specimens, pedicle groove is seen at posterior end.
Brachial valve interior displays median ridge which extends
almost entire length of shell. Muscle scars not visible.
Remarks
Lingulids are rare and occur either as sets of splayed
valves in the shales (Figure 4.6), or in concretions in the
siltstones. Although original shell material is present, often
it is only partially preserved, and internal structures cannot
be seen. Lingulids are far more common in the eastern part
of the Chagrin outcrop area, where sediments are coarser.
The burrowing lifestyle of lingulids is more suitable to
coarser sediments than to fine grained muds (Thayer and
Steele-Petrovi'c, 1975). The low bulk density of lingulids
would cause them to float on a fluid substrate, unable to
assume their normal feeding position. Lingulids were most
abundant in the lower half of both the Brecksville
Reservation and Brandywine Falls sections (Figures 2 and
3). Girty (1939) reported L. eriensis and L. arcta Girty
from the Chagrin Shale at Brecksville Reservation.
However, the two species of Lingula found at these study
sites cannot always be distinguished in the field, therefore
distribution patterns outlined in Figures 2 and 3 reflect
generic and not species patterns.
Lingula arcta Girty, 1939
Figures 4.3a, b
Description of material
Small size, elongate oval outline, length greater than
width. Lateral margins parallel. Surface ornament consists
of very fine, evenly spaced, concentric growth lines.
Pedicle valve umbo elevated posteriorly, and slightly
pointed. Brachial valve and valve interiors unknown.
Remarks
One good specimen of a partially exfoliated pedicle
valve was recovered from an in situ concretion at the
Brandywine Falls section (Figure 4.3). Girty (1939)
identified this species from the Chagrin Shale at
Brecksville Reservation, and noted that it often occurred
with L. eriensis. He stated that the two species were
distinct, based upon shape and surface ornamentation.
Growth lines on L. arcta are very fine and closely spaced.
1990
Brachiopods and Bivalves in the Chagrin Shale
13
FIGURE 5. Aulacella sp. !, interior, pedicle valve, Brecksville Reservation, interval 28, CMNH 8406; 2, interior mold of brachial valve,
Brecksville Reservation, interval 27, CMNH 8407; 3a, b, posterior view, and brachial valve exterior, Brecksville Reservation, float, CMNH
8408 ; 4, brachial valve interior, Brecksville Reservation, interval 26, CMNH 8409; 5, pedicle valve interior, Brecksville Reservation,
float, CMNH 8410; 6, interior mold of pedicle valve, Brecksville Reservation, interval 28, CMNH 8411 ; 7, pedicle valve exterior,
Brecksville Reservation, float, CMNH 8412. Bar scale = 1 cm.
while those on L. eriensis are coarser, farther apart, and less
regular. Girty (1939) also reported shell fragments of what
could be L. limatnla from the Chagrin Shale at Brecksville.
The latter species is characterized by a markedly papillose
shell surface. Neither of the lingulids collected from the
Chagrin Shale in this study exhibit this character.
Genus Trigonoglossa Dunbar and Condra, 1932
Trigonoglossa sp.
Figure 4.4
Description of material
Shell medium size, triangular outline, gently convex
pedicle valve, brachial valve unknown. Posterior end
pointed. Length greater than width; greatest width in
anterior part of specimen. The axial region of the specimen
appears to be inflated, forming a gentle fold. Valve surface
marked by strong, evenly spaced, concentric lines
prominently elevated above surface of the valve, separated
by flat interspaces. Very faint radial striations. Valve
interiors unknown.
Remarks
One partial, somewhat crushed specimen was
recovered from Brecksville Reservation. The prominent
growth lines and triangular shape distinguish it from
species of Lingula.
Order Acrotretida Kuhn, 1949
Suborder Acrotretidina Kuhn, 1949
Superfamily Discinacea Gray, 1 840
Family Discinidae Gray, 1840
Subfamily Orbiculoideinae Schuchert, 1929
Genus Orbiculoidea d’Orbigny, 1847
Orbiculoidea sp.
Figures 4.1, 4.2
Description of material
Small to large size, circular outline, convex brachial
valve, concave pedicle valve. Brachial valve apex pointed,
posteriorly eccentric; diameter of figured specimen (Figure
4.2) approximately 3.2 cm. Ornament includes strong,
elevated, evenly spaced, concentric growth lines. Smooth,
wide depressions separate growth lines. Very faint radial
striations are present, especially in the apical area, dying
out anteriorly. Posterior margin rounded, not truncated.
Muscle scars and other internal structures not observed.
Remarks
Three occurrences of Orbiculoidea were noted only in
the section at Brecksville (Figure 2). One partial brachial
valve, found atop a silt bed, was complete enough for
identification. In contrast, Weidner (1983) found many well
preserved orbiculoids in concretions in other parts of the
Chagrin outcrop area. The pedicle valve of one such specimen
from Mill Creek is illustrated (Figure 4. 1 ) for comparison. The
14
SCHWIMMER AND F ELD MANN
No. 45
FIGURE 6. Schell wienella sp. 1, interior mold of pedicle valve, with dental plate impressions, Brandywine Falls, interval 57, CMNH
8413; 2, posterior mold of interarea, deltidium and dental plates, Brandywine Falls, float, CMNH 8414; 3a. b. interior mold of pedicle
valve, brachial valve, Brandywine Falls, float, CMNH 8415; 4, interior mold of pedicle valve and interarea, Brandywine Falls, float,
CMNH 8416; 5, interior mold of pedicle valve, Brandywine Falls, interval 58, CMNH 8417; 6, interior mold of brachial valve, cardinal
process, Brecksville Reservation, interval 24, CMNH 8418. Bar scale = 1 cm.
size, shape, and ornamentation appear to closely resemble the
Brecksville specimen (Figure 4.2). Pedicle foramen is narrow,
linear; 8 mm length; posteriorly located with respect to the
apex. Internal structures unknown.
Class Articulata Fluxley, 1 869
Order Orthida Schuchert and Cooper, 1932
Suborder Orthidina Schuchert and Cooper, 1932
Superfamily Rhipidomellacea Alichova, 1960
Family Onniellidae Opik, 1933
Genus Aulacella Schuchert and Cooper, 1931
Aulacella sp.
Figures 5. 1-5.7
Description of material
Small subcircular shell; valves unequally biconvex;
width greater than length; maximum width at mid-valve;
hinge narrow, short, approximately half width of shell;
delthyrium open; valve surfaces finely costellate; costellae
extend from beak to anterior margin, and increase by
intercalation; pedicle valve fold and brachial valve sulcus
strong posteriorly, becoming weak anteriorly; anterior
margin rectimarginate; valves inflated in areas of umbones,
flattening laterally and anteriorly. Concentric ornamentation
present near anterior margin. Pedicle valve interior with
large, flabellate diductor muscle scars, which almost
enclose small, spatulate adductor scars; a forked
myophragm is seen anteriorly, separating diductor scars.
Brachial valve interior with prominent, trilobed, cardinal
process, and diverging brachiophore bases beneath elongate
socket ridges; posterior flabellate, and anterior subquadrate,
adductor muscle scars located on either side of dorsal fold.
Remarks
Aulacella can be distinguished from Cariniferella by the
presence in the latter of a convexo-concave profile, a sharp
fold, and a deep, narrow, sulcus (Schuchert and Cooper,
1932). Although Prosser (1912) reported species of
Cariniferella in the Chagrin, none was found in this study.
Aulacella is ubiquitous in the lower half of the Chagrin at
Brecksville, decreasing in numbers and in occurrence in the
upper half. In contrast, Aulacella is less abundant and is
found in only seven intervals in the lower third of the section
at Brandywine Falls, and in only one instance in the upper
third (Figures 2 and 3). Aulacella is the only punctate taxon
of Chagrin brachiopods. This feature may have enabled it to
respire at reduced capacity, even when valves were shut
during episodes of turbulence, which would normally clog
the lophophore (Thayer, 1986; Shumway, 1982).
Order Strophomenida Opik, 1934
Suborder Strophomenidina Opik, 1934
Superfamily Davidsoniacea W. King, 1850
Family Meekellidae Stehli, 1954
Subfamily Meekellinae Stehli, 1954
Genus Schellwienella I. Thomas, 1910
SCHELLWIENELLA Sp.
Figures 6. 1-6.6
1990
Brachiopods and Bivalves in the Chagrin Shale
15
FIGURE 7. Retichonetes sp. la,b, partially exfoliated brachial valve with four hinge spines on either side of pedicle valve umbo, and
enlarged view of pedicle valve interarea, Brecksville Reservation, float, CMNH 8419; 2, interior mold of pedicle valve, with costellae and
growth lines creating reticulated appearance, Brandywine Falls, interval 59, CMNH 8420; 3, interior mold of brachial valve, median
septum, Brandywine Falls, interval 73, CMNH 8421 ; 4, partially exfoliated, pyritized brachial valve and pedicle valve spines,
Brandywine Falls, interval 62, CMNH 8422; 5, interior mold of pedicle valve, flabellate diductor scars, divergent dental plates,
Brecksville Reservation, interval 19, CMNH 8423; 6, interior mold of brachial valve, Brecksville Reservation, interval 23, CMNH 8424; 7,
mold of brachial valve, median septum, alveolus, and divergent lateral septa, Brandywine Falls, interval 73, CMNH 8425 ; 8, mold of
brachial valve interior, pedicle valve interarea, Brecksville Reservation, interval 23, CMNH 8426. Bar scale - 1 mm.
Description of material
Small to medium size, semicircular outline, slightly
resupinate profile; width greater than length. Valves
generally flat, slightly convex in umbonal regions; hinge
width slightly less than greatest shell width, which occurs
at mid-length. Interarea straight, relatively high; triangular
delthyrium covered by convex pseudodeltidium. Pedicle
and brachial valves finely costellate; costellae increase by
intercalation, and appear to curve gently as they approach
posterior lateral margins; concentric ornamentation sparse.
Fold and sulcus either weak or absent; anterior margin
rectimarginate. Pedicle valve interior exhibits short dental
plates diverging at wide angles; diductor muscle scars faint,
large, flabellate. Brachial valve interior with small, bilobed,
cardinal process, thin, short median ridge, bisecting
slender, spatulate, adductor muscle scar.
Remarks
Schellwienella is distinguished from Schuchertella by
the presence of short, diverging, dental plates. These plates
are lacking in the latter genus. Most of the Chagrin
specimens of Schellwienella are incomplete; generally the
anterior and lateral margins are broken off, and original
shell material is absent. In one specimen, however, (Figure
6.3a), the original shell material comprising the dental
plates is present. A spirolophous lophophore was probable
in all stocks of the superfamily Davidsoniacea, although
impressions or calcareous supports are rare (Muir- Wood
and Williams, 1965). Schellwienella is moderately
abundant in the lower half of the Brecksville section, but
occurs in only one interval in the upper half. At
Brandywine Falls, this genus is more abundant, occurring
in the lower two-thirds of the section, but with only one
occurrence in the upper third of the site (Figures 2 and 3).
Suborder Chonetidina Muir- Wood, 1955
Superfamily Chonetacea Bronn, 1862
Family Chonetidae Bronn, 1862
Subfamily Devonochonetinae Muir- Wood, 1962
Genus Retichonetes Muir- Wood, 1962
Retichonetes sp.
Figures 7. 1-7.8
Description of material
Small size, semicircular outline; gently biconvex;
width greater than length; hinge line straight, and equal to
greatest width of shell; interareas moderately high;
minimum of four spines project at high angles (45-60
degrees) from pedicle valve interarea, adjacent to umbo;
delthyrium covered by pseudodeltidium. Valve exteriors
16
SCHWIMMER AND FELDMANN
No. 45
FIGURE 8. Productellids. 1-3, Productellid, gen. indet. 1, interior mold of pedicle valve, sulcus, anterior rugae, Brandywine Falls,
interval 57, CMNH 8427; 2a, b, interior mold of pedicle valve, with beak and divergent “septa” , Brecksville Reservation, interval 24,
CMNH 8428; 3, interior mold of pedicle valve, spine ridges, Brandywine Falls, float, CMNH 8429; 4-14, Acanthatia sp. 4, interior mold
of brachial valve, Brecksville Reservation, interval 24, CMNH 8430; 5, interior mold of brachial valve, endospine impressions,
Brandywine Falls, interval 58, CMNH 8431 ; 6, pedicle valve exterior, elongate hinge spines (arrow), prostrate body spines, borehole,
Brecksville Reservation, interval 4, CMNH 8432 ; 7, mold of pedicle valve exterior, hinge and lateral spines, Brandywine Falls, float,
CMNH 8433; 8, interior mold of brachial valve, cardinal process, breviseptum, Brandywine Falls, float, CMNH 8434; 9, Brachial valve
exterior, spine impressions, Brecksville Reservation, interval 28, CMNH 8435; 10, exterior mold of pedicle valve, small ears, spine
impressions, Brandywine Falls, interval 62, CMNH 8436; 11, interior mold of brachial valve, breviseptum, Brandywine Falls, float,
CMNH 8437; 12, interior mold of brachial valve, bilobed cardinal process, endospine impressions, Brandywine Falls, float, CMNH 8438;
13, interior mold of brachial valve, alveolus; interior mold of pedicle valve interarea, pseudodeltidium, hinge spines, lateral spines
(arrow), Brecksville Reservation, interval 27, CMNH 8439; 14, interior mold of pedicle valve, hinge spines, prostrate spines, Brecksville
Reservation, float, CMNH 8440. Bar scale = 1 cm.
1990
Brachiopods and Bivalves in the Chagrin Shale
17
capillate; capillae increase by intercalation; fold and sulcus
absent; anterior margin rounded. Pedicle valve interior with
short median septum, originating at posterior margin, and
extending approximately one-fourth shell length; dental
plates present, diverging at wide angle; diductor muscle scars
large, flabellate. Brachial valve interior with small, lobate
cardinal process; muscle scars small, slightly flabellate.
Remarks
Prosser (1912) reported and described Chonetes
minutus Prosser, 1912 from the Chagrin Shale in Ashtabula
County, and Chonetes scitulus Hall, 1857 from the Chagrin
Shale at Tinker’s Creek in Cuyahoga County. Retichonetes is
differentiated from Chonetes by the reticulate appearance of
the valves, and by the presence of an anterior median septum
and two short lateral septa in the brachial valve, as opposed
to three or more fine, long, diverging septa in the brachial
valve of Chonetes (Muir- Wood, 1962). Retichonetes occurs
throughout both of the measured sections (Figures 2 and 3).
Suborder Productidina Waagen, 1883
Superfamily Productacea Gray, 1840
Family Productellidae Schuchert, 1929
Subfamily Productellinae Schuchert, 1929
Productellid gen. indet.
Figures 8. 1-8.3
Description of material
Several medium size pedicle valves are not assignable
to a genus or to a species. They retain the convex,
subtrigonal, productellid shape, but an absence of brachial
valves, interareas, and cardinalia preclude more specific
identification. The inflated pedicle valve appears to be
rugose anteriorly, and to be ornamented with evenly spaced
spine ridges. The beak extends over the hinge area, and
curves inward. Width is greater than length, with the
greatest width at midlength. A mold of the exterior of the
pedicle valve confirms the presence of ears. Although one
specimen (Figure 8.1) has a prominent sulcus, a feature
common in leioproductids, the lack of a median ridge
ornamented with a spine row makes assignment to this
group questionable. Another specimen (Figure 8.2b) has
two converging “septa” originating at the umbo and
extending down the center of the valve to midlength.
Specimens of this taxon appear to be present, in small
numbers, throughout both Brecksville and Brandywine
Falls (Figures 2 and 3).
Subfamily Chonopectinae Muir- Wood and Cooper, 1960
Genus Acanthatia Muir- Wood and Cooper, 1960
Acanthatia sp.
Figures 8.4-8.14
Description of material
Small to medium size, semicircular outline. Width
greater than length. Hinge length slightly less that greatest
width, which occurs at midlength. Pedicle valve
moderately convex, brachial valve concave. Pedicle valve
ears small, if present, and not differentiated from visceral
disc. Pedicle valve interarea longer and higher than
brachial valve interarea, and characterized by
pseudodeltidium (Figure 8.13). Pedicle valve ornamented
by evenly-spaced, concentrically arranged prostrate spines
over surface, and long balancing or attachment spines at
hinge. Pedicle valve slightly rugose, especially anteriorly.
Brachial valve with depressions corresponding to pedicle
valve spines (Figure 8.12). Pedicle valve interior often
marked by very fine punctation, representing endospines.
Brachial valve interior with bilobed cardinal process, and
very thin breviseptum. Alveolus not well preserved.
Interior surface marked by numerous, small endospines.
Muscle scars not seen.
Remarks
Acanthatia is characterized by its semicircular outline,
lack of costation, lack of spine ridges and distinct interarea
in each valve. The finely punctate interior surfaces are
evidence of endospines, which represent projections of the
taleolae (pseudopunctae) into the body cavity. This genus
occurs throughout both sections in this study. Prosser
(1912) reported Productella hirsuta Hall, 1867 from the
Chagrin Shale at Tinker’s Creek, Cuyahoga County.
Productella is characterized by suberect or recumbent
spines which arise from scattered pustules (Muir- Wood and
Cooper, 1960). The specimens in this study show no
evidence of surface pustules. Further, spines are arranged
in rows, and are not scattered.
Order Rhynchonellida Kuhn, 1949
Superfamily Rhynchonellacea Gray, 1848
Family Trigonirhynchiidae McLaren (in Schmidt and
McLaren), 1965
Genus Centrorhynchus Sartenaer, 1970
Centrorhynchus sp.
Figures 9.1-9.12
Description of material
Small size, biconvex, rounded outline; length
approximately equal to width; greatest height at midvalve.
Ornament of simple, sharply angular costae, which arise in
umbonal regions, and terminate at anterior margins. Costae
internally thickened in axial regions, leaving a more
rounded impression on interior molds. General costal
formula 3/4; 1-1/0-0; 7+/8+ (Figure 9.1). Pedicle valve
umbo inflated, beak slightly incurved over interarea.
Pedicle valve sinus wide, beginning about midvalve, and
18
SCHWIMMER AND FELDMANN
No. 45
FIGURE 9. Centrorhynchus sp. la-e, ventral, dorsal, lateral, posterior, and anterior views, Brecksville Reservation, interval 28, CMNH
8441 ; 2a, b, interior mold of brachial valve and latex mold of same specimen, Brandywine Falls, interval 39, CMNH 8442 ; 3, posterior
view, Brecksville Reservation, interval 28, CMNH 8443; 4a-d, posterior, dorsal, lateral and ventral views, Brecksville Reservation,
interval 28, CMNH 8444 ; 5, posterior view, pedicle foramen, Brecksville Reservation, float, CMNH 8445; 6, crenulate anterior margin,
Brecksville Reservation, interval 28 ,CMNH 8446; 7a, b, pedicle and brachial valve interiors, Brecksville Reservation, interval 5, CMNH
8447; 8, brachial valve interior, Brandywine Falls, float, CMNH 8448; 9, mold of pedicle valve, Brandywine Falls, float, CMNH 8449;
10, mold of pedicle valve, Brandywine Falls, float, CMNH 8450; 11, interior mold of brachial valve, Brandywine Falls, float, CMNH
8451 ; 12, interior molds of pedicle and brachial valves, Brandywine Falls, float, CMNH 8452. Bar scale = 1 cm.
extending to anterior margin. Tongue well-defined.
Anterior margin crenulate (Figure 9.6). Pedicle valve
interior with short dental plates. Brachial valve interior
with covered, U-shaped septalium, strongly arched
anteriorly (Figure 9.2). Septum stout, extending
approximately one half shell length.
Remarks
Angular costae, a distinctive general costal formula, a
robust septalium and cover, generally rounded outline,
inflated pedicle valve, and suberect pedicle valve beak are
diagnostic characters of the genus Centrorhynchus.
Sartenaer (1970) pointed out that this Famennian genus
appears in New York and Pennsylvania. Identification of
the Chagrin Shale specimens would be reinforced if
specimens in better condition could be recovered and
serially sectioned. Distribution is shown in Figures 2 and 3.
This rynchonellid is the most numerous and persistent
taxon found in the study area.
Family Camarotoechiidae Schuchert, 1929
Subfamily Camarotoechiinae Schuchert, 1929
Genus Leiorhynchus Hall, 1860
“Leiorhynchus" sp.
Figures 10.1-10.9
1990
Brachiopods and Bivalves in the Chagrin Shale
19
FIGURE 10. “Leiorhynchus” sp. 1, interior mold of brachial valve, Brecksville Reservation, interval 5, CMNH 8453; 2a, b, interior of
pedicle valve umbo and interior mold of pedicle valve and brachial valve, Brecksville Reservation, interval 5, CMNH 8454 ; 3, interior,
brachial valve umbo, Brandywine Falls, float, CMNH 8455; 4. posterior view, Brecksville Reservation, interval 5, CMNH 8456; 5,
antero-lateral mold of pedicle valve, Brecksville Reservation, float, CMNH 8457; 6, mold of posterior, Brandywine Falls, float, CMNH
8458; 7, interior mold of brachial valve, septum and muscle scars, Brecksville Reservation, interval 2, CMNH 8459; 8a, b, internal mold
of brachial valve and latex mold of same specimen, Brecksville Reservation, interval 5, CMNH 8460; 9, interior mold of brachial valve,
Brandywine Falls, float, CMNH 8461 Bar scale = / cm.
Description of material
Medium size, biconvex, globular; outline circular.
Pedicle valve beak strongly incurved and touching brachial
valve. Pedicle foramen, if present, very small, and rounded.
Pedicle sinus and brachial fold are weak and wide,
originating from about midlength of the valve. The tongue is
not visible. Surface ornament of a few coarse median costae
which originate in umbonal regions and extend to anterior
commissure. Brachial valve fold with at least five costae,
pedicle valve sulcus with fewer. Costae on lateral slopes
weak or absent. Pedicle valve interior with median septum
extending one third shell length. Shell appears thickened in
posterior region. Pedicle valve interior with short, divergent
dental plates and robust teeth. Brachial valve interior with
amphora-shaped septalium, with deep umbonal cavities on
either side. Septum extends at least one half shell length.
Remarks
Prosser (1912) reported three species of Liorhynchus
(= Leiorhynchus ) from the Chagrin Shale of Lake County.
Although Leiorhynchus is not considered a Famennian genus
(Sartenaer 1967), preservation in the Chagrin specimens is
too poor and features are too indistinct for a more concrete
diagnosis. No complete, articulated specimens were found,
and the general costal formula could not be determined. This
taxon is fairly abundant at Brecksville Reservation,
occurring throughout the section (Figure 2). On the other
hand, at Brandywine Falls, “ Leiorhynchus ” was recovered
from only four intervals (Figure 3). It is possible that this
taxon appeared more frequently, but was unrecognizable due
to conditions of preservation.
Order Spiriferida Waagen, 1883
Suborder Athyrididina Boucot, Johnson, and Staton, 1964
Superfamily Athyridacea M’Coy, 1844
Family Athyrididae M'Coy, 1844
Subfamily Athyridinae M’Coy, 1844
Genus Athyris M'Coy, 1844
Athyris sp.
Figures 11.1-5, 1 1 .7
Description of material
Small to medium size, subcircular, biconvex, inflated in
umbonal areas, flattened on lateral slopes and around
anterior margin. Hinge length less than greatest width, which
occurs at midvalve. Pedicle beak extends over hinge, and
incurves slightly; foramen rounded, moderately large.
Pedicle valve weakly sulcate, from about midvalve to
anterior margin; ornament of concentric growth lines, some
lamellose. Pedicle interior with divergent dental plates
extending approximately one third shell length; spiralium
impressed upon interior mold of one specimen (Figure 1 1.7).
20
SCHWIMMER AND FELDMANN
No. 45
FIGURE 11. Athyris-Composita. 1-5,7, Athyris sp. 1, pedicle valve exterior, dental plate impressions, Brecksville Reservation, interval
18, CMNH 8462 ; 2, interior mold of brachial valve, median septum and elongate muscle scar, Brandywine Falls, interval 59, CMNH
8463; 3, pedicle valve interior, dental plates, Brandywine Falls, float, CMNH 8464 ; 4, mold of posterior umbones, Brandywine Falls,
float, CMNH 8465; 5, mold of brachial valve interior, median septum and elongate muscle scar, Brandywine Fads, float, CMNH 8466; 7,
interior mold of spiralium, Brandywine Falls, float , CMNH 8468; 6,8-11, Composita, sp. 6, brachial valve exterior, Brecksville
Reservation, interval 29, CMNH 8467; 8, pedicle valve interior, dental plates, chordate muscle scar, Brecksville Reservation, interval 28,
CMNH 8469; 9a-c, dorsal, ventral, lateral interior molds. Brandywine Falls, float, CMNH 8470; 10, pedicle valve exterior, dental plates,
Brecksville Reservation, interval 29, CMNH 8471; 11, interior mold of pedicle valve, chordate muscle scar, Brandywine Falls, float,
CMNH 8472. Bar scale = 1 cm.
Brachial valve interior with median septum extending half of
shell length, bounded by thin, elongate muscle scars. Tooth
sockets elongate.
Remarks
Athyris and Composita are members of the same
subfamily, and externally, resemble one another. The former
genus is more flattened laterally, and less convex overall
than the latter. Differences between the two genera can be
found internally, especially in the length and angle of
divergence of the dental plates. Unfortunately, in the field, it
is rare to be able to distinguish between the two genera,
when weathered molds or fragments are the only material
available. For this reason, Athyris and Composita are
considered together when plotting occurrence at the two
study sites (Figures 2 and 3). Both appear to be fairly
abundant throughout the section at Brecksville Reservation.
At Brandywine Falls, both are found in the lower half of the
unit, and occur again sporadically in the upper third.
Genus Composita Brown, 1849
Composita sp.
Figures 11.6, 11.8-11.11
Description of material
Small to medium size, ovate, biconvex, length greater
than width; maximum width at midvalve. Hinge line short.
Ornament of concentric growth lines, becoming somewhat
lamellose anteriorly; faint radial striations originate at beak
and extend to margins. Pedicle valve with poorly defined
fold; anterior margin appears uniplicate. Pedicle beak
extends over hinge line; foramen rounded, open. Pedicle
valve interior strong teeth supported by divergent dental
plates, which extend about one fourth sheil length,
recurving slightly at the anterior ends. Muscle scars cordate
(Figure 11.11). Brachial valve interior with median septum
originating at beak and extending one half shell length,
bounded by thin, elongate muscle scars (Figure 1 1 ,9a).
Remarks
Composita closely resembles Athyris, as pointed out in
remarks about the latter taxon.
1990
Brachiopods and Bivalves in the Chagrin Shale
21
FIGURE 12. Spinospirifer and Ambocoelia. 1-4, Spinospirifer. 1, pedicle valve mold, Brandywine Falls, interval 73, CMNH 8488; 2,
interior mold of pedicle valve with punctae, Brandywine Falls, interval 59, CMNH 8473 ; 3, interior mold of pedicle valve, punctae,
Brecksville Reservation, interval 3, CMNH 8474; 4, pedicle valve mold, Brandywine Falls, interval 43, CMNH 8475; 5-9, Ambocoelia. 5,
latex mold of pedicle valve, Brecksville Reservation, interval 19, CMNH 8476 ; 6, interior mold of brachial valve, Brandywine Falls,
interval 68, CMNH 8477; 7, interior mold of pedicle valve, Brandywine Falls, interval 58, CMNH 8478; 8, interior mold of brachial
valve, Brecksville Reservation, interval 19, CMNH 8479; 9, interior mold of pedicle valve, and brachial valve interior, Brecksville
Reservation, interval 5, CMNH 8480. Bar scale = 1 cm.
Suborder Spiriferidina Waagen, 1883
Superfamily Cyrtiacea Fredericks, (1919) 1924
Family Ambocoeliidae George, 1931
Genus Ambocoelia Hall, 1860
Ambocoelia sp.
Figures 12.5-12.9
Description of material
Small size, pedicle valve strongly convex, brachial valve
weakly convex. Trigonal to subtrigonal outline; length greater
than width; hinge line less than greatest width, which occurs at
midvalve. Interareas not visible. Surface ornament of
concentric growth lamellae, especially evident in the anterior
part of the shell, as well as concentrically arranged spine rows.
Pedicle valve with deep, narrow sulcus originating at beak and
extending to rectimarginate anterior commissure. Posterior
two-thirds of pedicle valve, and umbonal region, inflated;
anterior third of valve, and lateral flanks, flattened. Brachial
valve circular, with shallow sulcus extending from umbonal
region to anterior margin; slightly inflated in posterior half of
umbonal region; anterior half of valve, and flanks, flattened.
Brachial valve interior with short divergent crura which are
attached to socket ridges by hinge plates. Cardinal process
small, triangular. Cardinal extremities rounded.
Remarks
A distinctive quadrate muscle pattern, described by
Hall (1860), is characteristic of Ambocoelia, as well as a
deeply triangular pedicle interarea. Very few specimens of
this taxon were collected in the study area; unfortunately,
neither of these features was exposed in the specimens
collected, nor was the brachial valve muscle pattern. The
taxon is common in the lower half of the section at
Brecksville, but is sparse in the upper third. Distribution at
Brandywine Falls is limited to the lower and upper thirds
of the section (Figures 2 and 3).
Superfamily Spiriferacea King, 1846
Family Mucrospiriferidae Pitrat, 1965
Genus Spinospirifer Martynova, 1961
Spinospirifer sp.
Figures 12.1-12.4
Description of material
Transverse, rhomboidal shells. Ornament of simple
plications; shell spinose. Pedicle valve with sulcus defined
by one larger plication on either side, and at least three
smaller plications on the sulcus. Pedicle valve interior with
short, divergent dental plates. Brachial valve, interareas,
and other internal structures unknown.
Remarks
A few specimens of this taxon were found to be
spinose upon microscopic examination. Material was
scarce, and consisted mostly of fragments, making
identification difficult. It is possible that this taxon may be
22
SCHWIMMER AND FELDMANN
No. 45
FIGURE 13. Cyrtospirifer leboeufensis. 1, partially exfoliated pedicle valve, Brandywine Falls, float, CMNH 8481; 2, interior mold of
pedicle valve, Brecksville Reservation, interval 11 , CMNH 8482; 3a,b, interior mold ofspiralium and counterpart, Brandywine Falls, interval
45, CMNH 8483; 4, posterior view, pedicle valve interarea, upper half of delthyrium covered, Brandywine Falls, float, CMNH 8484; 5,
posterior view, interarea, Brecksville Reservation, interval 24, CMNH 8485; 6, pedicle valve interior, convergent dental plates, flahellate
muscle scar, Brandywine Falls, float, CMNH 8486; 7, pedicle valve interior, Brandywine Falls, float, CMNH 8487. Bar scale - 1 cm.
FIGURE 14. Cyrtospirifer spicatus. 1, mold of pedicle valve exterior, elongate mucrons, Brandywine Falls, float, CMNH 8489; 2, partial
mold of pedicle valve interior, divergent dented plates and spatulate muscle scar, Brandywine Falls, float, CMNH 8490; 3, pedicle valve
interior, dental plates diverging at 45 degree angle, Brecksville Reservation, interval 28, CMNH 8491 ; 4, interior mold of pedicle valve,
divergent dental plates, spatulate muscle scar, Brandywine Falls, float, CMNH 8492; 5, interior mold of pedicle valve, spiralium
impression (arrow), Brandywine Falls, float, CMNH 8493; 6, posterior view of narrow interarea, Brandywine Falls, float, CMNH 8494.
Bar scale = 1 cm.
1990
Brachiopods and Bivalves in the Chagrin Shale
23
FIGURE 15. Sphenospira alta. I. interior mold of pedicle valve interaea with horizontal and vertical striations, delthyrium, stegidium,
Brecksville Reservation, interval 21 , CMNH 8495; 2, brachial valve exterior, Brecksville Reservation, interval 8, CMNH 8496; 3, interior
mold of pedicle valve interarea, stegidium, Brecksville Reservation, interval 8, CMNH 8497; 4, interior mold of pedicle valve interarea,
stegidium. Brecksville Reservation, interval 28, CMNH 8498. Bar scale = I cm.
more abundant in the study area than reported, but due to
the paucity of well-preserved material from the unit, may
not have been identifiable. Partial molds and fragments are
particularly difficult to identify.
Family Cyrtospriferidae Termier and Termier, 1949
Genus Cyrtospirifer Nalivkin (in Fredericks, 1924)
Cyrtospirifer leboeufensis Greiner, 1957
Figures 13.1-13.7
Description of material
Medium to large size; generally transverse;
subpentagonal profile; unevenly biconvex, with pedicle
valve slightly more inflated. Hinge line straight, equal to or
slightly less than greatest width of shell; interareas
moderately high, vertically striated; brachial interarea
anacline, pedicle interarea apsacline; large, open, triangular
delthyrium. Valve exteriors ornamented with low, rounded,
costae which originate at beak and extend to anterior
margin; lateral costae simple. Costae on fold and sulcus
increase by bifurcation. External lamellae common on
anterior half of shell. Pedicle valve exterior with greatest
convexity in area of umbo; beak projects over hingeline;
long, V-shaped sulcus is moderately deep on the posterior
half of the valve, becoming shallow and even obsolete for
the anterior half. Anterior margin gently uniplicate.
Brachial valve exterior with a sharply defined fold
originating at the beak and terminating at the anterior
margin. Pedicle valve interior with pair of dental plates
extending from the beak, initially diverging at a wide angle
then recurving around the anterior end of the spatulate
diductor muscle scar, which bears radial striations on the
anterior two-thirds of its length. The adductor muscle scars
are seen on a slender median groove, located between two
low ridges on the diductor scar, which extends two-thirds
the length of the shell. Spiral brachidia impressed upon the
internal mold of the lateral mantle cavity in some
specimens. The brachial valve interior not observed.
Remarks
Cyrtospirifer is particularly abundant and well
preserved in the Chagrin Shale, and often original shell
material can be recovered. The preservation of delicate
dental plates suggests moderate to rapid burial, before
disarticulation and breakage could occur. Greiner (1957)
noted in C. leboeufensis that the pedicle valve dental plates
join, forming a low ridge around the anterior portion of the
diductor muscle scar. This ridge is not prominent in the
Chagrin specimens. He further stated that this species is
distinguished by its size, shape, and distinct, elongated
muscle scar, which is partly enclosed by the dental plates
(Figure 13.7). Prosser (1912) reported Cyrtospirifer from
Brecksville Reservation and from Brandywine Falls.
24
SCHWIMMER AND FELDMANN
No. 45
FIGURE 16. Toryniferella sp. la,b, interior mold of pedicle valve, median septum and dental plates, with enlargement of double-barreled
spines, Brandywine Falls, interval 69, CMNH 8499; 2, posterior view of pedicle valve beak and interarea, Brandywine Falls, interval 57,
CMNH 8500; 3, interior mold of pedicle valve. Brandywine Falls, float, CMNH 8501; 4, interior mold of pedicle valve, Brecksville
Reservation, interval 19, CMNH 8502; 5, interior mold of pedicle valve. Brandywine Falls, interval 69, CMNH 8503; 6, interior mold of
brachial valve, Brandywine Falls, float, CMNH 8504 ; 7, interior mold of brachial valve, weak fold. Brandywine Falls, interval 43,
CMNH 8405. Bar scale = 1 cm.
Cyrtospirifer spicatus Greiner, 1957
Figures 14.1-14.6
Description of material
Medium to large shell, extremely transverse; biconvex;
hinge length increased by long, attenuated mucrons. Fold and
sulcus fairly well-defined from beak to anterior margin.
Interareas, costae, and other exterior features similar to C.
leboeufensis. Pedicle valve interior distinguished by short
dental plates, which extend one-third the length of the shell
and diverge at a 45 degree angle. Spatulate muscle scar not
enclosed by dental plates.
Remarks
Distinguishing characters for this species include faint
growth lines as micro-ornament, the extremely transverse
profile with exaggerated mucrons, the short, diverging dental
plates, and the shape of the diductor muscle scar (Greiner,
1957). Cyrtospirifer leboeufensis and C. spicatus are for the
most part indistinguishable in the field at either Brecksville or
Brandywine Falls. Often, only fragments or partial specimens
are recovered, and identifications to the species level are not
possible. Therefore, in the brachiopod distribution data (Figures
2 and 3), the species are combined and charted together, at the
generic level. Cyrtospirifer exhibits almost continuous
occurrence in both sections studied. A few intervals in the upper
third of the Brandywine Falls section lack species of this genus.
Genus Sphenospira Cooper, 1954
Sphenospira alta (Hall, 1867)
Figures 15.1-15.4
Description of material
Large size, triangular outline; valves unevenly
biconvex; pedicle valve pyramidal, brachial valve shallow;
greatest width at hinge; pedicle valve interarea finely
striated with vertical and faint horizontal lines; posterior
two-thirds of delthyrium covered by delthyrial plate, while
stegidium covers anterior one-third, growing in increments
as pedicle atrophies (Cooper, 1954). Lateral slopes, fold and
sulcus are costate; costae originate at beak and terminate at
anterior margin, increasing by bifurcation on lateral slopes.
Pedicle and brachial valve interiors not observed.
Remarks
Prosser (1912) reported this species from Chippewa
Creek at Brecksville Reservation. This species was also
collected and described from Chippewa Creek by Cooper
in 1954. Specimens in the National Museum of Natural
History, Washington exhibit divergent dental plates in the
interior of the pedicle valve, as well as radiating muscle
scars between the plates. Museum specimens also show the
cardinal process, teeth, and sockets. Sphenospira alta
occurs, in low abundance, in the lower half of the section at
Brecksville. Only two occurrences in the upper half were
1990
Brachiopods and Bivalves in hie Chagrin Shale
25
recorded. Two specimens were found in situ in the lower
half of the Brandywine Falls area, although several very
large specimens were recovered from float in the creek.
Occurrences are charted in Figures 2 and 3.
Superfamily Reticulariacea Waagen, 1883
Family Elythidae Fredericks, 1924
Genus Toryniferella Weyer, 1967
Toryniferella sp.
Figures 16.1-16.7
Description of material
Medium to large size (width greater than 25 mm in
adults); transversely elliptical outline; biconvex; rounded
margins; width greater than length; hinge length less than
greatest width, which occurs at mid-valve; low interareas;
open delthyrium. Valve exteriors ornamented by prominent
concentric growth lines, each bearing numerous, regularly
spaced, fine, double-barreled spines (Figure 16.1). Pedicle
valve exterior with high umbonal region, and small, pointed,
well-defined beak, which incurves over interarea; sulcus
shallow or obsolete. Brachial valve exterior with low fold, set
apart by a shallow groove on either side; umbo elevated,
broad. Pedicle valve interior displays long, median septum,
originating at umbo, extending approximately two-thirds of
valve length. Median septum flanked by initially diverging
dental plates, which extend one-half valve length, and then
recurve slightly toward septum. Brachial valve interior with a
thin, median ridge which begins approximately one-fourth
the shell length from the posterior margin, and extends one-
half of valve length. Two thin, short, parallel crura extend
from hinge, tenninating before median ridge begins. Crural
plates form single plate across inner surface of crura, but do
not attach to the valve floor.
Remarks
Toryniferella is easily distinguished from Reticularia
praematura (Hall, 1866), which was reported in the Chagrin
by Prosser (1912), by the presence of biramous, or double-
barreled spines in the former genus, and uniramous spines in
the latter genus. Toryniferella can be differentiated from
Torynifer based on brachial valve morphology. Torynifer
possesses a median septum in the brachial valve, which
supports the cardinalia (Carter, 1988). The Chagrin specimens
have no brachial median septum. Toryniferella is distinguished
from Kitakamithyris, another elythid, by the presence in the
former of crural plates which are not separated or divided, and
which do not connect to the floor of the shell (Weyer, 1967).
Specimens of Toryniferella in the Chagrin Shale are
generally incomplete. Original shell material is almost
absent, and dorsal (brachial) beaks are often broken off. At
this time, identification below the genus level is not possible.
Toryniferella is a relatively recently named genus
(Weyer, 1967), and has not been reported widely in the
literature. It is possible that reclassification of some
specimens of Torynifer spp. and Reticularia praematura
(Hall, 1866), the latter species having been reported from
the Chagrin by Prosser (1912), may result in further
additions of members to the genus. Toryniferella occurs in
abundance in the lower half of the section at Brandywine
Creek, and again in the upper third of the same section. In
contrast, at Brecksville, this genus is sparse, occurring in
four intervals in the lower half of the section, and in five
intervals in the upper third (Figures 2 and 3).
Phylum Mollusca Cuvier, 1797
Remarks
The Chagrin megafauna in the Cuyahoga River Valley
is almost exclusively restricted to brachiopods. The only
other faunal elements, with the exception of two
unidentifable, pyritized, epibionts, are four taxa of
bivalves. Their distribution is plotted in Figures 2 and 3.
Class Bivalvia Linne, 1758 (Buonanni, 1681)
Subclass Pteriomorphia Beurlen, 1944
Order Pterioida Newell, 1965
Suborder Pteriina Newell, 1965
Superfamily Pteriacea Gray, 1847 (1820)
Family Pterineidae Miller, 1877
Genus Leptodesma Hall, 1883
Leptodesma sp.
Figures 17.6, 17.8-17.10
Description of material
Small to medium size, biconvex. Hinge line straight
from anterior extremity to posterior wing. Posterior wing
terminates as a short spine. Surface ornament of fine to
coarse, irregularly spaced concentric lines. Dentition,
hinge, ligament and internal structures unknown. No shell
material is preserved.
Remarks
Very small specimens of Leptodesma are often found
concentrated on bedding planes in the Chagrin Shale.
These could represent spatfalls, or juvenile specimens,
which were buried before growth could proceed. This
taxon is found at both locations, mainly in the lower halves
of each section (Figures 2 and 3).
Superfamily Pectinacea Rafinesque, 1815
Family Pterinopectinidae Newell, 1938
Genus Pterinopecten Hall, 1883
Pterinopecten? sp.
Figures 17.1-17.4
26
SCHWIMMER AND FELDMANN
No. 45
FIGURE 17. Bivalvia. 1-4, Pterinopecten? sp. 1, interior mold, Brandywine Falls, float, CMNH 8506; 2, interior mold, Brandywine Falls,
interval 58, CMNH 8507; 3, mold of exterior, with reticulate pattern formed by costa and growth lines, Brandywine Falls, interval 58,
CMNH 8508; 4, exterior mold, subequal auricles, Brandywine Falls, interval 74, CMNH 8509; 5, Sanguinolites sp., interior mold of right
valve, growth lines increase by intercalation and bifurcation (arrow), Brandywine Falls, float, CMNH 8510; 7, Spathella? sp., interior
mold, right valve, Brandywine Falls, float, CMNH 8512; 6, 8-10, Leptodesma sp. 6, interior mold of left valve, with posterior spine,
Brecksville Reservation, interval 25, CMNH 8511 ; 8, partial interior mold of left valve, Brandywine Falls, interval 62, CMNH 8513; 9,
interior mold of disarticulated specimen, Brandywine Falls, interval 58, CMNH 8514; 10, molds of juvenile specimens, Brandywine Falls,
float. CMNH 8515. Bar scale = 1 cm.
Description of material
Small to medium size, generally flattened. Pectinid
shell shape with subequal anterior and posterior ears.
Hinge line straight, extending length of ears. Surface
ornament of prominent radial ribs, crossed by concentric
growth lines, resulting in a reticulated appearance. Shell
material, dentition, hinge features, ligament and
musculature unknown.
Remarks
Specimens were all collected from the Brandywine Falls
location, either as float or in situ (Figure 3). No specimen is
complete; information was obtained from partial molds.
Order Modiomorphoida Newell, 1969
Superfamily Modiomorphacea S.A. Miller, 1877
Family Modiomorphidae S.A. Miller, 1877
Genus Spathella Hall, 1885
Spathella? sp.
Figure 17.7
Description of material
One medium size specimen was collected as float from
a slab at Brandywine Falls. Right valve ovate, ornamented
by evenly spaced, concentric growth lines. Beak
prosogyrous. One shard of shell material remains on the
specimen. Lunule and escutcheon, ligament, hinge, and
musculature unknown.
Subclass Anomalodesmata Dali, 1889
Order Pholadomyoida Newell, 1965
Superfamily Pholadomyacea Gray, 1847
Family Grammysiidae S.A. Miller, 1877
Genus Sanguinolites M'Coy, 1844
Sanguinolites sp.
Figure 17.5
1990
Brachiopods and Bivalves in the Chagrin Shale
27
Description of material
One specimen was recovered from float at Brandywine
Falls. Specimen large, elongate, with small umbones
located in anterior third of shell. Beak prosogyrous.
Ornament of very coarse comarginal rugae, which increase
by bifurcation and intercalation (Figure 17.5).
Brachiopod distribution
The total number of genera and their distribution
throughout both stratigraphic sections shows that
brachiopods are fairly abundant in the lower one half to
two thirds of each section (Figure 18). This is followed by
a sharp drop in the number of genera to two at Brecksville
and to zero at Brandywine. An especially silty interval is
present at this position at Brandywine, and could account
for the decrease in taxa, and for their inability to recolonize
during the time of silt deposition. Silt was transported into
the area by periodic storms (Weidner and Feldmann, 1983),
and tended to obliterate brachiopod assemblages with rapid
burial. There is also a possibility that turbidites were the
dominant mechanism of rapid sedimentation. Lewis (1988,
p. 24) evaluated different studies, and favors a storm
mechanism and deposition of silts in the form of
tempestites. After intervals of low abundance, abundance
increases in the upper third of each section. Diversity is
never as great as in the lower parts of both sections (Figure
18). Diversity declines near the contact with the Cleveland
Shale, concommittant with a change to anoxic conditions.
Brachiopods such as Cyrtospirifer and Centrorhynchus
are well distributed throughout both sections (Figures 2 and
3). This supports a conclusion that these two taxa were
eurytopic, and represented pioneer species. Cyrtospirifer
possessed a complex, spirolophous lophophore, which
enhanced food and oxygen gathering capabilities.
Centrorhynchus also had a spirolophous lophophore, but its
resurgence after periodic disasters might possibly be due to
an initial, epiplanktonic mode of existence, analogous to that
suggested by Ager (1962; 1965). Some genera, such as
Sphenospira and Aulacella, are abundant at one site and
sparse at the other, while the faunal distributions of other
genera, such as Retichonetes , Acanthatia, and
Schellwienella, appear similar in both sections.
The composition and vertical distribution of the
Chagrin fauna reflects colonization by opportunistic,
pioneering species. Levinton (1970) distinguished between
equilibrium species populations and opportunistic species
populations. The former are resource-limited, or near the
carrying capacity of the environment, which is stable and
less likely to fluctuate over time. As a result, population
levels tend to remain constant. Opportunistic species, on
the other hand, are not resource limited, and increase
rapidly in numbers. Under unstable environmental
conditions, these species are considered physiological
generalists; they tend to increase in numbers when space,
temperature, salinity and other environmental factors
become favorable.
Levinton (1970) offered several criteria for identifying
opportunistic species. First, there is random orientation of
specimens and lack of size sorting in individual beds,
although dominant taxa tend to be grouped by size. In the
Chagrin, this is especially true of the cyrtospiriferids and
the rhynchonellids, which are numerically dominant and
ubiquitous at the two sections.
Second, distribution over a limited area is
characteristic, with adjacent, nonfossiliferous horizons.
Within the study area, some layers are fossiliferous, while
adjacent layers and areas may be fossil poor. The pattern of
NUMBER OF GENERA
0 6 12 0 6 12
BRANDYWINE
Figure 18. Vertical distribution of taxa in the Cuyahoga
Valley. The number of taxa within each 20 cm interval is
plotted on a scale of zero to twelve . Both sections exhibit
abundant taxa in the lower halves, a sharp drop in abundance
in the upper one half to one third, and an increase in
abundance in the upper portion. Taxa decrease again near the
Chagrin-Cleveland boundary.
28
SCHWIMMER AND FELDMANN
No. 45
distribution appears random. On a regional scale, very few
Chagrin Shale outcrops in northeast Ohio contain the
abundant brachiopod fauna found in the Cuyahoga River
Valley. Adverse environmental conditions or lack of
colonizing stock may account for this condition, although
differential preservation must also be considered.
Third, stable faunal assemblages may be invaded by
one dominant opportunistic species, leading to
overwhelming domination of the assemblages (85-100%).
This does not appear to be the case with the Chagrin fauna.
Stable faunal assemblages never seemed to develop. After
periods of rapid silt deposition, many of the species which
were able to recolonize appear to be opportunists, with
special morphologies for coping with stressed
environments. New taxa did not appear in the section over
time, possibly because the initial colonists had not
sufficiently stabilized the substrate, and because the
physical environment remained changeable. Most Chagrin
brachiopods represent pioneer species, never rising much
past an entry level for community succession. The same
faunal elements continue to reestablish themselves after
storm events. Very few taxa show distribution patterns
which are anomalous in that they do not reflect pioneering
characteristics throughout both sections (Figures 2 and 3).
Aulacella is reported only once from the upper two thirds
of the Brandywine Falls section, although it appears
throughout the Brecksville section. Sphenospira is present
in the lower half of Brecksville, and in two intervals in the
upper half of that section, and in two intervals in the lower
half of Brandywine Falls section. Schellwienella occurs in
the lower halves of both sections, and once in the upper
parts of both sections. The lingulids show the same general
distribution pattern as Schellwienella. Orhiculoidea is
found only at Brecksville and is not considered in this
analysis. The other Chagrin brachiopods remain fairly
constant in their distributions throughout the sections.
Pioneer assemblages do not appear to have been replaced
by more stable aggregations. Storm events were too
frequent to allow for an orderly progression of faunal
elements and community succession over time.
Alexander (1977) studied opportunistic brachiopods in
Idaho and Utah. Many developed compressed body plans
for suspension on a fluid substrate, which also served to
increase oxygen diffusing mantle surfaces under oxygen
deficient conditions. According to Alexander (1977),
stressed communities are characterized by one dominant
type of opportunist which can adapt to adverse conditions;
by an absence of rugose and colonial corals, sponges, and
bryozoans, which thrive in clear, oxygenated, less turbid
waters; and by a variety of vacant niches, as evidenced by a
lack of mobile faunal elements. The Chagrin environment
was probably stressed, and faunal diversity is low,
consisting of sessile organisms which can float in, or on, an
unstable, muddy substrate, and which can feed and breathe
under low oxygen, turbid conditions (Schwimmer, 1988).
In this respect, the Chagrin brachiopods are opportunists,
which exploited stressed habitats, and managed to establish
and maintain pioneer communities.
The brachiopods collected in the Cuyahoga Valley
Chagrin sequence provide the best evidence for a
paleoenvironmental interpretation of the Chagrin Shale,
due to their almost exclusive occurrence in the sediments at
both locales. Their morphological adaptations and inferred
life styles enhance our understanding of conditions at the
time of deposition. Further, their distribution patterns at the
two sites may aid in correlation of the two sections. It is
anticipated that detailed collection at sites in the central
and western portions of the Chagrin outcrop belt would
yield similar brachiopod faunas and distributions.
Age of the Chagrin Shale
On the evidence of its conodont and brachiopod
faunas, the Chagrin Shale is Late Devonian (Famennian) in
age. Schopf and Schwieterling (1970) and Murphy (1973)
used the presence of the alga Foerstia, obtained from
locales in Ashtabula County, to correlate the Chagrin with
other Upper Devonian units in northwestern Pennsylvania
and southwestern New York. Murphy (1973) reported
Foerstia near the base of the Ellicott Shale Member of the
Chadakoin Formation in Pennsylvania and New York. This
formation is included in the Conneaut Group of the upper
Cassadagan Stage. Schopf (in Feldmann et al., 1978)
commented that the Foerstia zone probably was within the
Upper Cassadagan Stage. Feldmann et al. (1978)
speculated that the Chagrin could extend into the
Bradfordian Stage (Fammenian).
Brachiopods found in the Chagrin Shale support a
Famennian age. Dutro (1981) reported that Cyrtospirifer
spicatus Greiner and Sphenospira alta (Hall) occur in the
Venango and Cattaraugus formations, assigned to the
Famennian Conewangoan Stage of Cooper et al. (1942),
equivalent to the Bradfordian Stage of Rickard (1975).
The most definitive age data for the Chagrin
Formation come from a recent conodont analysis, reported
by Zagger and Banks (1988), from the Skinner's Run
pyrite bed, mentioned earlier. Anita Harris (U.S.
Geological Survey, 1988, pers. comm.) identified
specimens of Polygnathus exerplexus Sandberg and
Zeigler, 1979 and Bispathodus aculeatus (Branson and
Mehl, 1934) from the Zagger and Banks (1988) collection.
These indicate correlation with the Middle expansa
subzone (Sandberg and Zeigler, 1984) of late Famennian
age. Polygnathus exerplexus is restricted to the Lower to
Middle expansa Zone, and Bispathodus aculeatus,
although with a greater stratigraphic range, first occurs in
the Middle expansa subzone (Harris, 1988, pers. comm.).
1990
Brachiopods and Bivalves in the Chagrin Shale
29
Conclusions
Nineteen taxa of brachiopods and four taxa of bivalves
are described here from the Upper Devonian Chagrin Shale
of the Cuyahoga River Valley in northeast Ohio. The
Chagrin sediments were deposited on a relatively shallow
marine shelf, below normal wave base. Episodic storms are
thought to have carried sediments westward from the
prograding Catskill Delta complex, leading to the periodic
burial of brachiopod assemblages. Brachiopods are
generally preserved as molds, with occasional shell
material. Some molds are coated with pyrite. Pyrite also
occurs in sediments as nodules or burrow infillings.
Lingulid brachiopods are preserved in phosphatic
concretions. The infrequent fragmentation and abrasion of
specimens suggests rapid burial rates. The composition and
distribution patterns of brachiopod assemblages implies
population by pioneering species, which were able to
survive under adverse conditions. Stressed conditions
precluded colonization by stenotopic faunal elements.
Brachiopod distribution patterns at the two sites suggest a
rough correspondence in assemblage composition.
Conodont evidence provides an Upper Famennian age for
the Chagrin, as does the occurrence of the brachiopod
Sphenospira alta.
Acknowledgements
We wish to acknowledge the Cleveland Metroparks
and The National Park Service for granting collecting
permits. Joseph Hannibal, The Cleveland Museum of
Natural History, made the Museum collections available to
us during this project, and made the EDX analysis by Scott
Pluim, Tulsa, available. Lisa Stillings, formerly of Kent
State University, performed phosphorus lab tests. We are
appreciative of the counsel provided by Dr. J. Thomas
Dutro, Jr. of the United States Geological Survey, and Dr.
John L. Carter of the Carnegie Museum. Dr. Jed Day,
Illinois State University, and an anonymous reviewer
reviewed this manuscript and suggested improvements.
Contribution 398, Department of Geology, Kent State
University, Kent, Ohio 44242.
References
Ager, D. V. 1962. The Occurrence of Pedunculate Brachiopods in
Soft Sediments. Geological Magazine 99:184-186.
Ager, D. V. 1965. The Adaptation of Mesozoic Brachiopods
to Different Environments. Palaeogeography,
Palaeoclimatology, Palaeoecology 1:143-172.
Alexander, R. R. 1977. Growth, Morphology and Ecology of
Paleozoic and Mesozoic Opportunistic Species of Brachiopods
from Idaho-Utah. Journal of Paleontology 51:11 33-1 149.
Alichova, T. N. 1960. Otryad Orthida: Mshanki, Brakhiopody [T.
G. Sarycheva, ed.], p. 183-197. In Osnovy Paleontologii ,
edited by Y. Orlov.
Baird, G. C., and C. E. Brett. 1986. Erosion on an Anaerobic
Seafloor: Significance of Reworked Pyrite Deposits from the
Devonian of New York State. Palaeogeography,
Palaeoclimatology, Palaeoecology 57:157-193.
Berner, R. A. 1984. Sedimentary Pyrite Formation: An Update.
Geochimica et Cosmochimica Acta 48:605-615.
Beurlen, K. 1944. Beitrage zur Stammesgeschichte der Muscheln.
Bayerische Akademie der Wissenschaften , Sitzungsberichte,
No. 1-2:133-145.
Branson, E. B., and M. G. Mehl, 1934. Conodonts from the
Grassy Creek Shale of Missouri. In Conodont Studies No. 3:
Missouri University Studies 8(3): 1 7 1 -259.
Brett, C. E., and G. C. Baird. 1986. Comparative Taphonomy: A
Key to Paleoenvironmental Interpretation Based on Fossil
Preservation. Palaios 1:207-227.
Broadhead, R.F., and PE. Potter. 1980. Petrology of the Devonian
Gas-Bearing Shale Along Lake Erie Helps Explain Gas Shows.
United States Department of Energy, Morgantown Energy
Technological Center, Morgantown, West Virginia. 49 p.
Bronn, H.G. 1862. Die Klassen und Ordnungen der Wiechtiere
(Malacozoa) 3(1): 1-518.
Brown, T. 1849. Illustrations of the Fossil Conchology of Great
Britain and Ireland with Descriptions and Localities of all
Species. London, 273 p.
Bruiguiere, J.G. 1797. Encyclopedic Methodique . Tableau
Ency dope clique et Methodique des Trois Regnes de la Nature.
Vers, Coquilles, Mollusques et Polypiers, Vol. 2. pis. 96-134.
Carter, J.L. 1988. Early Mississippian Brachiopods from the Glen
Park Formation of Illinois and Missouri. Bulletin of the
Carnegie Museum of Natural History No. 27, 82 p.
Cooper, G.A. 1954. Unusual Devonian Brachiopods. Journal of
Paleontology 28:325-332.
Cooper, G.A., C. Butts, K.E. Caster, G.H. Chadwick, W.
Goldring, E.M. Kindle, E. Kirk, C.W. Merriam, F.M. Swartz,
PS. Warren, A.S. Warthin, and B. Willard. 1942. Correlation
of the Devonian Sedimentary Formation(s) of North America.
Geological Society of America Bulletin 53:1729-1794.
Cooper, G.A., and J.T. Dutro, Jr. 1982. Devonian Brachiopods of
New Mexico. Bulletins of American Paleontology Vol. 82 and
83, 215 p.
Cushing, H.P., F. Leverett, and F.R. Van Horn. 1931. Geology and
Mineral Resources of the Cleveland District, Ohio. U. S.
Geological Survey Bulletin 818, 138 p.
Cuvier, G. 1797. Tableau Elementaire de T Histoire Naturelle des
Animaux. Paris, 1-710.
Dali, W. H. 1889. Scientific Results of Explorations U.S. Fish
Comm. Albatross. No. VII. Preliminary report on the
collection of Mollusca and Brachiopoda Obtained in 1887-
1888: U.S. National Museum Proceedings 12:217-362.
Dumeril, A.M.C. 1806. Zoologie Analytique on Metlwde
Naturelle de Classification des Animaux. Allais (Paris), 344 p.
Dunbar, C.O., and G.E. Condra. 1932. Brachiopoda of the
Pennsylvanian System of Nebraska. Nebraska Geological
Survey Bulletin, 2nd Series, 51:1 -377.
Dutro, J. T., Jr. 1981. Devonian Brachiopod Biostratigraphy of
New York State, p. 67-82. In Devonian Biostratigraphy of
New York State, edited by W. A. Oliver and G. Klapper.
30
SCHWIMMER AND FELDMANN
No. 45
Feldmann, R. M., R. G. Osgood, Jr., E.J. Szmuc, and D.W.
Meinke. 1978. Chagrinichnites brooksi, a New Trace Fossil of
Arthropod Origin. Journal of Paleontology 52:287-294.
Feldmann, R.M., J.T. Hannibal, and L.E. Babcock. 1986. Fossil
Worms from the Devonian of North America (Sphenothallus)
and Burma (“Vermes”) Previously Identified as Phyllocarid
Crustaceans. Journal of Paleontology 60:34 1 -346.
Fredericks, G.N. 1924. O Verkhe-Kammennougol’nykh
Spiriferidakh Urala. Geologicheskiy Komitet Izvestia
38:295-324.
George, T.N. 1931. Ambocoelia Hall and Certain Other British
Spiriferidae. Geological Society of London, Quarterly
Journal 87:30-61 .
Girty, G. H. 1939. Some Linguloid Shells from the Late Devonian
and Early Carboniferous Rocks of Pennsylvania and Ohio.
U.S. Geological Survey Professional Paper 193-C, p. 47-67.
Gray, J.E. 1840. Synopsis of the Contents of the British Museum.
42nd ed., London, 370 p.
Gray, J.E. 1847. A List of the Genera of Recent Mollusca, Their
Synonyms and Types. Zoological Society of London,
Proceedings 15:129-219.
Gray, J.E. 1848. On the Arrangement of the Brachiopoda. Annals
and Magazine of Natural History, Series 2, 2:435-440.
Greiner, H. 1957. Spirifer disjunctus : Its Evolution and
Paleoecology in the Catskill Delta. Yale University, Peabody
Museum of Natural Histoiy Bulletin 11:1-75.
Hall, J. 1860. Observations of Brachiopoda. New York State
Cabinet Natural History, Annual Report 13:65-75.
Hall, J. 1867. Descriptions and Figures of the Fossil
Brachiopoda of the Upper Helderberg, Hamilton, Portage,
and Chemung Groups. New York Geological Survey,
Palaeontology 4( 1 ): 1 -428 p.
Hall, J. 1883. Lamellibranchiata, Plates and Explanations.
Natural History New York, Paleontology 5(1 ): 1- 20, 79 pi.
Hall, J. 1885. Lamellibranchiata II, Descriptions and Figures of
the Dimyaria of the upper Helderberg, Hamilton, Portage and
Chemung Groups. New York Geological Survey, Paleontology
5( 1 ):269-562.
Hannibal, J.T., and R.M. Feldmann. 1983. Arthropod Trace
Fossils, Interpreted as Echinocarid Escape Burrows from the
Chagrin Shale (Late Devonian) of Ohio. Journal of
Paleontology 57:705-716.
Huxley, T.H. 1869. An Introduction to the Classification of
Animals. London, 147 p.
King, W. 1846. Remarks on Certain Genera Belonging to the
Class Paliobranchiata. Annals and Magazine of Natural
Histoiy 18:26-42, 83-94.
King, W. 1850. A Monograph of the Permian Fossils of England.
London Palaeontographical Society, 258 p.
Kuhn. O. 1949. Lehrbuch der Paldozoologie. Stuttgart, 326 p.
Levinton, J.S. 1970. The Paleoecological Significance of
Opportunistic Species. Lethaia 3:69-78.
Lewis, T.L. 1988. Late Devonian and Early Mississippian Distal
Basin-Marginal Sedimentation of Northern Ohio. The Ohio
Journal of Science 88(l):23-39.
Linne, C. 1758. Systema Naturae, 10th Edition, Tomus 1,
Holmiae, Laurentii Salvii: 1-824.
Manheim, F., G.T. Rowe, and D. Jipa. 1975. Marine Phosphorite
Formation off Peru. Journal of Sedimentary Petrography
45:243-251.
Martynova, M.V. 1961. Stratigrafiia i Brakhiopody Famenskogo
Iarusa Zapadnoi Chasti Tsentral’nogo Kazakhstana. Materialy
po Geologii Tsentral' nogo Kazakhstana 2:1-210.
M’Coy, F. 1844. A Synopsis of the Characters of the
Carboniferous Limestone Fossils of Ireland. Dublin, 207 p.
Menke, C.T. 1828. Synopsis Methodica Molluscorum Generum
Omnium et Specierum Earum Quae in Museo Menkeano
Adservantur. Pyrmonti, 91 p.
Miller, S.A. 1877. The American Paleozoic Fossils. Published by
the Author, Cincinnati, 245 p.
Muir- Wood, H.M. 1955. A History of the Classification of the
Phylum Brachiopoda. British Museum of Natural History,
London, 24 p.
Muir- Wood, H.M. 1962. On the Morphology and Classification
of the Brachiopod Suborder Chonetoidea. British Museum of
Natural History, London, 132 p.
Muir-Wood, H.M., and G.A. Cooper. 1960. Morphology,
Classification and Life Habits of the Productoidea
(Brachiopods). Geological Society of America Memoir 81,
447 p.
Muir-Wood, H.M., and A. Williams. 1965. Strophomenida, p.
H36 1-370. In Treatise on Invertebrate Paleontology, Part H,
Brachiopoda. edited by R. C. Moore. The Geological Society
of America and University of Kansas Press.
Murphy, J.L. 1973. Protosalvinia (Foerstia) Zone in the Upper
Devonian Sequence of Northeastern Ohio, Northwestern
Pennsylvania, and Western New York. Geological Society of
America Bulletin 84:3405-3410.
Newell, N.D. 1938. Late Paleozoic Pelecypods: Pectinacea.
Kansas Geological Survey 10:1-123.
Newell, N.D. 1965. Classification of Bivalvia. American
Museum Novitates, No. 2206, 25 p.
Newell, N.D. 1969. Order Modiomorphoida, p. N393. In Treatise
on Invertebrate Paleontology, Part N, Mollusca 6, Vol. 1 ,
Bivalvia, edited by R. C. Moore. Geological Society of
America and University of Kansas Press.
Opik, A. 1933. Uber einige Dalmanellacea aus Estland.
University of Tartu (Dorpat). Acta et Commentationes, Series
A, 25( 1 ): 1-25.
Opik, A. 1934. Ober Klitamboniten. University of Tartu (Dorpat),
Acta et Commentationes, Series A, 26(3): 1 - 239.
Orbigny, A. d’. 1847. Considerations Zoologique et Geologiques
sur les Brachiopodes ou Palliobranches. Academie des
Sciences Paris, Comptes Rendus 25:193-195, 266-269.
Pepper, J.F., W. De Witt, Jr., and D.F. Demarest. 1954. Geology of
the Bedford Shale and the Berea Sandstone in the Appalachian
Basin. United States Geological Survey Professional Paper
259, 109 p.
Pitrat, C.W. 1965. Family Mucrospiriferidae, p. H686. in Treatise
on Invertebrate Paleontology, Part H, Brachiopoda, edited by
R. C. Moore. Geological Society of America and University of
Kansas Press.
Potter, P.E., J.B. Maynard, and W.A. Pryor. 1980. Final Report of
Special Geological, Geochemical and Petrological Studies of
1990
Brachiopods and Bivalves in the Chagrin Shale
31
the Devonian Shales in the Appalachian Basin. Prepared for
United States Department of Energy Eastern Gas Shales
Project. University of Cincinnati, 86 p.
Prosser, C.S. 1912. The Devonian and Mississippian Formations
of Northeastern Ohio. Geological Survey of Ohio, Bulletin
15, 574 p.
Rafinesque-Schmaltz, C.S. 1815. Analyse de la Nature on Tableau
de I'Univers et des Corps Organises, etc. Palermo, 224 p.
Rickard, L.V. 1975. Correlation of the Devonian Rocks in New
York. New York State Museum and Science Service Map and
Chart Series 24, pi. 3.
Sandberg, C.A., and W. Zeigler. 1984. Palmatolepsis- Based Revision
of Upper Part of Standard Late Devonian Conodont Zonation, p.
179-189. In D.L. Clark ( ed . ) , Conodont Biofacies and
Provincialism. Geological Society of America Special Paper 196.
Sartenaer, P. 1961. Late Upper Devonian (Famennian)
Rhynchonellid Brachiopods. Institut Royal des Sciences
Nature lies de Belgique Bulletin 37:1-10.
Sartenaer, P. 1967. Correlation, p. 1043-1060. In International
Symposium on the Devonian System, edited by D. H. Oswald.
Alberta Society of Petroleum Geologists, Calgary.
Sartenaer, P. 1969. Late Upper Devonian Rhynchonellid
Brachiopods from Western Canada. Geological Survey of
Canada, Bulletin 169, 269 p.
Sartenaer, P. 1970. Nouveaux Genres Rhynchonellides
(Brachiopodes) du Paleozoique. Institut Royal des Sciences
Naturelles de Belgique Bulletin 46: 1 -32.
Schmidt, H., and D.J. McLaren. 1965. Paleozoic Rhynchonellacea,
p. H552-H597. In Treatise on Invertebrate Paleontology, Part
H, Brachiopoda, edited by R. C. Moore. Geological Society of
America and University of Kansas Press, Lawrence.
Schopf, J.M., and J.F. Swieterling. 1970. The Foerstia Zone of
the Ohio and Chattanooga Shales. U.S. Geological Survey
Bulletin 1294-H:1-15.
Schuchert, C. 1929. Brachiopoda, p. 1-140. In Schuchert and
C.M. Levene. Brachiopoda. Fossilium Catalogus I: Animalia,
Pars 42, Junk, ( Berlin).
Schuchert, C., and G.A. Cooper. 1931. Synopsis of the
Brachiopod Genera of the Suborders Orthoidea and
Pentameroidea, with Notes on the Telotremata. American
Journal of Science, Series 5, 22:241-251.
Schuchert, C., and G.A. Cooper. 1932. Brachiopod Genera of the
Suborders Orthoidea and Pentameroidea. Yale University
Peabody Museum of Natural History, Memoir 4, No. 1 , 270 p.
Schwimmer, B.A. 1988. Brachiopoda from the Chagrin Shale
(Late Devonian) of Ohio. Geological Society of America,
Program with Abstracts 20(5):387.
Schwimmer, B.A., J.T. Hannibal, D.J. Stukel II, and R.M.
Feldmann. 1987. The Paleontology and Depositional
Environment of the Chagrin Shale (Famennian) in
Northeastern Ohio. Second International Symposium on
the Devonian System, Program and Abstracts, Calgary,
Alberta, p. 204.
Shumway, S.E. 1982. Oxygen Consumption in Brachiopods, and
the Possible Role of Punctae. Journal of Experimental
Marine Biology and Ecology 58:207-220.
Slansky, M. 1986. Geology of Sedimentary Phosphates. Elsevier,
Amsterdam, 210 p.
Stehli, F.G. 1954. Lower Leonardian Brachiopoda of the Sierra
Diablo. American Museum of Natural History Bulletin
105:257-358.
Stukel, D.J. II. 1987. Ichnology and Paleoenvironmental
Analysis of the Late Devonian (Fammenian) Chagrin Shale of
Northeast Ohio. Kent State University, Masters Thesis
(unpubl.), 95 p.
Szmuc, E.J., R.G. Osgood, Jr. and D.W. Meinke. 1976.
Lingulichnites , a New Trace Fossil Genus for Lingulid
Brachiopod Burrows. Lethaia 9: 163-167.
Termier, H., and G. Termier. 1949. Essai sur I'Evolution des
Spiriferides. Maroc Service Geologique du Division des
Mines et de la Geologie, Notes et Memoires 74:85- 1 1 2.
Thayer, C.W. 1986. Respiration and the Function of Brachiopod
Punctae. Lethaia 19:23-31.
Thayer, C.W., and H.M. Steele-Petrovic. 1975. Burrowing of the
Lingulid Brachiopod Glottidia pyramidata: Its Ecologic and
Paleoecologic Significance. Lethaia 8:209-221 .
Thomas, I. 1910. British Carboniferous Orthotetinae. Great
Britain Geological Survey Memoir 1 ( 2 ):83- 1 34.
Waagen, W.H. 1882-1885. Salt Range Fossils, Part 4, (2)
Brachiopoda. Palaeontologia lndica, (Memoirs of the
Geological Survey of India), Series 13, 1:329-770 (1882),
2:391-546 (1883), 5:729-770 (1885).
Weidner, W.E. 1983. Paleoecological Interpretation of
Echinocarid Arthropod Assemblages in the Late Devonian
(Famennian) Chagrin Shale, Northeast Ohio. Kent State
University, Masters Thesis (unpubl.), 66 p.
Weidner, W.E., and R.M. Feldmann. 1983. Paleoecological
Interpretation of Echinocarid Arthropod Assemblages in the
Late Devonian (Fammenian) Chagrin Shale, Northeastern
Ohio. Journal of Paleontology 59:986-1004.
Wetzel, R.G. 1983. Limnology. Saunders College Publishing,
Philadelphia, 753 p.
Weyer, D. 1967. Kitakamithyris Minato, 1951 (Brachiopoda,
Spiriferida) aus dem Etroeungt (Oberdevon) und Tornai
(Unterkarbon) des Rheinischen Schiefergebirges. Geologie
[Berlin] 16:433-451.
Zagger, G., and P. Banks. 1989. Age and Origin of the Skinner’s
Run Pyrite Bed, Cuyahoga County, Ohio. The Ohio Journal
of Science 89(2):9.
KIRTLANDIA
The Cleveland Museum of Natural History
June 1990 Number 45 :33-38
Estimation of Numbers for a
Riverine Necturus Population
Before and After
TFM Lampricide Exposure
TIMOTHY O. MATSON
The Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Abstract
The purpose of this research was to assess the impact of the lampricide, 3-trifluormethyl-
4-nitrophenol, upon the mudpuppy, Necturus maculosus, in the Grand River, Lake County,
Ohio. Pre-treatment and entire season population estimates for 1987 and 1988 were made
using Schnabel estimation. Population estimates ranged from 556-1,118 per km in 1987 but
declined to 280-397 per km in 1988. Comparison of pre-treatment or entire season estimates
between years indicates a minimum 29% decrease in population size. Individuals within the
range of total lengths examined appear to exhibit equal sensitivity to TFM. Intensive seining
and manual turning of rock slabs were effective methods for capturing Necturus: trapping and
electroshocking proved ineffective.
Introduction
The lampricide 3-trifluoromethyl-4-nitrophenol (TFM)
has been used in control of sea lamprey in the Great Lakes
drainage since 1958 (National Research Council of
Canada, 1985). The chemical has been found toxic to a
number of vertebrate taxa, and the degree of toxicity is
dependent upon application concentration, temperature, pH
and hardness of the water. Relatively little field research
has been published concerning effects of TFM on
amphibians even though field treatment summaries
frequently refer to mortality of amphibians. Gilderhus and
Johnson (1980) note mortality in anuran tadpoles in 16% of
treatment summaries; mortality in Necturus was noted in
32% of observations from Lake Superior tributaries, 36%
34
Matson
No. 45
from Lake Michigan tributaries, and of those observations
18% referenced high mortality.
In October 1986, TFM was applied to Conneaut Creek
located in Ashtabula County, Ohio and Crawford County,
Pennsylvania. Mortality in Necturus was high, but no
population data were available either before or after the
TFM application. The Grand River in Lake and Ashtabula
Counties of Ohio was scheduled to receive TFM treatment
during spring of 1987. To assess the impact of the lampricide
upon the population of Necturus maculosus , pre-treatment
and post-treatment population estimates were made and then
compared. For additional comparative purposes, the study
was continued during 1989, and statistical comparison of
population estimates was made between years.
Study Area and Methods
Field work extended from March through July 1987
and from April through July 1988. High water, shortage of
time, and difficulty in capturing animals limited the
number of sites where the population was studied to one.
The 600 m study site was located in the Grand River,
Madison Twp., Lake County, Ohio (Thompson, Ohio 7.5
min. topographic map, 1960, photorevised 1970). Two
meanders, one riffle, three gravel bars, and several pools
were major channel features. Pools present were of three
types: isolated, flood-plain pools (IFPP); rapid-bounded
pools (RBP); and sequential, incised channel-margin pools
(SICMP). Water depth ranged from 10 cm to
approximately 2 m during the summer months but was
subject to dramatic fluctuations from low water under
drought conditions during summer, 1988, to spring flood
stage in early March or April. Substrate in the river channel
consisted of sand, gravel, cobbles, large glacial erratics, and
slabs of siltstone from the Chagrin Shale, some of which
exceeded two meters in length. Pools had substrates of sand,
silt and large boulders or siltstone slabs, whereas others
collected detritus and silt to a depth of 10-15 cm.
Four methods were used to capture Necturus during
1987: intensive seining with 3.7 m and 7.6 m bag seines,
manual overturning of rock slabs capturing animals in a
1 .8 m seine, trapping with baited minnow-like traps, and
electroshocking using both back-pack and small pull-
behind units. Intense seining was productive from late
March to early May and consisted of seining SICMP and
RBP repeatedly with assistants churning the substrate
before the net. This produced chocolate-colored water with
much floating and suspended organic material. Overturning
slabs was productive from mid-May through July when
DATE OF CAPTURE
FIGURE 1. Histogram of the number of Necturus maculosus captured on different dates during 1987 and 1988. Exposure to TFM
occurred on 27 April 1987 and is indicated by the arrow.
1990
Estimation of Numbers of Necturus
35
1987 1988
DATE OF CAPTURE
FIGURE 2. Number of Necturus maculosus within size cohorts captured on different dates during 1987 and 1988. Exposure to TFM
occurred on 27 April 1987 and is indicated by the arrow.
water depth was reduced. Trapping and electroshocking
were unsuccessful for capturing Necturus. Trapping
produced only 1 capture in 126 trap nights, and
electroshocking was totally ineffective. Seining and slab
turning were the only methods used in 1988.
All animals captured were marked by toe-clipping.
Individuals marked in the pre-treatment phase during 1987
were marked by removing the outermost toe on the left
hind foot; individuals captured during the post-treatment
phase of the study were marked in similar fashion on the
right. All animals were toe clipped on the left front foot in
1988. The Schnabel method of population censusing and
estimation was used to estimate population size before and
after treatment of the river with TFM. Chapman’s Poisson
table was used to calculate 95% confidence interval
estimates for <50 recaptures. The standard normal
approximation to the Poisson was used to estimate 95%
confidence intervals for >50 recaptures (Seber, 1973).
Population estimates were compared using the normal
approximation z-statistic (Seber, 1973). Null hypotheses
tested were: 1) There was no difference in population size
between pre-treatment and post-treatment estimates during
1987; 2) There was no difference in population size
between intervals corresponding to the pre-treatment
period between years; 3) There was no difference in
population size for entire data sets between years.
Total length (TL) of each animal was measured to the
nearest millimeter (mm) in the field to examine differential
mortality associated with size and to compare catchability
using various methods.
The site was searched for dead Necturus during and
following TFM exposure on 27 and 28 April.
Results
Intensive seining yielded the greatest number of
captures and detected most small larvae and juveniles.
Seining accounted for 181 of 212 (85.4%) animals
examined in the field in 1987 and for 244 of 290 (84.1%)
captures in 1988. Overturning rock slabs resulted in capture
of 77 animals (15.3%). All but four animals captured using
this technique had minimum TL of 150 mm.
Few animals were obtained from 21 to 29 March.
Number of captures increased in early April, peaked during
mid-month, and then decreased to fluctuate about mid-
March levels through July (Figure 1). Captures of various
size classes of Necturus were not evenly distributed
temporally (Figure 2). No juveniles or small larvae were
obtained before 29 March due at least in part to technique
and high water, and only 51 animals with TL<150 mm were
captured after 1 May. Most animals captured between 29
March and 6 May were <150 mm TL and accounted for the
elevated number of captures during this period (Figs. 1, 2).
TFM treatment was initiated upstream of the site on 26
April and reached the site early morning on 27 April.
Repeated searching of the site on the 27th and 28th of April
for dead Necturus yielded six individuals of which three
were marked. One individual was an adult in excess of 175
mm TL, while the remaining five were under 75 mm TL.
One hundred eighty-one (181) Necturus were captured
during the pre-treatment phase of the study in 1987, and 31
individuals were obtained during the post-treatment phase.
Population estimates and 95% confidence interval estimates
are presented in Table 1. The comparison of pre-treatment
with post-treatment population estimates for 1987 was
invalid because only two Necturus were recaptured during
the post-treatment phase having only the post-treatment
36
Matson
No. 45
Table 1. Data obtained during two years of study of the Grand River Necturus Population. Comparisons of
population estimates between years for both the pre-treatment and the entire season were significantly different
(P<0.001).
Number
Number
Number
Population
95%
examined
Marked
Recaptured
Estimate(N)
C.I.E.
Pre-treatment
1987
178
153
25
805/km
513, 1188/km
1988
205
120
85
262/km
212, 323/km
Entire Season
1987
209
171 *
35
803/km
556, 1118/km
1988
290
165
125
333/km
280, 397/km
* 3 marked animals found dead after TFM treatment
mark. Comparison of pre-treatment population estimates
between years was statistically significant (PcO.OOl), as was
the entire season estimate (PcO.OOl) between years.
Discussion
Cagle (1954) found baited nets, dip nets, seines and
trotlines to be ineffective in capturing Necturus
maculosus louisianensis in Louisiana. Setlines produced
satisfactory catches of adults, and intensive dip netting and
seining produced larvae. Shoop (1965) reported traps
constructed of 6 mm wire mesh unsuccessful in capturing
Necturus maculosus louisianensis and N. beyeri. However,
unusually high success was achieved using baited setlines.
Although no direct indication of success was stated, dip
netting and seining were used by Shoop (1965) for
capturing Necturus in Louisiana. Results reported here
show that intensive seining produced the greatest number
of captures and almost all animals <75 mm TL. Neither
setlines nor trotlines were used in this study as both
techniques would yield a bias toward large juveniles and
adults and would result in a serious underestimate of
population size and in increased mortality because
Necturus tends to swallow baited hooks.
Shoop (1965) used electroshocking with some success
to capture Necturus in excess of 140 mm snout-vent length
in Big Creek, Louisiana. Eleven of 50 individuals were
obtained by electroshocking; the remaining individuals
being taken on baited set lines. No indication of time, effort
required or success rate was stated. Fitch (1959) was
successful in collecting 28 Necturus with the aid of an
electroshocker in April at Sugarloaf Lake, Michigan. My
results contrast with those in Louisiana and Michigan in
that electroshocking was found totally ineffective.
Employment of the technique under different water
conditions in the Grand River or where animals are not
concealed beneath slabs may produce some catches, but it
seems improbable that the method is useful for capturing
many individuals as is required for population estimates.
This assertion is supported by the observation that after
using an electroshocker in a SICMP with no Necturus
apparent, the pool was intensively seined and 13 individuals,
small larvae, juveniles and adults, were captured.
The dramatic increase in number of Necturus caught
from late March through the end of April is difficult to
explain. Bishop ( 1926) mentioned movements or migrations
in streams and lakes, and vernal movements from lakes into
streams were noted by Pope (1962). Movements into lake
shallows during April and May were reported by Fitch
(1959) and by Gibbons and Nelson (1968); adult Necturus
were reported to retreat to deeper, colder waters in summer,
while larvae remained in shallower, warmer waters (Harris,
1958). Movement into shallows does not appear to be the
factor involved at this site in the Grand River. On the
contrary, nearly all Necturus were obtained during March,
April, and early May in SICMP where water depth was
greater than in the main river channel.
Field technique improved over time during March, 1987,
and poor technique may partially explain the low number of
captures early in the study. High water in early spring
hampered activities in the river in both years and contributed
to the low number of captures. However, the same seining
techniques which produced high catches in April were less
successful from late April through July in the same pools.
Furthermore, the capture pattern in 1988 is nearly identical to
1990
Estimation of Numbers of Necturus
37
that of 1987. Small larvae and juvenile classes that were
common in SICMP collections from late March through late
April or early May were absent thereafter. Movement from
pools into the channel may explain these observations.
Consequently, for population size estimates or age-size
population studies this procedure produces a biased estimate
against large juveniles and adults; whereas slab turning
produces a bias against larvae and juvenile classes.
Inspection of the data from 1987 with respect to the
size and number of individuals captured over time
suggested that Necturus <150 mm TL may have suffered
higher mortality than larger animals (Figure 2). This
hypothesis was formulated on the observation that after I
May 1987 few animals <150 mm TL were captured. If this
hypothesis was correct, then proportionately fewer animals
within this size cohort would have been expected to appear
in the 1988 capture data. On the contrary, more Necturus
within this cohort were captured in 1988; apparently
differential mortality did not occur. The total number of
Necturus examined in 1988 was higher than that in 1987
(Table 1); improved capture techniques and reduction of
water flow and depth resulting from drought are probable
causes for higher capture in 1988.
My data do not agree with those derived from caged
Necturus experiments in the Grand River (Daugherty and
Klar, 1988) where observed mortality was restricted to
individuals with TL <50 mm. The experimental procedure
is not explicitly described by Daugherty and Klar, and it is
not stated if individuals of different size classes were
placed into a single cage or were placed by similarity of
size into separate cages. Furthermore, it is unclear how
mortality was observed, i.e. if the animals were found dead
or if some were injured or disappeared, perhaps due to
cannibalism. They suggest overcrowding in cages as
possible cause for observed mortality. This is unlikely for
the day or two the animals were confined unless cages
were very small and crowding was intense resulting in
physical interaction or injury (personal observation). It is
possible that mortality observed in caged experiments did
reflect increased susceptibility to TFM by small Necturus
and that small individuals in pools where much of my data
were obtained were able to escape some of the TFM effects
by burrowing into the substrate.
Both pre-treatment and entire season comparisons
between years indicate large reduction in population size
from 1987 to 1988 following exposure to TFM within the
treatment range of 2.1 mg/1 to 6.6 mg/1 (U.S. Fish and
Wildlife chemical treatment summary). The 29% reduction
in population size hinges in part upon accuracy of
population estimates and was based upon the percentage
decrease between the lowest 95% confidence interval
estimate for the entire year of 1987 (556) and the upper
estimate for 1988 (397). Schnabel estimates may reflect
only the general order of magnitude of population size
(Schnabel, 1938) and are dependent upon validity of
certain assumptions (Seber, 1973; Caughley, 1977).
Recruitment from reproduction was not a factor since no
hatchlings, which are readily identifiable, were included in
the estimates. Immigration and emigration may be factors
that effect the population estimates; I am currently
investigating these aspects of Necturus biology. Assuming
immigration and emigration rates are equal and remain
similar between years, they likely have negligible impact
on the comparison of population size between years. The
assumptions of random sampling and equal catchabi 1 i ty of
marked and unmarked individuals probably are valid. Toe
clipping temporarily marks Necturus ; however, clipped
animals were readily identifiable as marked animals within
the few months following marking. Therefore, loss of
marks was not considered to effect estimates.
Consequently, based upon two years of study, population
size and degree of population reduction can only be
approximated. Because mortality of Necturus during the
TFM treatment was apparent (personal observation; U.S.
Fish and Wildlife chemical treatment collection summary,
1987), notable decrease in population size can, at least in
part, be attributed to the lampricide.
Acknowledgements
Partial funding for the research was received from the
Ohio Department of Natural Resources, Division of
Wildlife and from the Kirtlandia Society of The Cleveland
Museum of Natural History. I thank R.A. Pfingsten, D.
Waller and an anonymous reviewer for reviewing the
manuscript and making helpful suggestions for its
improvement. I am indebted to Nick Ashbaugh, Katrine
Bosley, and to the numerous volunteers who generously
gave their time in the field.
References
Bishop, S.C. 1926. Notes on the Habits and Development of the
Mudpuppy Necturus maculosus (Rafinesque). New York
State Museum Bulletin No. 268: 1-60.
Cagle, F.R. 1954. Observations on the Life History of the
Salamander Necturus maculosus louisianensis . Copeia
1954:257-260.
Caughley, G. 1977. Analysis of Vertebrate Populations John
Wiley and Sons.
Daugherty, W.E. and G.T. Klar. 1988. Sea Lamprey Management
in the United States in 1987. U.S. Fish and Wildlife annual
report to Great Lakes Fishery Commission.
Fitch, K.L. 1959. Observations on the Nesting Habits of the
Mudpuppy, Necturus maculosus Rafinesque. Copeia
1959:339-340.
Gibbons, J.W., and S. Nelson, Jr. 1968. Observations on the
38
Matson
No. 45
Mudpuppy, Necturus maculosus, in a Michigan Lake.
American Midland Naturalist 80:562-564.
Gilderhus, P.A., and B.G. Johnson. 1980. Effects on Sea Lamprey
(Petromyzon marinas ) Control in the Great Lakes on Aquatic
Plants, Invertebrates and Amphibians. Canadian Journal of
Fisheries and Aquatic Science 37:1985-1905.
Harris, J.P. 1959. The Natural History of Necturus: I. Habitats and
Habits. Field and Laboratory 27: 1 1-20.
National Research Council of Canada. 1985. TFM and Bayer 73
Lampricides in the Aquatic Environment . Publication No.
NRCC 22488.
Pope, C.H. 1964. Amphibians and Reptiles of the Chicago Area.
Chicago Natural History Museum Press.
Schnabel, Z.E. 1938. The Estimation of the Total Fish Population
of a Lake. American Mathematical Monthly 45:348-352.
Seber, G.A.F. 1973. The Estimation of Animal Abundance and
Related Parameters. Hafner Press.
Shoop, C.R. 1965. Aspects of reproduction in Louisiana
Necturus Populations. American Midland Naturalist
74:357-367.
United States Fish and Wildlife Service. 1987. Chemical
Treatment Collection Summary. Grand River. Lake
County, Ohio.
United States Fish and Wildlife Service. 1987. Chemical
Treatment Summary. Grand River. Lake County. Ohio.
•NATURAL HISTORY*
Published by
The Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106
Production by Wendy Donkin
NUMBER 46
ff i/ELAND, OHIO
KIRTLANDIA
Anthropology
P ale o epidemiology of For otic Hyperostosis
in the Libben and Bt-5 Skeletal Populations 1
Robert P. Mensforth
Paleontology
North American Late Devonian
Cephalopod Aptychi 49
Calvin J. Frye and Rodney M. Feldmann
•NATURAL HISTORY*
KIRTLANDIA
The Scientific Publication of The Cleveland Museum of Natural History
David S. Brose and
Joseph T. Hannibal, Editors
Brief History and Purpose
Kirtlandia, a publication of The Cleveland Museum of Natural History, is named in honor of Jared
Potter Kirtland, a noted nineteenth-century naturalist who lived in the Cleveland, Ohio area. It began
publication in 1967 and is a continuation of the earlier series Scientific Publications volumes 1 to 10
(1928-1950), and new series volumes 1 to 4 (1962-1965). Supported by the Kirtlandia Society of The
Cleveland Museum of Natural History, Kirtlandia is devoted to the publication of scientific papers in
the various fields of inquiry within the Museum’s sphere of interest: Cultural and Physical
Anthropology; Archaeology; Botany; Geology; Paleobotany; Invertebrate and Vertebrate
Paleontology; Systematic Ecology; and Invertebrate and Vertebrate Zoology. Issues will vary from
single monographs to collections of short papers, review articles, and brief research notes.
Kirtlandia is abstracted in Biological Abstracts and indexed in Bibliography and Index of Geology and
Zoological Record.
Associate Editors
James K. Bissell, The Cleveland Museum of Natural History
Bruce Latimer, The Cleveland Museum of Natural History
Martin Rosenberg, Case Western Reserve University
Sonja Teraguchi, The Cleveland Museum of Natural History
Editorial Advisory Board
Rodney Feldmann, Kent State University
Michael C. Hansen, Ohio Geological Survey
Richard Meindl, Kent State University
G. Michael Pratt, Heidelberg University
David H. Stansbery, Ohio State University
Frederick H. Utech, Carnegie Museum of Natural History
Ed Voss, University of Michigan
Andrew M. White, John Carroll University
Kirtlandia No. 46
ISSN 0075-6245
© 1991 by The Cleveland Museum of Natural History
Cleveland, Ohio
Copies of Kirtlandia, and many issues of the Scientific Publications series of
The Cleveland Museum of Natural History, are available for sale.
Write to: Library, The Cleveland Museum of Natural History, 1 Wade Oval Drive, University Circle,
Cleveland, Ohio 44106-1767 for a current price list.
KIRTLANDIA
The Cleveland Museum of Natural History
August 1991 Number 46: 1 -47
Paleoepidemiology
of Porotic Hyperostosis
IN THE LlBBEN AND BT-5
Skeletal Populations
Robert P. Mensforth
Department of Anthropology
Cleveland State University
Cleveland, Ohio 44115
Abstract
The frequencies with which porotic hyperostosis occurred in the Libben Late Woodland (n = 580) and Bt-
5 Late Archaic (n = 247) skeletal populations were examined. Aside from temporospatial and cultural
affiliations, the skeletal samples representing these two prehistoric band level societies principally differ with
respect to level of subadult mortality. That is, juvenile death rates were substantially greater for the Libben
group. The goals of the study were to (1) identify those factors which played a paramount role in the etiology
of the skeletal lesion in the two groups and (2) evaluate the extent to which porotic hyperostosis serves as a
useful bioassay of health status in earlier human groups.
The epidemiological patterns that were observed support the conclusion that Libben and Bt-5 porotic
hyperostosis was a fundamental consequence of iron deficiency anemia. In addition, the only significant
difference in the frequency of porotic hyperostosis that was observed between the two groups was confined to
subadults. Here, Libben children displayed a significantly greater frequency of unremodeled lesions and total
lesions compared to those at Bt-5. Evidence is presented which suggests that dietary inadequacy and
parasitism played a minor role in the etiology of porotic hyperostosis in the two groups. Alternatively, local
environmental circumstances associated with habitation and exploitation of the Black Swamp may have
played a fundamental role in elevating infectious disease loads resulting in a greater prevalence of iron
deficiency anemia in Libben infants and children compared to those at Bt-5. Factors which contribute to a low
level of non-specificity, and age related differences in biological and demographic sensitivity for porotic
hyperostosis are discussed.
?
Mensforth
No. 46
Introduction
Diseases which become manifest in the human skeleton
have piqued the curiosity of physical anthropologists for a
considerable period of time. These include both (a) specific
and (b) non-specific skeletal and dental lesions. A specific
skeletal lesion is one that can be attributed to a single
disease entity (eg., multiple myeloma), whereas a non-
specific lesion is one that can be induced by a wide variety
of disease states and which must therefore be regarded as
multicausal . The etiology of a specific skeletal lesion can
usually be established with a high level of confidence, and
often provide us with information about the health status of
a particular individual. However, such abnormalities are
infrequent in small human groups. Because of this, specific
skeletal lesions provide us witli very limited knowledge
about the major forces of morbidity and mortality which
operated in earlier human groups. The paleoepidemi-
ological utility of specific skeletal lesions is therefore
markedly restricted.
Alternatively, non-specific skeletal lesions, though
multicausal in nature, are sufficiently frequent to yield
epidemiological information on a populational basis. Thus,
it is not remarkable that paleoepidemiologists now employ
a variety of non-specific indicators of disease and
nutritional stress in order to evaluate the potential
significance of pathological responses that occurred in
earlier human groups. Such responses are presumed to
reflect demographic adaptations to diverse and often
changing environmental, sociocultural, political, and
socioeconomic circumstances. The stress indicators which
have been used most frequently include: (1) enamel
hypoplasias, (2) enamel hypocalcifications, (3)
radiographic evidence of long bone growth perturbations as
reflected in Harris lines, (4) patterns of long bone
diaphyseal lengths attained at each age, (5) cortical bone
remodeling dynamics, and (6) skeletal lesions such as
porotic hyperostosis and periosteal reactions (Goodman et
al ., 1984; Cohen and Armelagos, 1984). It is important to
realize, however, that non-specific stress indicators are not
of equal paleoepidemiological utility. This is due to the fact
that the underlying genetic and environmental factors
which interact to determine the phenotypic expression of
skeletal and dental stress indicators varies considerably.
Thus, each stress indicator is characterized by its own
tissue-specific set of benefits and liabilities as a potential
bioassay. Skeletal biologists who employ non-specific
indicators of disease and nutritional stress to evaluate
population fitness in earlier human groups must therefore
address two cardinal issues. The first concerns etiology of
the stress indicator in question. Thus, in order to identify
probable cause of a non-specific lesion on a populational
basis, differential diagnoses must be used to evaluate a
series of tenable explanatory hypotheses. Those hypotheses
which best explain the age and sex specific
epidemiological patterns that are observed must then be
ascribed greater valence as a source of inference.
The second issue concerns the extent to which a
particular population pathology serves as a useful bioassay
of disease and/or nutritional stress in earlier human groups.
This problem can be approached by examining the
demographic sensitivity of our paleoepidemiological tool.
Therefore, if a skeletal response is to be regarded as a
useful indicator of disease stress, it would be predicted that
differences in the age and sex specific frequency
distributions of the lesion will correspond in direction,
though not necessarily in magnitude, to differences in age
and sex-specific mortality rates that are observed for two or
more groups. For those population studies where the
relationship between lesion frequencies and mortality rates
are found to be discordant, or totally non-existent, then the
epidemiological utility of the bioassay must seriously be
questioned. In other circumstances the relationship
between lesion frequencies and mortality rates may be
weak, but nonetheless apparent. Here, we must carefully
consider the extent to which potential interpretive value of
a lesion outweighs the manifest limitations that
characterize such a lesion. Furthermore, in those
circumstances where the demographic sensitivity of a
particular bioassay is found to be compromised, the
paleoepidemiologist must critically evaluate the potential
roles that other factors may have played in generating the
patterns of lesion distribution that were observed. The
latter include various physiological and developmental
phenomena, methodological and sampling problems,
cultural practices, etc.
Descriptive research in paleopathology has thus far
contributed greatly to our knowledge about the distribution,
and nature, of diseases that have influenced the course of
human evolution in time and space. The primary value of
such studies is clearly recognized. However, the view
posited here is that paleoepidemiology should be regarded
as a subdiscipline of paleodemography, and should not be
considered as a parallel, or independent, line of inquiry.
Moreover, studies in paleopathology which are not
conducted in a demographic context must also be regarded
as devoid of evolutionary significance. Thus, it is
suggested here that non-specific stress indicators of
maximum value will be those which possess three
fundamental properties described as follows:
1) The stress indicator should exhibit a high degree
of biological sensitivity to one or more
environmental perturbations.
2) The stress indicator should be characterized by a
high level of demographic sensitivity such that
population patterns in the frequency of occurrence
1991
PA LEOEP1DEMIOLOG Y OF POROTIC HYPEROSTOSIS
3
C
FIGURE 1. Libben children, la, Libben child (KSU-08027, 7 years) that displays extensive unremodeled parotic hyperostosis affecting the
superior orbital plates, lb, Libben child (KSU-04035, 12-18 months) which exhibits a pronounced unremodeled porotic hyperostotic
lesion affecting the right parietal bone fragment. Note the enlarged pore channels enveloped by a well defined microporous cribriform
mesh, lc. Shown here are the complimentary photographic and radiographic images of a right parietal bone cross-section. This Libben
child (KSU-01258, 3 years) widening of diploic spaces, irregular trabeculation, and the hair-on-end striations that are characteristic of
erythroid bone marrow hyperplasia.
of a lesion (i.e., morbidity) can clearly be related
to patterns of age and sex specific mortality.
3) The stress indicator should be characterized by a
low level of non-specificity. That is, we should be
able to infer probable cause of the stress indicator,
on a populational rather than individual basis, with
a relatively high level of confidence.
Given these perspectives, the purposes of this
monograph are to (1) examine in detail the factors which
promote the development of erythroid marrow hyperlasia
associated anemias in human groups, (2) establish the
demographic significance of such anemias in human
groups, (3) develop hypothetical model of iron deficiency
anemia related porotic hyperostosis that would be expected
to occur in earlier human groups, (4) conduct a case study
which explores the extent to which this model provides
insight with regard to the etiology of porotic hyperostosis
as seen in two prehistoric band level societies, and (5)
evaluate the demographic sensitivity and paleoepidemi-
ological utility of porotic hyperostosis as a bioassay of
population fitness in earlier human groups.
Porotic Hyperostosis in Prehistory
Porotic hyperostosis is a descriptive term for cranial
lesions that display a coral, cribriform, or sieve-like
porosity (Angel, 1966; Cule and Evans, 1968; El-Najjar,
1975). Common referents for the skeletal lesion that have
been reported in the earlier literature include: cribra
orbitalia (Welcker, 1888), symmetrical osteoporosis
(Hrdlicka, 1914; Williams, 1929), osteoporosis of the
cranium (Muller, 1935), cribra cranii (Williams, 1929;
Henschen, 1961), and spongy hyperostosis (Putshcar,
1966). The lesions are often symmetrical in disposition,
exhibit variable degrees of osseous tissue hypertrophy, and
frequently display peripheral hypervascular channels
(Hengen, 1971). These skeletal changes are usually visible
upon close macroscopic examination (see Figures la-c).
Radiographic and histological studies of porotic
hyperostosis show that the lesions are characterized by
irregular trabeculation, widening of diploic spaces,
thinning of the endo-and-ectocranial tables (primarily the
outer), and a radial pattern of bone spiculation known as
hair-on-end striations which are oriented perpendicular to
the endocranial table of an affected bone (Moseley, 1965;
4
Mensforth
No. 46
Greenfield, 1975; El-Najjar and Robertson, 1976).
Porotic hyperostotic skeletal changes can be observed
most frequently affecting the anterior portion of the
superior orbital plates (Hengen, 1971). Lesions affecting
the pericranial surfaces of the frontal, parietal, and occipital
bones are more commonly seen in subadults (Carlson et
at ., 1974). Endocranial skeletal changes tend to be
uncommon relative to those which are described above
(Henschen, 1961; Hengen, 1971). Severe manifestations of
the lesion are accompanied by excessive osseous tissue
hypertrophy, obliteration of the outer table, marked
resorption and thinning of the inner table, and involvement
of temporal, sphenoid, and facial bones (Moseley, 1965).
Skeletal changes that are identical to those described as
porotic hyperostosis have been identified in a number of
pathological conditions (see Table 1). Radiographic
evidence of porotic hyperostosis has been observed most
frequently as a consequence of iron deficiency anemia and
the chronic hemolytic anemias (Sheldon, 1936; Caffey,
1937; Eng, 1958, Sax, 1963; Baker, 1964; Powell et al. ,
1965; Aksoy et al., 1966; Lanzkowski, 1968; Agarwal et
al., 1970; Shahidi and Diamond, 1960; Britton et al., 1960;
Burko et al., 1961; Moseley, 1974; and Williams et al.,
1975). The lesions are less frequently seen in association
with cyanotic congenital heart disease (Nice, 1964),
hereditary spherocytosis (Trucco and Brown, 1967),
polycythemia vera (Dykstra and Halberstma, 1940), and
pyruvate kinase deficiency (Becker et al., 1971). For all of
the conditions listed above the skeletal changes that affect
the cranium are the direct result of erythroid bone marrow
hyperplasia which occurs in response to an underlying
anemic stimulus (Moseley, 1974). Thus, porotic
hyperostosis is best considered a non-specific consequence
of bone marrow proliferation.
The earliest descriptive reports that identified porotic
hyperostosis in prehistoric human groups were published in
the late 19th and early 20th centuries (Welcker, 1888;
Wood-Jones, 1910; Hrdlicka, 1914; Moore, 1929; Hooton,
1930; and Muller, 1935). Although several hypotheses
were posited to explain the skeletal lesion, the pathogenesis
of porotic hyperostosis remained vague and enigmatic prior
to the mid 1960‘s (Henschen, 1961; Nathan and Haas,
1966; Angel, 1964; and Moseley, 1965). Historical reviews
of the earlier accounts are available in the more recent
literature (see El-Najjar and Robertson, 1976; El-Najjar et
al., 1976). In general, the earlier studies demonstrated that
porotic hyperostosis occurred in skeletal groups of
widespread temporospatial distribution, and that the lesions
were somewhat more common in populations that had
lived in the equatorial regions of the Old and New Worlds.
Currently, two hypotheses are regarded as primary
explanations for the etiology of porotic hyperostosis that
has been observed in earlier human groups. One considers
TABLE 1. List of the Major Clinical Conditions Where
Radiographic Evidence of Skeletal Changes in the
Cranium Have Been Reported as a Primary Consequence
of Erythroid Bone Marrow Hyperplasia
I. Congenital Hemolytic Anemias:
A. Thalassemias
1 . Thalassemia Major (i.e., Mediterranean
Disease, Cooley’s Anemia, Erythroblastic
Anemia)
2. Thalassemia Intermedia - Severe
Heterozygous
3. Thalassemia Minor - Mild Heterozygous
B. Sickle Cell Disease
1. Sickle Cell Anemia (Hemoglobin S-
Homozygous)
2. Hemoglobin C - Homozygous
3. Hemoglobin E - Homozygous
4. Hemoglobin S-C
5. Hemoglobin S - Thalassemia
6. Other Infrequent Abnormal Hemoglobins
C. Hereditary Nonspherocytic Hemolytic Anemias
1. Glucose-6-Phosphate Dehydrogenase
Deficiency
2. Pyruvate Kinase Deficiency
3. Other Rare Enzyme Deficiencies
D. Hereditary Spherocytosis (i.e.. Spherocytic
Anemia and Congenital Hemolytic Jaundice
E. Hereditary Elliptocytosis (rare)
II. Iron Deficiency Anemia
III. Cyanotic Congenital Heart Disease (rare)
IV. Polycythemia Vera in Childhood (rare)
Adapted from Moseley ( 1965), The paleopathologic riddle of
“symmetrical osteoporosis.” American J. of Roentgenology
95:135-142.
the lesion to be a fundamental response to some variant of
hemolytic anemia that occurred as a consequence of
endemic falciparum malaria (Angel, 1966; Zaino, 1964;
Ascenzi and Salistreri, 1977). The alternative hypothesis
suggests that porotic hyperostosis is a direct result of iron
deficiency anemia (Moseley, 1965, Moseley 1966; Hengen,
1971). Historically, both views have been offered to
explain the occurrence of porotic hyperostosis in Old and
New World skeletal populations alike.
The geographic distribution of the lesion in some areas
of the Old World due not conflict with the hemolytic
anemia hypothesis (Angel, 1967). However, a substantial
1991
Paleoepidemiolog y of P orotic Hyperostosis
5
body of clinical, experimental, and epidemiological
research supports the view that iron deficiency anemia
played a substantial, if not the major, role in the patho-
genesis of porotic hyperostosis in all earlier human groups
(Mensforth et al. 1978, and references therein). The major
proponents of these divergent perspectives, and evidence
offered to support such inferences, will thus be considered.
Angel (1967) has favored the view that porotic
hyperostosis that is seen in several Old World circum-
Mediterranean skeletal groups represents a bony response
to the hemolytic anemias of thalassemia and/or sickle cell
anemia. The lesion is interpreted to be direct skeletal
evidence of balanced polymorphic adaptations which arose
in response to the selection pressures of falciparum malaria
in pre-modem Old World agricultural communities (Angel,
1966). Evidence given in support of this hypothesis
includes geographic distribution and frequency of the
lesion in circum-Mediterranean skeletal groups, and overall
patterns of cranial and post-cranial skeletal involvement.
For example, the greater incidence of slight and healed
lesions that were observed in these groups are considered
to be those skeletal changes which would be expected to
occur in thalassemic heterozygotes. Angel (1967)
suggested that the low incidence of severe porotic
hyperostosis was due to an early and intense selection
against individuals that were homozygous for the
condition. Thus, the marked skeletal changes that are often
observed in modern thalassemic homozygotes are
attributed to modern medical intervention which has
allowed affected individuals to survive for longer periods
of time. As a consequence, these individuals are assumed
to sustain more pronounced degrees of osseous tissue
hypertrophy than would have been the case for pre-modem
thalassemic homozygotes. In addition, cortical thinning of
the long bones and ribs are cited as post-cranial evidence of
erythroid marrow hyperplasia which occurred in response
to the hemolytic anemia of thalassemia in these earlier
human groups (Angel, 1966; Angel, 1967).
However, some more recent workers concerned with the
origin of porotic hyperostosis in Old World circum-
Mediterranean archeological populations consider the
skeletal evidence for thalassemia to be highly suspect
(Ascenzi and Salistreri, 1977). As an alternative, it is
suggested that the lesion may be the result of compensatory
erythroblastic anemia which occurs in association with
malarial infection alone (Germana and Ascenzi, 1980).
The etiology of porotic hyperostosis in New World
skeletal groups has likewise been attributed to the
hemolytic anemias of thalassemia or sickle cell anemia
(Wakefield et al., 1937; Zaino, 1964; Zaino, 1967; Zaino,
1968; Zaino and Zaino, 1969). Evidence in support of this
view is weak, and is primarily limited to the observation
that a greater incidence, and degree of severity, of the
lesion characterizes several New World skeletal groups (for
discussion see Moseley, 1965). Angel (1967) considered
iron deficiency anemia, primarily as a result of hookworm
infestation, to be a more probable cause of porotic
hyperostosis in New World populations. Paleo-
epidemiological research discussed thus far has focused on
a general relationship whereby some form of parasitism
(i.e., falciparum malaria or hookworm infestation) has been
regarded as the principal factor involved in the etiology of
porotic hyperostosis that occurred in Old and New World
skeletal populations alike.
The major alternative hypothesis is that the majority of
porotic hyperostosis that has been observed in human
skeletal groups occurred as a result of iron deficiency
anemia in direct response to elevated levels of infectious
disease and nutritional stress (Moseley, 1965; Moseley,
1966; Hengen, 1971; Carlson et al., 1974; El-Najjar et al.,
1975; El-Najjar and Robertson, 1976; El-Najjar et al., 1976;
Lallo et al., 1977; Mensforth et al., 1978; Schutte, 1979;
Cassidy, 1980; Von Endt and Ortner, 1982; Palkovich,
1985, 1987; Stuart-Macadam, 1985, 1987a, 1987b, 1989;
Walker, 1986; Hodges, 1987; Fairgrieve, 1990).
Considerable evidence supports this inference. First,
contemporary epidemiological studies have demonstrated
that iron deficiency anemia is the most common nutritional
deficiency that affects human groups on a world-wide basis
(Witts, 1966; World Health Organization, 1968; Robbins,
1974; Baker, 1978; Betts and Weidenbenner, 1986). Porotic
hyperostosis likewise has a very widespread distribution in
Old and New World skeletal populations (Moseley, 1965;
Hengen, 1971; El-Najjar et al., 1976; Lallo et al., 1977).
Second, the incidence of iron deficiency anemia in modem
groups, and the frequency of porotic hyperostosis in
skeletal groups, is greater for those populations that subsist
(or subsisted) on dietary staples that are low in bioavailable
iron (Carlson et al., 1975; El-Najjar and Robertson, 1976;
Lallo et al., 1977; Martinez-Torres and Layrisse, 1974;
Witts, 1966; Jeliffe and Blackman, 1962; Scrimshaw and
Young, 1976; Grantham-McGregor et al., 1974; Burks et
al, 1976; and Ashworth et al., 1973). Third, no evidence
exists to support the view that any of the hemolytic
anemias which arose in response to falciparum malaria in
the Old World were operative as selective factors in the
pre-Columbian New World (El-Najjar, 1976 and references
therein). Nonetheless, hookworm infection most likely
contributed to a greater prevalence and incidence of iron
deficiency anemia in certain Old and New World human
groups (Hengen, 1971).
Benefits and Liabilities of Porotic Hyperostosis Studies
Paleoepidemiological analyses that have employed
samples of sufficient size to investigate porotic
hyperostosis in earlier human groups have demonstrated a
6
Mensforth
No. 46
marked concordance with the iron deficiency anemia
hypothesis (El-Najjar, 1976; Lallo et al., 1977; Mensforth
et al. , 1978; Walker, 1986; Stuart-Macadam, 1989). For
example, El-Najjar (1976) examined the frequency of
occurrence for porotic hyperostosis in 4,146 individuals
from fourteen New World skeletal groups. This
macrosample was further subdivided into maize-
dependent (n= 1,722) and non-maize-dependent (n= 2,424)
groups. The rank order frequency of occurrence for porotic
hyperostosis in the maize-dependent sample was subadults
(31.7%), adult females (25.3%), and adult males (21.6%).
Likewise, the rank order frequency of occurrence for
porotic hyperostosis in the non-maize-dependent sample
was subadults (8.6%), adult females (5.2%), and adult
males (2.0%).
Thus, two patterns emerge that are in accord with the
iron deficiency anemia hypothesis. First, both subsamples
exhibit a rank order frequency of porotic hyperostosis
which identifies those age/sex groups at greatest risk of
acquiring iron deficiency as a result of intrinsic
physiological factors alone. Second, the markedly higher
incidence of porotic hyperostosis that characterizes the
maize-dependent skeletal groups is a pattern which would
be expected to occur as a consequence of diets that are low
in bioavailable iron.
However, there are theoretical biases and method-
ological problems which limit the inferential utility of
studies such as the one described above. The first problem
concerns the relationship between the frequency of porotic
hyperostosis and levels of morbidity and mortality that
characterize our comparisons of prehistoric human groups.
We may find that one skeletal groups exhibits a frequency
of porotic hyperostosis that is statistically significantly
greater than another skeletal group. However, differences
in lesion frequency alone do not allow us to assess the
demographic impact that such disease loads may have had
on earlier human groups. That is, if morbidity is high and
mortality is low, than many individuals who survive
periods of elevated risk would be expected to exhibit
lesions that had undergone various stages of healing or
remodeling. Alternatively, certain high risk age/sex groups
may display elevated frequencies of active lesions at time
of death. Thus, before we accept porotic hyperostosis as a
useful bioassay of disease and nutritional stress we must
first determine whether or not differences in the frequency
of the skeletal lesion correlate with differences in mortality
rates that are observed. If a positive correlation exists, we
must then evaluate the strength of such a relationship. If
the relationship is strong, we can then ascribe demographic
sensitivity, and evolutionary significance, to the non-
specific stress indicator.
The second problem that characterizes many
paleoepidemiological inquiries is an over emphasis on
dietary hypotheses. Thus, it is routinely assumed that the
nutritional status of prehistoric horticultural and
agricultural groups will be impaired by the ingestion of
domesticated cultigens. It is well recognized that maize,
and other Old and New World cereal grains, are low in
bioavailable iron and other essential nutrients (Scrimshaw
and Young, 1976). However, we often overlook floral and
faunal evidence which indicate that many prehistoric
horticultural and agricultural groups supplemented their
diets with substantial amounts of nutrient rich plant and
animal foods that were acquired via foraging, hunting,
and/or trade. Furthermore, it is very probable that few
prehistoric groups practiced the intense levels of
monocropping that occur in many contemporary under-
developed societies where levels of sanitation and hygiene
are poor, population density is high, and the incidence of
protein deficiency and protein-calorie malnutrition is
exceptionally high (Scrimshaw and Young, 1976; Beisel,
1982; Herbert, 1985; Herbert, 1987).
Perhaps equally important as diet are the demographic
changes that occurred in human groups which made
widespread use of domesticated cultigens. Overwhelming
empirical evidence has shown that increases in local and
regional population densities, combined with the hygiene
and sanitation problems of sedentary village life, results in
a significant rise in the frequency, duration, and severity of
infectious diseases in agricultural communities
(Scrimshaw and Young, 1976; Herbert, 1985, and
references therein). Moreover, several researchers have
documented that half of all nutritional crises are
precipitated by infectious disease episodes alone (Gordon
et al., 1963; Scrimshaw and Suskind, 1959; Maynard and
Hammes 1970; Martorell, 1980). Thus, the extent to which
infectious diseases may have contributed to the increased
incidence of porotic hyperostosis that occurred in certain
prehistoric human groups has not received due
consideration in light of its demonstrated importance
(Lallo et al., 1977; Mensforth et al., 1978).
The primary methodological problems which limit the
utility of many porotic hyperostosis studies are ( 1 ) inadequate
demographic documentation for the samples being
investigated, (2) routine use of broad age intervals for
reporting skeletal lesion frequencies, and (3) lack of
information about lesion activity status (i.e., active versus
healed lesions) at time of death for affected individuals. First,
it is important to realize that many human skeletal collections
that were assembled during the earlier part of the 20th
century are demographically unsound as a result of biased
sampling techniques. These include selective recovery of (a)
cranial versus post-cranial skeletal remains, and (b) adults
versus subadults. Furthermore, the growing interest in human
skeletal diseases that fascinated physical anthropologists at
the turn of the century may have introduced a further bias
1991
Paleoepidemiology oe Porotic Hyperostosis
7
whereby pathological specimens were recovered in greater
numbers relative to non-pathological individuals. The extent
to which the latter bias may affect many extant human
skeletal samples is difficult to assess in retrospect.
Nonetheless, it is quite clear that the demographic
composition of many skeletal collections is markedly
unbalanced. Thus, the quality of age and sex information that
has been reported for such materials remains highly suspect
(Weiss, 1972; Weiss, 1973; Ruff, 1981).
Second, paleoepidemiologists have often reported
skeletal lesion frequencies for very broad, and non-
standardized, age groups (i.e., subadults, adult males, and
adult females). The use of broad age intervals not only
masks useful age-specific patterns in the frequency of
occurrence of skeletal lesions, but also masks the effects of
census errors. These affects are usually most pronounced
for subadults where infant underenumeration can markedly
skew the skeletal lesion frequencies that are observed in
such groups. Thus, erroneous conclusions about differential
health status may arise when skeletal lesion frequencies for
demographically balanced skeletal samples are compared
to the skewed lesion frequencies of selectively biased
skeletal samples. Moreover, the width in years that is used
to define the subadult age interval varies from one
paleoepidemiological report to the next. Therefore, lesion
frequencies that are given for subadults grouped in 0-10
year, 0-15 year, and 0-20 year intervals must be adjusted
before meaningful comparisons of health status can be
made. However, the information required to carry out such
adjustments is all too often lacking in the original reports.
Third, most paleoepidemiological studies make no
attempt to discriminate between skeletal lesions that are
active versus those which are inactive at time of death.
This is peculiar given the fact that unremodeled lesions
provide us with our only reasonable estimates of the age-
specific frequency of occurrence, and levels of mortality,
which may have occurred as a direct, or indirect, result of
the disease process which produced the skeletal lesion. The
total lesion frequencies most often reported are combined
measures of unremodeled and remodeled lesions that were
observed in a group. These may provide us with a general
index of overall morbidity that accompanied a particular
stress indicator. However, combined measures which
include remodeled lesions may also introduce a significant
amount of noise. That is, our ability to estimate the real
total frequency with which a skeletal lesion occurred in an
earlier human group will be distorted by age progressive
bone remodeling. The latter may be further confounded by
a sex differential in bone remodeling rates as is the case
with Harris lines (see Garn, 1968). In general, these effects
will be most pronounced in adults where the (a) frequency
of unremodeled lesions that are observed is likely to be
lower than the frequency of total lesions that are observed.
and (b) the frequency of total lesions that are observed is
likely to be lower than the frequency of total lesions that
occurred in the group.
Significance of the Iron Deficiency Anemia Hypothesis
Contemporary epidemiological surveys have shown that
iron deficiency anemia is so common in human groups
throughout the world that the prevalence of iron deficiency
is now regarded as one of the best indices of the nutritional
health status of a population (Witts, 1966; World Health
Organization, 1968; Kilpatrick, 1970; Baker, 1978; Wen-
guang et al., 1986). The biological significance of iron
resides in the fact that it is required to sustain important
physiological processes. These include hemoglobin
synthesis, tissue respiration, enzyme activity, and
oxidative-reduction reactions (Macdougall et al., 1975).
Iron also plays a paramount role in the maintenance of host
immunological competence (Prasad, 1979), and the
maintenance of normal epithelial tissue structure and
function (Naiman et al., 1969). Furthermore, studies have
shown that an individual’s ability to maintain adequate
body tissue iron stores and dietary absorption of the
nutrient are directly related to health status (World Health
Organization, 1968 and references therein). Therefore, the
intrinsic physiological and extrinsic environmental factors
that are involved in the etiology of iron deficiency anemia
will be considered.
Intrinsic Physiological Risk Factors that Elevate Risk
of Iron Deficiency Anemia
Clinical studies have shown that iron deficiency anemia
occurs most frequently in those age and sex groups of a
population where physiological demands for the nutrient
are greatest (Heath and Patek, 1937; Hallberg et al, 1970;
Robbins, 1974; Olivares et al., 1986). Thus, four high risk
groups can be identified in any human group. These can be
further subdivided into two major and two minor high risk
groups. The two groups at greatest risk of acquiring iron
deficiency anemia are (1) infants and young children that
range in age from 6 months to three years, and (2) adult
females in their child-bearing years (Finch, 1968a, and
1968b). The two minor high risk groups are (1) adolescent
males and females, and (2) men and women over 60 years
of age (Scott et al., 1970).
Birthweight and rate of somatic growth are now
recognized as the two most important constitutional factors
which elevate the risk of iron deficiency anemia in infants
and young children (Josephs, 1956; Jacobs and
Wormwood, 1982; Laskari, 1984; Betts and Weidenbenner,
1986). Infants are born with iron concentrations that are
proportional to body weight, and iron stores established at
birth are the most important source of the nutrient for the
first six months of life (Mackay, 1931; Mackay, 1933).
Mensforth
No. 46
Indeed, in otherwise healthy full-term infants a rapid rate
of somatic growth alone often results in significant tissue
iron depletion by six months of age (Josephs, 1956).
Thereafter, exogenous sources of iron are required in order
to meet the infant’s physiological demands for the nutrient.
Thus, it is not remarkable that the combined effects of
(1) rapid somatic growth, (2) frequent infections, and (3)
weaning diets of poor nutritional quality often give rise to a
high frequency of iron deficiency anemia in infants and
young children (Josephs, 1936; Josephs, 1953; Betke,
1970; Smith, 1972; Scrimshaw and Young, 1976). In
modern well-nourished societies pediatricians often
consider iron deficiency anemia in infants to be a
transitional state in which the child is more anemic than ill
(Sturgeon, 1956; Betke, 1970; Smith, 1972; Thomas et ai,
1977). In developing countries, however, the condition is
often much more prevalent, severe, and contributes
significantly to elevated levels of subadult morbidity and
mortality (Akel et al. , 1963; Manchandra et ai, 1969;
Maynard and Hammes, 1970; Ashworth, 1973; Grantham-
McGregor et ai, 1974; Burks et ai, 1976).
With regard to adult females, pregnancy is the major
physiological circumstance that elevates risk of acquiring
iron deficiency anemia (Witts, 1966; Yusufji et ai , 1973).
Population studies have reported a world-wide prevalence
of nutritional anemias in pregnancy that range between
fifteen and ninety-nine percent (Witts, 1966; World Health
Organization, 1968; Finch et ai, 1968; Scott et a!., 1970).
Hunter’s ( 1960) survey of nutritional anemias in pregnancy
further showed that ninety-eight percent were the direct
consequence of iron deficiency, with folate and B12
deficiencies occupying a minor role in overall etiology.
The important intrinsic factors which interact to increase
the risk of iron deficiency anemia during pregnancy
include ( 1 ) low tissue iron stores at the onset of pregnancy
due to the cumulative effects of regular menstrual iron
losses, (2) the physiologic hydremia of pregnancy where
expanding plasma volume has the net effect of decreasing
maternal hemoglobin concentration, packed red cell
volume, and red cell count, (3) maternal excretion of iron
in urine and sweat, (4) fetal and placental iron
requirements, and (5) blood loss at parturition (Finch,
1968a; Finch, 1968b; Pritchard and Scott, 1970).
More specifically, the third trimester fetal iron demands
are so great that maternal tissue iron stores rapidly become
depleted (Yusufji et ai, 1973). At this time maternal iron
balance becomes highly dependent upon dietary absorption
which may increase three to four times above normal in
order to meet excessive maternal and fetal iron
requirements (Apte and Iyengar, 1970). These observations
are concordant with the finding that iron deficiency anemia
in pregnant women is most prevalent during the third
trimester (Yusufji et ai, 1973). In addition, the period of
lactation following parturition is accompanied by additional
iron requirements that amount to 0.5- 1.0 mg. iron per day
above the normal 1.5 mg. of iron per day that is normally
absorbed from the diet (Finch, 1968a; Finch, 1968b).
Among the two minor population age/sex groups that
are at elevated risk of acquiring iron deficiency anemia, the
principal risk factor in adolescents is accelerated rate of
somatic growth (Saddi and Schapira, 1970; Kenney, 1985;
Liebman, 1985). The increased tissue iron requirements
that accompany the adolescent growth spurt can rapidly
deplete body iron stores and establish a latent iron deficient
state. In addition, the onset of menarche and subsequent
menstrual blood losses contribute to a sex differential
whereby adolescent females experience a higher incidence
of iron deficiency anemia relative to males (Rybo, 1970;
Saddi and Schapira, 1970). Finally, individuals of both
sexes that are over sixty years of age are at elevated risk of
developing iron deficiency anemia due to age progressive
degeneration in the absorptive capacity and efficiency of
the gastrointestinal tract (Finch, 1968a; Finch, 1968b).
Extrinsic Factors Which Elevate Risk of Iron
Deficiency Anemia
Diet. — The absorption of dietary iron is dependent upon
the biochemical properties of food items that are ingested,
and the organism’s physiological controls which are
designed to maintain a conservative equilibrium (Conrad,
1970; Davis, 1970). In humans, iron is absorbed primarily
in the duodenum, the most alkaline portion of the
gastrointestinal tract (Davis, 1970). In order to meet
nutritional requirements, iron must be exposed to the
mucosa of the small intestine in a soluble state, in
reasonable quantities, and for a sufficient length of time
(Conrad, 1970). Under normal circumstances the body
exerts rigid control over iron homeostasis through
conservation, reutilization, and by regulating the processes
by which iron losses are replenished (Conrad, 1970). For
example, healthy infants absorb approximately ten percent
of available dietary iron (Heinrich, 1970). In contrast, iron
deficient infants have the capacity to absorb two to three
times the normal amount as a means of compensating for
iron loss and tissue iron depletion (Andelman and Sered,
1966; Saddi and Schapira, 1970).
Studies concerned with the bioavailability of dietary
iron have demonstrated that the iron content of food varies
tremendously (Wretlind, 1970; Martinez-Torres and
Layrisse, 1974; Morck et ai, 1981; Palazzari et ai, 1986).
Even among similar food items the iron content may vary
depending on how the food is prepared and, for vegetable
food items, where it is grown (Bressani, 1958; Cook and
Monsen, 1976). For example, heme iron is more readily
absorbed than ferrous iron, and the latter is better absorbed
than ferric iron (Callender et ai, 1957; Hallberg and
1991
Paleoepidemiology of P orotic Hyperostosis
9
Sovell, 1967; Turnbull et al, 1967). Also, meat products
generally contain more iron that is in a readily absorbable
form (Layrisse et al, 1968; Layrisse et al, 1969; Martinez-
Torres and Layrisse, 1971).
The bioavailability of iron is also dramatically
influenced by other dietary constituents such as chelating
agents (Hwang and Brown, 1965; Kuhn et al, 1968; Davis,
1970). These compounds can either promote or inhibit iron
absorption. Ascorbic acid is a chelating agent which
promotes iron absorption by producing a water soluble iron
complex (Moore et al, 1940). Several sugars and amino
acids also facilitate iron absorption by decreasing the
precipitation and polymerization of dietary iron (Charley
et al, 1963; Pollack et al, 1964; Martinez-Torres and
Layrisse, 1970). In contrast, compounds such as phytates,
phosphonates, carbonates, and oxylates strongly inhibit the
absorption of dietary iron by effectively binding iron into
insoluble macromolecules (Hegsted et al, 1949; Sharpe et
al, 1950; Foy et al, 1959; Hussain and Patwardhan, 1959;
Conrad, 1970).
Qualitatively superior and inferior dietary regimens
have been reported in association with iron deficiency
anemia in children (Davidson et al, 1935; Josephs, 1956;
Dawson and Desforges, 1958; Woodruff, 1958). In general,
a consistent relationship between anemia and artificial, or
prolonged, milk feeding has been demonstrated in modem
groups (Mackay, 1931; Fullerton, 1937; Smith, 1972).
Epidemiological studies have also confirmed that
prolonged breast feeding and weaning diets of maize or
com gruels often occur in association with a high incidence
of iron deficiency anemia in infants and young children
(Ashworth, 1973; Jelliffe and Blackman, 1962; Grantham-
McGregor et al, 1974). These findings are attributed, in
part, to the high phosphorous content in milk and com, as
well as the high concentration of phytic acid in com, which
inhibits the absorption of dietary iron (Lanzkowski and
McKenzie, 1959; Martinez-Torres and Layrisse, 1974).
Another factor to be considered with regard to
nutritional status, in infants and children in particular, is the
extent to which acute and chronic gastrointestinal
infections lead to the malabsorption of dietary iron and
other essential nutrients (Conrad, 1970; Carpenter and
Sack, 1981; Gryboski and Walker, 1983; Santos, 1986).
Diarrheal episodes promote malabsorption by increasing
intestinal motility (Gordon et al, 1963; Fagundes-Netto,
1984). Therefore, both the quantity of iron and the amount
of time that it is made available to the absorptive surfaces
of the intestinal tract are decreased. In addition, diarrheal
episodes are accompanied by dehydration, electrolyte
imbalance, negative nitrogen balance, loss of appetite, and
the substitution of solid foods by starchy gruels of lower
nutritional quality (Gordon et al, 1963; Scrimshaw, 1964).
Moreover, tissue iron depletion alone appears to be an
important factor which promotes malabsorption syndrome
in iron deficient individuals (Naiman, 1969).
Severe degenerative epithelial tissue changes (i.e.,
glossitis, stomatitis, and koilonychia) are not commonly
observed in iron deficient adults unless the condition
persists for an indefinite period of time (Halsted et al,
1965; Yusufji et al, 1973). However, iron deficient infants
and children are reported to experience a relatively high
incidence of gastrointestinal dysfunction that results in
malabsorption syndrome (Hawksley et al, 1934; Wilson et
al, 1962; Halsted et al., 1965; Guha et al, 1968; Naiman
et al, 1969). Tissue iron depletion alone is a major factor
responsible for defects in epithelial tissue structure and
function that occurs in iron deficient children (Naiman et
al, 1969). The most important cytological defects include:
atrophy of the gastric and duodenal mucosa, diminished
synthesis and secretion of gastric acid, and reduced
gastrointestinal enzyme activity (i.e., mucosal disacharidase
and cytochrome oxidase activity) (Guha et al, 1968;
Halsted et al, 1965; Mahoney and Hendricks, 1975).
Atrophy of the gastric mucosa is usually the most frequent
and pronounced tissue change that occurs in iron deficient
children (Davidson and Markson, 1955; Badenoch et al,
1957; Rawson and Rosenthal, 1960). The functional
consequence of atrophic gastritis is a reduction in gastroferrin
synthesis and gastric acid secretion (Stewart, 1937; Shearman
et al, 1966; Ghosh et al, 1972; Smith, 1972). The latter
abnormality has important implications for those individuals
who already exist in an overt state of iron deficiency. Under
normal circumstances gastric acidity promotes the absorption
of dietary iron and calcium by preventing the formation of
insoluble macromolecules (Cook et al., 1964; Mahoney and
Hendricks, 1975). The altered gastric pH which occurs more
commonly in iron deficient children exacerbates the
circumstance by further reducing the amount of dietary iron
that is available to the host in a soluble form which can
readily be absorbed in the duodenum.
Gastric mucosal atrophy also effects erythropoiesis by
interfering with the synthesis and secretion of intrinsic factor
(Guyton, 1976). This compound is produced by parietal cells
of the gastric mucosa and plays a significant role in
facilitating the absorption, and preventing the digestion, of
vitamin B j 2 in the intestinal tract. Vitamin B 1 2 is required for
red blood cel! maturation. In the absence of intrinsic factor
vitamin B j 2 absorption is impaired. The process of
hemoglobin synthesis is not affected, but a macrocytic
anemia is a common consequence (Guyton, 1976).
Though less pronounced, atrophy of the duodenal
mucosa and reduced enzyme activity of the small intestine
promotes the malabsorption of fats, some carbohydrates,
and contributes to occult blood loss in iron deficient
subjects (Wilson et al, 1962; Naiman et al. , 1969).
Furthermore, iron deficiency anemia in children is regarded
10
Mensforth
No. 46
as one of several protein-losing enteropathies (Lahey,
1962). Pathological effects include malabsorption of
dietary amino acids and leakage of plasma proteins in the
GI tract. Thus, cytological damage to the GI tract which
occurs in iron deficient children has a negative impact on
nutritional status in general, and further reduces the
bioavailability of dietary iron at a time when physiological
demands for the nutrient are great.
Infectious disease: impaired immune response and
nutritional immunity. — The synergistic relationships
between iron deficiency anemia and infectious disease has
received considerable attention in recent years. Clinical and
experimental studies have repeatedly demonstrated that iron
deficient humans and laboratory animals experience a
greater incidence and severity of infectious disease episodes
compared to normal healthy subjects (Shaw and Robertson,
1964; Werkman et al., 1964; Kilpatrick, 1970; Baggs and
Miller, 1975). In humans, this relationship is particularly
marked for children under five years of age (Andelman and
Sered, 1966; Arbeter et al. , 1971; Joynson et al., 1972;
Chandra, 1973; Chandra and Saraya, 1975; Macdougall et
al., 1975; Scrimshaw and Young, 1976; Krantman et al.,
1982; Chandra, 1985; Walter et al, 1986). Indeed, the high
incidence of respiratory and gastrointestinal infections that
occur in iron deficient infants and children significantly
contribute to growth retardation and elevated levels of
subadult morbidity and mortality in many contemporary
under-developed societies (Witts, 1966; Gordon et al.,
1967; Jose and Welch, 1970; Maynard and Hammes, 1970;
Krantman et al, 1982; Forman et al, 1984; McMurray,
1984; Hercberg et al., 1986). Thus, the interactions of
anemia and infection have important demographic
consequences for populations that are subjected to
intensified levels of disease and nutritional stress.
One of the most significant findings in recent years
concerns the fact that impaired immune response in an
early manifestation of tissue iron depletion (Chandra,
1973, 1985; Macdougall et al, 1975; Prasad, 1979). The
principal implication is that host resistance to infectious
disease is compromised prior to the onset of overt iron
deficiency anemia. Studies have shown that both (1) cell
mediated immunity and (2) bactericidal capacity of
leukocytes are suppressed in iron deficient subjects
(Arbeter et al., 1971; Macdougall et al, 1975; Srikantia
et al, 1976; Weinberg, 1977; Root and Cohen, 1981).
Though similar defects in immune response occur in
protein-calorie malnutrition, the site of the biochemical
defect in iron deficiency differs and does not appear to be
directly related to protein deprivation (Macdougall et al.,
1975; Purtillo and Connor, 1975; Bhaskaram and Reddy,
1975; Bhaskaram et al., 1977; Prasad, 1979; Hoffman-
Goetz and Kluger, 1979; Chandra and Au, 1980; Gross
and Newberne, 1980).
Suppressed leukocyte function in iron deficient
subjects is a direct consequence of tissue iron depletion
(Chandra, 1973, 1985). One proposed defect involves
myelo-peroxidase (MPO) synthesis (Baggs and Miller,
1975; Yetkin et al, 1979; Parry et al, 1981). MPO is an
enzyme that is dependent on heme iron for its activity. It
plays a substantial role in the metabolic pathway that
mediates the phagocytic capacity of polymorphonuclear
(PMN) leukocytes. Simply stated, certain phagocytic cells
of the mammalian reticuloendothial system (i.e.,
eosinophils, neutrophils, and macrophages) require MPO
to synthesize hydrogen peroxidase. The latter enzyme is
involved in the intracellular processes of lysosomal
pathogen killing (Baggs and Miller, 1975). The reduction
in leukocyte MPO concentrations, and the subsequent
decrease in leukocyte bactericidal activity, appears to be a
major factor contributing to the high incidence and
greater severity of infections that have been observed in
iron deficient children (Arbeter et al, 1971; Macdougall
et al., 1975).
Recent research has also demonstrated that reduced
quantities of the iron-containing enzyme ribonucleotide
reductase, and other cytochrome heme enzymes, impair
DNA and protein synthesis in iron-deficient subjects
(Joynson et al., 1972; Jacobs and Joynson, 1974; Beisel,
1982). The resultant effects include suppressed (a)
lymphocyte production, (b) cytotoxic T lymphocyte
activity, and (c) neutrophil ferritin and lactoferrin
production (Prasad, 1979; Root and Cohen, 1981; Beisel,
1982; Kuvibidila et al., 1981, 1983a, 1983b; McMurray,
1984). Thus, impaired immune response is a direct, and
early, consequence of tissue iron depletion that compro-
mises host resistance to infectious disease.
In addition, the competitive relationship between host
and microbe iron requirements, and the subsequent host
response to infectious episodes, can substantially reduce
the amount of iron that would normally be available to the
host for the maintenance of physiological processes. It is
now well recognized that iron is a prime nutrilite that is
required by bacterial and viral pathogens in order to
survive and multiply in mammalian host tissues
(Brendstrup, 1950; Weinberg, 1966; Weinberg, 1974).
Moreover, many pathogens have evolved the capacity to
synthesize and secrete siderophores (Weinberg, 1966;
Garibaldi, 1972; Weinberg, 1977; Kochan, 1977b). These
are iron-binding compounds which enable the microbe to
compete effectively with host iron-binding proteins and
host tissues for essential iron (Weinberg, 1977 and
references therein). Studies have demonstrated that both
the (1) rate of growth and (2) virulence of bacterial
pathogens are directly related to the amount of free iron
that is available to the microbes (Martin et al, 1963;
Shade, 1963; Kaye and Flook, 1963; Brubaker et al, 1965;
1991
Paleoepidemiology of Porotic Hyperostosis
Weinberg, 1966; Bullen and Rogers, 1968; Polk and Miles,
1971; Kochan, 1977a, 1977b, 1978; Kluger and
Rothenburg, 1979; Hoffman-Goetz el al. , 1981;
Hoshishima et al, 1985).
Under ordinary circumstances, iron-binding proteins
play a critical role in regulating the amount of free iron that
is available to pathogens (Weinberg, 1974). Serum proteins
such as transferrin, and large concentrations of lactofeirin
contained in human breast milk, exert a strong
bacteriostatic effect on microbial growth (Bullen and
Rogers, 1968; Bullen et al. , 1968; McFarlane et al., 1970;
Fletcher, 1971; Bullen et al., 1972; Hanson and Winberg,
1972; Purtillo and Connor, 1975; McFarlane, 1976; Faulk,
1976). This bacteriostatic effect is greatly diminished, or
completely abolished, in individuals with severe protein
deficiency where cause of death is most often due to
overwhelming infections (McFarlane and Hamid, 1973).
Iron-binding proteins are therefore considered to be an
important non-specific factor in host resistance to
infectious disease.
The hypoferremia which is induced by host response to
local and systemic infections also plays a key role in iron
economy (Grieger and Kluger, 1978). It has long been
recognized that infectious disease episodes are
accompanied by a dramatic reduction in serum iron
concentrations (Cartwright et al. , 1946; Greenberg et al,
1947; Kuhns et al. , 1950). This effect is mediated by
plasma transferrins which bind free iron and store it in the
reticuloendothelial system (i.e. , liver, spleen, and bone
marrow) for the duration of the infection (Cartwright et
al., 1946; Vannotti, 1957; Kochan, 1978). The
physiological response whereby the host induces a
hypoferric state in an attempt to deprive the pathogen of
necessary iron has been termed nutritional immunity by
Weinberg (1974). Therefore, iron which would normally
be mobilized for hemoglobin synthesis is sequestered
from both host and pathogen for the duration of the
infectious episode.
Mild infections of short duration usually have no
significant effect on hemoglobin levels in adults (Kuhns et
al, 1950; Chandra, 1985). If the infection persists a
normochromic normocytic, or slightly hypochromic
microcytic, anemia can develop. This is commonly referred
to as the anemia of infection (Cartwright et al, 1946).
However, among infants and children both mild and severe,
acute and chronic, infections can result in a marked
decrease in hemoglobin levels (Mackay, 1933; Davidson
and Fullerton, 1938; Manchandra et al, 1969). In effect,
erythropoiesis is suppressed and hemoglobin synthesis is
markedly inhibited (Cartwright et al. , 1946). Furthermore,
the regain of normal erythropoietic activity following
recovery from an infectious episode can be delayed for
several weeks to several months in young children
(Fullerton, 1937). Thus, respiratory and gastrointestinal
tract infections that occur during periods of rapid growth
can interfere with the bioavailability of iron and
precipitate, or exacerbate, an iron deficient state. Iron
deficiency anemia may ensue even though the amount of
iron available in the diet is more than adequate.
Some clinical researchers have challenged the concept of
nutritional immunity and have questioned the adaptive
significance of hypoferremia due to infectious episodes
(Srikantia et al, 1976; Prasad, 1979). These researchers
emphasize the fact that impaired immune response is an
early consequence of tissue iron depletion. It is argued that
a hypoferric state would further compromise immunological
competence of the host. However, these investigators have
failed to consider that during infectious episodes iron,
particularly heme iron, is diverted to those tissues of the
immune system, the reticuloendothelial system (RES),
which require it most in order to maintain, or enhance, cell-
mediated immunity, phagocytic activity, and the bactericidal
capacity of PMNs.
Studies have shown that infections inhibit RES
erythropoietic activity and hemoglobin synthesis (Kuhns
et al, 1950 and references therein). At the same time,
however, infections stimulate RES myeloid tissues
(Cartwright et al, 1946; Smith, 1972). The implication is
as follows. Although heme iron may not be available for
hemoglobin synthesis, it may well be used to promote
protein synthesis, lymphocyte production and cytotoxic
competence, and leukocyte phagocytic activity and
bactericidal capacity (Vannotti, 1957; Kochan, 1978).
Viewed in this perspective, host response to infectious
disease (i.e., hypoferremia) and its subsequent effects on
the differential bioavailability of iron appear to be
complimentary physiological adaptations.
Cultural practices and parasitism. — In addition to
constitutional factors, diet, and microbial infection, a
number of cultural and environmental variables are also
known to play a role in the etiology of iron deficiency
anemia (Shah and Seshadri, 1985). For example, Gordon
and associates (1963) have shown that (1) culturally
prescribed weaning practices and (2) age and sex specific
food restrictions and/or taboos, may contribute to patterns
of chronic malnutrition in certain human groups. Likewise,
an important environmental variable in tropical and
subtropical regions of the world is parasitic hookworm
infestation (Roche and Perez-Gimenez, 1959; Bradfield et
al, 1968; Venkatachalam, 1968). This promotes intestinal
blood loss which can lead to chronic iron deficiency
anemia in some individuals.
With respect to trends in morbidity and mortality, it is
worthy of comment that adults and subadults generally
tolerate uncomplicated iron deficiency anemia rather well.
In children the anemic condition may further operate as a
12
Mensforth
No. 46
homeostatic mechanism that functions to balance iron
metabolism during developmental periods of fluctuating
supply and demand. Such a physiological homeostat has a
strong selective value for rapidly growing organisms that
must constantly balance iron economy with respect to (1)
nutritional requirements and bioavailability on the one
hand, and (2) host nutritional immunity in response to
infectious diseases on the other. Arbeter and associates
(1971) have expressed a similar view with respect to the
iron deficiency anemia that accompanies protein
deficiency. These workers suggest that the anemia may be
an adaptation to the lowered metabolism of protein-
deprived tissues.
The epidemiological relationships discussed thus far can
be summarized as follows:
1) Although prevalence may vary, iron deficiency
anemia occurs in all human groups, and is the
single most common nutritional disorder that
affects humans on a world-wide basis.
2) Intrinsic physiological risk factors play a major
role in predisposing certain age/sex groups to iron
deficiency anemia. The age/sex groups at greatest
risk are infants and young children six months to
three years of age, and adult females in their peak
period of fertility. Those age/sex groups at lesser
risk are adolescents, females in particular, and
elderly individuals.
3) Infants and young children represent those
individuals at greatest risk of elevated morbidity
and mortality as a consequence of physiological
dysfunctions accrued in association with iron
deficiency anemia. These include impaired
immune response and malabsorption syndrome
which may occur as a result of marked tissue iron
depletion alone.
4) The prevalence of iron deficiency anemia is
sensitive to differences in extrinsic environmental
factors that vary in human groups. Thus, a high
incidence of iron deficiency anemia has repeatedly
been observed in societies that (a) subsist on diets
low in bioavailable iron, and (b) experience
elevated levels of infectious diseases.
Thus, it is clear that iron deficiency anemia has a
measurable impact on the health status and demographic
characteristics of extant human populations. Similarly, if
porotic hyperostosis in earlier human groups occurred as a
primary consequence of iron deficiency anemia, it would
be reasonable to infer that strong demographic correlates
exist which argue in favor of the skeletal lesion’s utility as
a bioassay of disease and nutritional stress.
Hypothetical Model of Skeletal Lesion Differential
Sensitivity
Paleoepidemiologists are clearly aware that
hematological data on the living provide the most accurate,
and sensitive, measure of the prevalence, morbidity, and
mortality associated with iron deficiency anemia in extant
human groups. Skeletal lesions, by comparison, must
always be regarded inferior with respect to their ability to
assay such conditions in extinct human societies. The
extent to which a skeletal lesion conforms to predicted
patterns of a particular disease age/sex distribution will
thus strongly influence the degree to which such lesions
provide information useful for generating inferences about
the health status of earlier human groups.
With regard to the property of sensitivity. Figures 2a-c
illustrate the lesion frequency distributions which would be
expected for a hypothetical skeletal response that exhibited
high sensitivity, low sensitivity, differential sensitivity
based on age and/or sex, and no sensitivity to a
hypothetical disease and/or nutritional stress identified here
as Disease X. For this simple model let us assume that (a)
no sex differences characterize Disease X, (b) three discrete
age groups are at risk, and (c) those at greatest risk are
subadults, those at moderate risk are adolescents, and those
characterized by a lower risk are middle aged adults. For
this model the magnitude of Disease X is irrelevant. Hence,
frequency can represent the number of individuals affected
for a designated radix, or frequency can simply refer to the
percent of individuals affected at each age where age is in
years. Here, we are only concerned with the extent to
which our Active skeletal lesion mimics the epidemi-
ological pattern of the hypothetical disorder.
Relative to the epidemiological distribution of Disease
X , as measured by modern clinical techniques, Figure 2a
identifies skeletal lesion frequencies that display high
sensitivity (dot screen), and low sensitivity (black screen),
to the disorder. Nonetheless, in each case the pattern of
lesion distribution corresponds well with that of Disease X ,
and correctly identify those age groups at risk and the
differential magnitude of such risk. Figure 2b illustrates the
circumstance where a skeletal lesion displays differential
age-related sensitivity. Here, the Active skeletal response
(dot screen) provides a poor assay of the frequency with
which Disease X affected subadults and middle aged adults.
Nonetheless, the skeletal lesion provides a useful index of
the extent to which Disease X affected adolescents in the
population. Finally, Figure 2c illustrates the circumstances
where the skeletal lesion itself may be either frequent (dot
screen) or infrequent (black screen), but in each case the
pattern of lesion distribution bears no relationship to the
underlying epidemiological profile of Disease X.
With respect to various non-specific skeletal and dental
indicators of disease and nutritional stress that are currently
1991
Paleoepidemiology of P orotic Hyperostosis
13
AGE
in use, it is clear that many factors may synergize to
generate one, or more, of the patterns illustrated above,
compromise their sensitivity as a bioassay, and restrict their
epidemiological utility. Primary among these are ( 1 ) tissue-
FlGURE 2. Hypothetical models of differential skeletal lesion
responsiveness to a hypothetical clinical disorder (i.e., disease or
nutritional stress). Shown in 2a is the pattern we would expect to
observe for a skeletal lesion that displayed high fidelity where age
groups at risk are clearly identified. The dot screen distribution
portrays a lesion of high sensitivity to the underlying disorder,
whereas the black-screen distribution indicates a low sensitivity
(i.e., infrequent) lesion pattern. Model 2b illustrates a skeletal
lesion ( dot-screen ) that exhibits differential age-related sensitivity
to clinical Disease X. Here, the skeletal lesion is sensitive to, and
provides a useful bioassay of, the disease as it affects adolescents.
However, the lesion pattern is relatively insensitive to the
frequency with which Disease X affects subadults and middle
aged adults. Model 2c illustrates skeletal lesions which are both
common (dot-screen) and uncommon (black-screen). However,
neither pattern of skeletal lesion distribution corresponds with the
epidemiological profile of Disease X. Skeletal lesions such as
these (i.e., those which are insensitive to the underlying clinical
disorder) must be regarded as poor, or useless, bioassays of
health status in earlier human groups.
specific rates of growth, development, maturation, and
aging, (2) age and sex related differences in bone
remodeling rates, and (3) limited ability to distinguish
between active versus inactive lesion status at time of
death. Lesions which are active at time of death provide
our best measures of potential age-specific mortality
associated with a particular disorder, whereas combined
measures of active and healed lesion frequencies provide
us with our best approximation of overall morbidity for the
disorder in earlier human groups. However, differential age
and sex related bone-specific remodeling rates will distort
indices of morbidity derived from skeletal lesions. Thus,
the magnitude of such distortion, and the ways in which
paleoepidemiologists deal with such phenomena, must be
addressed for each particular skeletal lesion that is used to
assay disease response in earlier human groups.
Hypothetical Model of
Iron Deficiency-Related Porotic Hyperostosis
If it is assumed that iron deficiency anemia was the
primary factor involved in the etiology of porotic
hyperostosis in a prehistoric human group, a hypothetical
model of the expected age and sex specific frequency
distributions for unremodeled and remodeled skeletal
lesions can be posited. Figure 3a identifies the age and sex
specific frequency of occurrence that we would expect to
see for porotic hyperostotic lesions that were unremodeled
(i.e., active) at time of death. Again, depending upon the
sensitivity of the skeletal lesion to the underlying disorder,
the overall magnitudes (i.e., frequencies) of the skeletal
lesion would be expected to vary from one population to the
next due to differential interaction of intrinsic and extrinsic
risk factors. Nonetheless, it would be predicted that the ages
at onset, peak incidence, and remission would conform to
the iron deficiency anemia hypothesis and bracket those
POROTIC HYPEROSTOSIS
14
Mensforth
No. 46
(%) SNOIS31 Q313aOIAI3dNn 30 A0N3nO3d3 (%) A0N3H03H3 NOIS33 Q3"l3QOI6l3d
1991
Paleoepidemiology oe P orotic Hyperostosis
15
age/sex groups at greatest risk of acquiring the nutritional
disorder in each group. These are: A, infants and children
that are six months to three years of age; B, adolescents; C,
adult females during their peak child-bearing years; and D,
post-reproductive adult males and females.
Remodeled lesion frequency distributions shown in
Figure 3b should differ from unremodeled lesions in two
fundamental ways. First, the ages at onset, peak incidence,
and remission for remodeled lesions should display a lag
phase and translate to the right of each high risk component
in the age distribution. Second, remodeled lesion
frequencies should display continuous decay as a result of
age progressive bone remodeling.
The cumulative age and sex specific frequency
distributions for remodeled porotic hyperostosis that we
would expect to observe in a prehistoric skeletal series are
shown in Figure 3c. This is an additive representation of
information summarized in Figure 3b. Here, it is
recognized that several factors will confound our ability to
identify remodeled porotic hyperostotic lesions that
occurred in association with each high risk age/sex group.
The most important among these concerns neurocranial
bone remodeling rates. As was discussed earlier, porotic
hyperostosis is almost exclusively confined to bony
elements of the neurocranium. Growth studies have shown
that the human neurocranium achieves 95 percent of adult
size by approximately seven years of age (Malina, 1975).
Thereafter, neurocranial bone remodeling rates diminish
considerably to reach a low level throughout life.
It is thus reasonable to assume that (a) many lesions of
slight expression that occurred in the early years of life will
be completely remodeled away by adulthood, and (b) a
substantial proportion of remodeled lesions that are seen in
adults will represent stress episodes that occurred in late
infancy, childhood, adolescence, or any combination of
these. This results in a circumstance where remodeled
lesions will accumulate in young adults, even though these
lesions are continuously undergoing reduction in frequency
due to a slow rate of age progressive bone remodeling. The
remodeled lesion frequency distributions illustrated in
Figure 3c therefore represent a more realistic expectation
of the cumulative remodeled lesion distributions that would
be observed for porotic hyperostosis in earlier human
skeletal groups.
With regard to the demographic focus of the
hypothetical porotic hyperostosis model, the two most
important age/sex groups of interest to the
paleoepidemiologist are infants and young children and
adult females in their child-bearing years. Elevated levels
of subadult mortality that occur as a result of disease and
nutritional stress will directly influence both the (a) number
of individuals that survive to adulthood, and (b) mean
fertility rate that adults must achieve in order to replace the
population in succeeding generations. In addition, the latter
requirement influences the degree to which nutritional
stress may affect the reproductive performance of adult
females. It is thus reasonable to presume that such
additional compromises may further elevate the mean
fertility requirements of the group.
Background for the Libben and Bt-5 Skeletal Populations
Libben is a multi-component Late Woodland cemetery
site located in the Black Swamp on the banks of the Portage
River in Ottawa County, Ohio. The site was excavated in
1967-1968 and yielded the human skeletal remains of 1,327
individuals that ranged in age from 4 months in utero to 50+
years (Lovejoy et al. , 1977).1 Radiocarbon dates ranging
from A.D. 850 to A.D. 1250 indicate that the Libben site
was formed over a 400 year period, with primary use of the
cemetery concentrated in a 200 year span from A.D. 900 to
A.D. 1100. Based on ceramic and lithic analyses the Libben
site has been assigned to the Younge phase of the Western
Basin Tradition.
In contrast to earlier reports (Prufer and Shane, 1976;
Lovejoy et al ., 1977), no evidence presently exists to
support the conclusion that Libben was a semi-permanent
village occupied on a year-round basis. It is now clear that
the site represents one of a small number of pan-regional
cemeteries that was inhabited seasonally, and discontin-
uously, for several generations (D. Brose, personal
communication). The aboriginal peoples that created the
Libben site most likely consisted of a small number of
culturally affiliated bands which inhabited the northwestern
and southwestern shores of Lake Erie, and that placed
heavy reliance on local dietary resources (Keenlyside,
1978; Lennox, 1982; Krakker, 1983, 1984; Ferris and
Mayer, 1990). These included an abundance of freshwater
fish, small mammals, migratory birds, acorn and hickory
nuts, several species of berries (Harrison, 1978). In
addition, a small quantity of maize was recovered from the
site indicating that these peoples also indulged in
rudimentary maize horticulture, a finding concordant with
observations reported for several other Western Basin
Younge phase population groups (see Bowen, 1990; and
references therein). Nonetheless, it is quite clear maize
horticulture was not an important element in the
subsistence economy of the Libben people at this time.
Given the number of burials recovered (n=1327), the
annual crude death rate observed (CDR=.050), and a two
hundred year period of primary use, it is estimated that the
Libben cemetery was produced by a small regional
population of approximately 130 to 150 individuals
distributed among several seasonally mobile bands.
The Carlston Annis Bt-5 shell midden is a sister site of
Indian Knoll. It represents a Late Archaic habitation and
cemetery site situated on the banks of the Green River in
16
Mensforth
No. 46
TABLE 2. Comparison ofLibben and Bt-5 Age-Specific Mortality Rates
Libben Bt-5
Age Libben:Bt-5
Group
dx
>'x
dx
dx
l'x
dx
q ratio
0
226
1289
.175
76
354
.215
0.81
1
102
1063
.096
19
278
.068
1.41
3
68
961
.071
10
259
.039
1.82
5
117
893
.131
17
249
.068
1.93
10
94
776
.121
14
232
.060
2.02
15
92
682
.135
27
218
.110
1.23
20
63
590
.107
35
191
.183
0.58
25
78
527
.148
34
156
.218
0.68
30
1 15
449
.256
31
122
.254
1.01
35
154
334
.461
22
91
.242
1.90
40
97
180
.539
20
69
.290
1.86
45
50
83
.602
18
49
.367
1.64
+50
33
33
1 .000
31
31
1 .000
1.00
Total 1.289 354
Symbols: c/v, the absolute number of individuals dead at agev; / 'v, the absolute number of survivors to aget; and qx, the
probability of dying in the succeeding age class for those individuals that survive to agev.
the Western Coalfield region of Kentucky (Webb, 1950).
The site was excavated in the late 1930s and early 1940s
and yielded the human skeletal remains of approximately
390 individuals ranging in age from 7 months in utero to
70+ years (Webb. 1950; Mensforth, 1990). 2 Radiocarbon
dates ranging from B.C. 5350 to 2515 indicate that the
site was formed over a 2500 year period (Winters, 1974).
However, recent studies have shown that the site was
occupied most intensively over a 1500 year period from
approximately 3500 to 4000 y.b.p. (Marquardt and
Watson, 1983). Floral and faunal analyses indicate that
Bt-5 was a late summer and fall occupation characteristic
of a seasonally mobile semisedentary group. The Bt-5
hunter-gatherers likewise placed heavy reliance on local
dietary resources. These included mussels, deer, turkey,
waterfowl, and an abundance of hickory nuts and acorns
(Marquardt, 1972; Winters, 1974; Marquardt and Watson,
1983). Although seven small fragments of squash were
recovered at Bt-5, it is very unlikely that this
domesticated cultigen contributed to the subsistence
economy of these peoples. Alternatively, it has been
suggested that people of the Late Archaic Green River
Culture may have made occasional use of squash gourds
as containers, thus implying a utilitarian function
(Marquardt and Watson, 1983). Given estimates of
regional site density, length of occupation, and
ethnographic analogy, it has been suggested that
individual groups ranged in size from 30 to 50 individuals
per given time (May, 1969).
Paleodemographic reconstructions and composite life
table analyses with archetype fertility data are currently
available for Libben and Bt-5 (Lovejoy et ah, 1977;
Mensforth, 1990). Age-specific mortality rates for the
two skeletal populations and mortality ratios which
compare the two groups are given in Table 2. It can be
seen that infants and children are well represented in each
group. Thus, census error due to infant underenumeration
is not problematic.
The similarities which characterize the Libben and Bt-5
demographic profiles are as follows. Both skeletal groups
exhibit (1) type II survivorship curves, (2) high infant
mortality, (3) low mortality in the adolescent years, (3)
early onset of elevated mortality rates in young adults, and
(4) a sex differential in adult survivorship where males
over 35 years of age experience higher death rates relative
to females. However, the latter trends that were observed in
each group are statistically insignificant (Lovejoy et al.,
1977; Mensforth, 1990).
The manner in which Libben and Bt-5 demographic
parameters differ are as follows. Ratios that compare
Libben and Bt-5 age-specific mortality rates are given in
Table 2. Except for infants in the first year of life and 3rd
decade adults, it can be seen that Libben mortality rates
exceed those of Bt-5 at all ages. A comparison of
1991
Paleoepidemiology of Porotic Hyperostosis
17
survivorship at age 15 (1 f 5: Libben, 52.9; Bt-5, 61.6) shows
that 8.7 percent fewer Libben individuals survived to
adulthood relative to Bt-5. This difference in subadult
survivorship is statistically significant .05; total lesions %2=16.77, pc. 001).
With regard to degree of involvement, it was found that
the majority of Libben and Bt-5 porotic hyperostotic
lesions were the result of a minimal osseous tissue
response to the stimulus of erythroid marrow hyperplasia.
Among subadults, lesions of slight expression occurred in
85.1 percent (114/134) of Libben and 87.0 percent (20/23)
of Bt-5 individuals. Lesions of moderate expression were
observed in only 14.2 percent (19/134) of Libben and 13.0
percent (3/23) of Bt-5 subadults. The majority of moderate
tissue responses occurred in specimens that ranged in age
from six months to three years of age at time of death in
each skeletal series. Only one Libben child displayed
severe porotic hyperostosis (see Figure la). No Bt-5
subadults were affected to such a degree.
No adults in either skeletal series displayed severe
bony changes in response to erythroid marrow
hyperplasia. Lesions of slight and moderate expression
occurred in 95.7 percent (66/69) and 4.3 percent (3/69)
of Libben adults with porotic hyperostosis, respectively.
All lesions that were encountered in Bt-5 adults were
slight involvements. In addition, all lesions that were
identified in Libben and Bt-5 adults were confined to the
superior orbital plates.
1991
Paleoepidemiology of P orotic Hyperostosis
21
TABLE 4. Comparison of the Total Frequency With Which Porotic Hyperostosis
Was Observed in the Libben and Bt-5 Skeletal Groups
Unremodeled Remodeled Total
Lesions Lesions Lesions
Skeletal
Group
ni
n2
(%)
n ?
(%)
n4
(%)
Bt-5
247
25
10.1
26
10.5
51
20.6
Libben*
580
1 16
20.0
87
15.0
203
35.0
Total
827
141
17.0
113
13.7
254
30.7
Symbols: nr total number of specimens that could be examined for the presence/absence of porotic hyperostosis; n2, number of
specimens that were scored positive for unremodeled lesions; n number of specimens that were scored positive for remodeled
lesions; n4, total number of specimens that were scored positive for remodeled or unremodeled lesions.
* Chi square values indicate that the Libben group experienced a statistically significantly greater frequency greater frequency
of unremodeled and total porotic hyperostotic lesions compared to the Bt-5 group (unremodeled lesions, %2=1 1 .95;
remodeled lesions, ^2=2.94;total lesions, yf= 1 6.77).
Age and Sex Specific Frequency Distributions
for Unremodeled Porotic Hyperostosis
Data enumerating the age and sex specific frequency of
occurrence for porotic hyperostosis in the Libben and Bt-5
skeletal groups are given in Tables 5, 6, 7, and 8. Chi square
values that compare the total frequency with which
remodeled and unremodeled lesions were observed in
subadults and adults partitioned by sex in the two skeletal
series are listed in Tables 9 and 10, respectively.
The general similarities that were observed for each
groups are as follows: (1) subadults displayed the highest
frequency of unremodeled and total lesions at time of death,
(2) adult females showed the second highest frequency of
unremodeled and total lesions, and (3) adult males exhibited
the lowest frequency of unremodeled and total lesions. The
only exception to the pattern described above was that Bt-5
adult females displayed a slightly greater frequency of total
lesions than subadults in this group.
The Libben rank order frequency for total lesions
observed is subadults (43.2%), adult females (35.3%), and
adult males ( 16.1%). The comparable information for Bt-5
is adult females (25.0%), subadults (23.0%), and adult
males (12.7%). The Libben rank order frequency for
lesions that were unremodeled at time of death is subadults
(30.3%), adult females (14.3%), and adult males (2.2%).
The comparable Bt-5 rank order is subadults (14%), adult
females (8.8%), and adult males (2.8%).
Chi square values for within group comparisons (see
TABLE 5. Age-Specific Frequency of Occurrence for Porotic Hyperostosis in Libben Subadults (0-15 years)
Age
Group
ni
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
Extra-orbital
Lesions
n2
(%)
n3
(%)
n4
(%)
n5
(%)
0.0 - 0.5
61
0
0.0
0
0.0
0
0.0
0
0.0
0.5 - 1.0
38
10
26.3
0
0.0
10
26.3
7
18.4
1.0 - 3.0
71
45
63.4
6
8.4
51
71.8
18
25.4
3.0 - 5.0
26
8
30.8
7
26.9
15
57.7
2
7.7
5.0 -10.0
54
14
25.9
17
31.5
31
57.4
4
7.4
10.0 -15.0
60
17
28.3
10
16.7
27
45.0
2
3.3
Total
310
94
30.3
40
12.9
134
43.2
33
10.7
Symbols: n,, total number of specimens that could be examined for the presence/absence of porotic hyperostosis per age group;
n2, number of specimens that were scored positive for unremodeled lesions per age group; ip, number of specimens that were
scored positive for remodeled lesions per age group; n4, number of specimens that exhibited remodeled and/or unremodeled
lesions per age group; ny number of specimens that manifested extra-orbital lesions per age group.
22
Mensforth
No. 46
TABLE 6. Age-Specific Frequency of Occurrence for Porotic Hyperostosis in Libben Adults
Female Male
Age
Group
"l
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
»/
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
n2
(%)
n3
(%)
>U
(%>
n2
(%)
n3
(%)
n4
m
15-19
18
6
33.3
5
27.8
1 1
61.1
16
i
6.3
4
25.0
5
31.3
20-24
17
5
29.4
3
17.7
8
47.1
17
0
0.0
3
17.7
3
17.7
25-29
17
1
5.9
4
23.5
5
29.4
17
i
5.9
4
23.5
5
29.4
30-34
15
1
6.7
3
20.0
4
26.7
31
0
0.0
4
12.9
4
12.9
35 -39
17
1
5.9
4
23.5
5
29.4
30
1
3.3
2
6.7
3
10.0
40-44
26
2
7.7
8
30.8
10
38.5
11
0
0.0
i
1.9
1
1.9
45 -49
13
2
15.4
1
7.7
3
23.1
9
0
0.0
i
11.1
1
11.1
+50
10
i
10.0
0
0.0
1
10.0
6
0
0.0
0
0.0
0
0.0
Total
133
19
14.3
28
21.0
47
35.3
137
3
2.2
19
13.9
22
16.1
Symbols: n /, total number of specimens that could be examined for the presence/absence of porotic hyperostosis per age group; n2, number of specimens that were scored positive
for unremodeled lesions per age group; nj, number of specimens that were scored positive for remodeled lesions per age group; n4, total number of specimens that were scored
positive for remodeled or unremodeled lesions per age group.
TABLE 7. Age-Specific Frequency of Occurrence for Porotic Hyperostosis in Bt-5 Subadults (0-15 years)
Age
Group
"l
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
Extra-orbital
Lesions
n2
(%)
n3
(%)
n4
(%)
n5
(%)
0.0 - 0.5
38
0
0.0
0
0.0
0
0.0
0
0.0
0.5 - 1.0
17
3
17.7
0
0.0
3
17.7
2
12.5
1.0 - 3.0
14
7
50.0
1
7.1
8
57.1
0
0.0
3.0 - 5.0
10
1
10.0
3
30.0
4
40.0
0
0.0
5.0 -10.0
9
1
11.1
3
33.3
4
44.4
1
11.1
10.0 -15.0
12
2
16.7
2
16.7
4
33.4
0
0.0
Total
100
14
14.0
9
9.0
23
3.0
3
3.0
Symbols: n]t total number of specimens that could be examined for the presence/absence of porotic hyperostosis per age group;
n2, number of specimens that were scored positive for unremodeled lesions per age group; n ^ number of specimens that were
scored positive for remodeled lesions per age group; n4, number of specimens that exhibited remodeled and/or unremodeled
lesions per age group; number of specimens that manifested extra-orbital lesions per age group.
TABLE 8. Age-Specific Frequency of Occurrence for Porotic Hyperostosis in Bt-5 Adults
Female Male
Age
Group
"i
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
«/
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
n2
(%)
n
r (%)
n4
(%)
n2
(%)
(%)
n4
(%)
15-19
9
i
11.1
3
33.3
4
44.4
10
0
0.0
0
0.0
0
0.0
20-24
8
2
25.0
1
12.5
3
37.5
13
i
7.7
1
7.7
2
15.4
25-29
10
2
20.0
1
10.0
3
30.0
12
0
0.0
1
8.3
1
8.3
30-34
9
0
0.0
1
11.1
1
11.1
11
0
0.0
3
27.3
3
27.3
35-39
7
1
14.3
1
14.3
2
28.6
7
0
0.0
1
14.3
1
14.3
40-44
7
0
0.0
0
0.0
0
0.0
8
0
0.0
0
0.0
0
0.0
45-49
8
1
12.5
0
0.0
1
12.5
5
1
20.0
0
0.0
1
20.0
+50
18
2
11.1
3
16.7
5
27.8
5
0
0.0
1
20.0
1
20.0
Total
76
9
11.8
10
13.2
19
25.0
71
2
2.8
7
9.9
9
12.7
Symbols: /? /, total number of specimens that could be examined for the presence/absence of porotic hyperostosis per age group; n2, number of specimens that were scored positive
for unremodeled lesions per age group; nj, number of specimens that were scored positive for remodeled lesions per age group; n4, total number of specimens that were scored
positive for remodeled or unremodeled lesions per age group.
1991
Paleoepidemiology of P orotic Hyperostosis
23
TABLE 9. Chi Square Values for Within-Group Comparisons of the Frequency
of Occurrence for Porotic Hyperostosis in the Libhen and Bt-5 Skeletal Samples
Libben
Bt-5
Age and Sex
Group Comparisons
Unremodeled
Lesions
Total
Lesions
Unremodeled
Lesions
Toted
Lesions
Subadults vs. Adults
44.35°
19.81°
2.78
0.57
Subadults vs. Adult Males
44.26°
30.87°
6.12*
4.56*
Subadults vs. Adult Females
12.60°
2.40
0.18
0.10
Adult Males vs. Adult Females
13.19°
13.19°
4.32*
3.62
a: significant at the .001 level of probability.
b: significant at the .05 level of probability.
Table 9) show that subadults in each group experienced a
significantly greater frequency of unremodeled and total
lesions compared to adult males, and adult females
displayed a significantly greater frequency of unremodeled
lesions compared to adult males. In addition, Libben
subadults had a significantly greater incidence of
unremodeled lesions compared to adult females. Therefore,
both skeletal groups are characterized by marked age and
sex differences in the frequency of occurrence of porotic
hyperostosis. The only exception to this was that Bt-5
subadults and adult females showed no significant
differences in the frequency of unremodeled or total lesions
that were observed. Chi square values for between group
comparisons (see Table 10) show that the only significant
difference between Libben and Bt-5 was confined to
subadults. Here, Libben subadults displayed a significantly
greater frequency of unremodeled and total lesions
compared to Bt-5 subadults.
The age and sex specific frequencies with which
unremodeled porotic hyperostosis occurred in the Libben
and Bt-5 groups are illustrated in Figures 5a and 5b. These
frequency distributions identify those age/sex groups where
erythroid marrow hyperplasia was sufficiently active to
invoke a skeletal response at, or near, time of death. It was
previously shown that Libben and Bt-5 adults are
characterized by a low incidence of unremodeled porotic
hyperostosis. Therefore, unremodeled lesion frequency
data for adults over 20 years of age are plotted by decade
here in order to smooth effects of small subsample sizes
combined with low overall incidence.
Comparisons of Libben and Bt-5 age-related porotic
hyperostosis frequency distributions for subadults are
illustrated more clearly in Figure 6. These patterns show
that the lesion bears a strong relationship with
developmental age. The following general relationships
characterize subadults in both skeletal series. No lesions
were seen in individuals that were under six months of age
at time of death. A low incidence of unremodeled porotic
hyperostosis first occurred in the six to twelve month
period. During the one to three year period the frequency
of unremodeled lesions increased dramatically to reach
peak incidence in each group. Subsequently, the frequency
of active lesions decreased throughout the childhood years.
During the adolescent period the frequency of unremodeled
porotic hyperostosis showed a slight increase once again.
Figures 6a-c show that the age at onset, peak incidence,
and age at remission for unremodeled porotic hyperostosis
are virtually identical in Libben and Bt-5 subadults. Chi
square values listed in Table 1 1 indicate that Libben
subadult age-related patterns for unremodeled porotic
hyperostosis are statistically significant. The absence of
lesions prior to six months of age, peak incidence in the
one to three year period, and subsequent decline are all
well defined. Chi square values for the Bt-5 subadults show
a similar, though less pronounced, trend. This is primarily
TABLE 10. Chi Square Values for Between-Group
Comparisons of the Frequency of Occurrence for Porotic
Hyperostosis in the Libben and Bt-5 Skeletal Samples
Chi Square Values
Age
Group
Unremodeled
Lesions
Total
Lesions
Subadults
10.38*
13.09"
Adult Females
0.25
2.39
Adult Males
0.08
0.51
Total Adults
0.06
2.26
a: significant at the
b: significant at the
.001 level of probability.
.01 level of probability.
24
Mensforth
No. 46
Libben
A
Bt-5
B
F IGURE 5. Age and sex specific frequencies of occurrence for unremodeled porotic hyperostosis that were observed in the Libben (5a)
and Bt-5 (5b) skeletal samples. Individuals over 15 years of age are partitioned by sex.
a
AGE IN YEARS
b
AGE IN YEARS
c
F IGURE 6. Histograms and polygon frequencies that compare the age-specific frequencies of occurrence for unremodeled (6a-c) and
remodeled (6a-b) porotic hyperostosis in Libben and Bt-5 subadults ranging in age from birth to 15 years at time of death.
1991
Paleoepidemiology of P orotic Hyperostosis
25
TABLE 1 1. Chi Square Values for Within-Group Comparisons of the Age-Specific Frequency
of Occurrence for Unremodeled P orotic Hyperostosis in Libben and Bt-5 Subadults
Age Group
Age
Group
Skeletal
Group
1-3
3-5
5-10
10-15
0.5- 1.0
Libben
13.60"
0.35
0.26
0.05
Bt-5
3.68
0.29
0.19
0.00
b
(jj
o
Libben
6.51h
23.57"
16.02"
Bt-5
4.20b
3.65
3.17
b
L/i
o
Libben
1.22
0.21
Bt-5
0.01
0.21
5.0-10.0
Libben
0.69
Bt-5
0.13
a: significant at the .001 level of probability.
b: significant at the .05 level of probability.
due to the fact that a smaller number of Bt-5 subadults
were available for comparison in each age class.
Further reference to Figures 6a-c shows that Libben
subadults consistently displayed a greater frequency of
unremodeled and total lesions per age class compared to
Bt-5. Chi square values given in Table 12 indicate that
none of the subadult age class comparisons between the
two groups are statistically significant. However, this is not
surprising given that unremodeled lesions in both groups
displayed a marked conformity in age specific distribution,
and a much smaller number of Bt-5 subadults were
available for comparisons.
Among adult females in each group it can be seen that
the highest frequency of unremodeled porotic hyperostosis
occurred in individuals that were 15 to 30 years of age at
time of death. Thereafter, adult females displayed a low
frequency of unremodeled lesions. In contrast, adult males
in each group showed a low frequency of unremodeled
lesions at all ages. Therefore, the patterns that were
observed for porotic hyperostosis indicate that ( 1 ) the ages
at onset, peak incidence, and remission for unremodeled
lesions are markedly similar in the two band level societies,
and (2) the frequency distributions that characterize
unremodeled lesions identify those age/sex groups at
highest risk of acquiring iron deficiency anemia.
Age/Sex Frequency Distributions
for Remodeled Porotic Hyperostosis
Among those subadults that displayed porotic
hyperostosis in the two skeletal series (see Tables 5 and 7),
remodeled lesions were observed in 29.9 percent (40/134)
of Libben and 39. 1 percent (9/23) of Bt-5 individuals. Thus,
the majority of subadults with porotic hyperostosis in the
two skeletal series had active lesions at time of death. In
contrast, the vast majority of adults with porotic
hyperostosis had remodeled lesions at time of death. In
Libben adults with porotic hyperostosis, remodeled lesions
were seen in 86.4 percent ( 19/22) of affected males and 59.6
percent (28/47) of affected females (see Table 6). Similarly,
in Bt-5 adults with porotic hyperostosis, remodeled lesions
occurred in 77.8 percent (7/9) of affected males and 52.6
percent (10/19) of affected females (see Table 8).
TABLE 12. Chi Square
Values That Compare
the Age-
Specific and Total Frequency of Occurrence for Porotic
Hyperostosis Among Libben and Bt-5 Subadults
Chi Square Values
Age
Unremodeled
Remodeled
Total
Group
Lesions
Lesions
Lesions
0.5 - 1.0
0.49
—
0.49
1.0 -3.0
0.88
0.03
1.19
3.0 - 5.0
1.66
0.03
0.91
5.0 -10.0
0.93
0.01
0.53
10.0 -15.0
0.70
0.00
0.56
Total
\038b
1.09
13.09"
a: significant at the .00 1
level of probability.
b: significant at the .01 level of probability.
26
Mensforth
No. 46
TABLE 13. Porotic Hyperostosis Frequency Data and Chi Square Values That Compare Adult Females 12-30 Years of Age
Versus Those +30 Years of Age
Lesion Frequency
Unremodeled Total Chi Square Values
Adult
Female
Samples
Age
Group
n i
ll 2
Lesions
<%)
Lesions
(%)
Unremodeled
Lesions
Remodeled
Lesions
Libben
15-30
52
12
23.1
24
46.2
5.39b
4.37*
+30
81
7
8.6
23
28.4
Bt-5
15-30
27
5
18.5
10
37.0
1.79
3.24
+30
49
4
8.2
9
18.4
Combined
15-30
79
17
21.5
34
43.0
7 22 a
7.22"
+30
130
1 1
8.5
32
24.6
Symbols: /, total number of adult females that could be scored for the presence/absence of porotic hyperostosis per age group; ni number of adult females that were scored
positive for unremodeled lesions per age group; nj total number of individuals that were scored positive for remodeled or unremodeled lesions per age group; a, Chi square
values significant at the .01 level of probability; b, Chi square values significant at the .05 level of probability.
The similarities which characterize the age-specific
frequency of remodeled lesions that occurred in the two
groups are as follows. A low incidence of remodeled
lesions first appeared in the one to three year period. The
frequency of remodeled lesions gradually increased to
reach peak incidence in the five to ten year period in each
group. A low frequency of remodeled lesions was once
again seen in adolescents. Remodeled lesions in adults
were most commonly seen in individuals that ranged in
age from 1 5 to 35 years in each group. The only exception
to this was that Libben females over 35 years of age
showed a remodeled lesion frequency which was 14.8
percent higher compared to young adult females.
However, the latter difference is statistically insignificant
%2= 1 .04; p>.05).
Age Related Porotic Hyperostosis Frequency Distribution
in Adult Females
Archetype fertility data reported by Weiss (1973) show
that age specific fertility rates in contemporary primitive
societies are highest in adult females that are 15 to 30
years of age. As a corollary, it would be predicted that
peak periods of reproductive activity would predispose
young women to a greater incidence of iron deficiency
anemia. The latter would give rise to variable levels of
erythroid marrow hyperplasia and result in a greater
frequency of unremodeled porotic hyperostosis in these
individuals. Therefore, Libben and Bt-5 adult female age-
related porotic hyperostosis frequency distributions were
assessed in the following manner.
Information summarized in Table 13 compares the
frequencies with which unremodeled porotic hyperostosis,
and total lesions, occurred in Libben and Bt-5 adult
females that were 15 to 30 years of age versus those that
were +30 years of age at time of death. For the 15 to 30
year age group results show that 23.1 percent of Libben
females and 18.5 percent of Bt-5 females had unremodeled
lesions at time of death. Libben and Bt-5 adult females that
were over 30 years of age displayed unremodeled lesion
frequencies of 8.6 percent and 8.2 percent, respectively.
Thus, young adult females in both groups exhibited a
greater frequency of unremodeled lesions compared to
their older peers (Libben, 14.5% higher; Bt-5, 10.3%
higher). However, chi square values listed in Table 11
indicate that this difference is only significant for Libben
adult females.
Comparisons of total lesion frequencies that were
observed for Libben and Bt-5 adult female age
categories yielded similar results. However, these data
are inappropriate for the following reason. Remodeled
lesions that occurred in association with childhood and
adolescent stress episodes are likely to accumulate in
young adults where a substantial degree of age-
progressive bone remodeling has not yet occurred. In
contrast, it would be expected that many lesions of
youth would be completely remodeled away in older
individuals. Therefore (1) age-progressive bone
remodeling alone can give rise to age related differences
in the frequency of occurrence for porotic hyperostosis
in adults, and (2) combined measures of remodeled and
unremodeled porotic hyperostosis that disregard lesion
activity status at time of death will result in inflated
young adult lesion frequencies and deflated older adult
lesion frequencies.
1991
Paleoepidemiology of Porotic Hyperostosis
27
Age/Sex Frequency Distribution for the Libben and Bt-5
Porotic Hyperostosis Macrosample: a Test of the Iron
Deficiency Anemia Hypothesis
Paleoepidemiological analyses of skeletal pathologies
have generally emphasized the comparative approach. Here,
the primary goals are to identify and interpret the ways in
which two or more skeletal series differ. Nonetheless, the
only way to evaluate the extent to which one or more
factors may have played a common role in the etiology of a
stress indicator is to examine the similarities in lesion
frequency patterns that characterize two or more skeletal
groups. This can be accomplished, in part, by the use of
more refined age categories (eg., 5 year age intervals in
adults). However, when this procedure is employed with
skeletal samples that are derived from small anthropological
populations an additional problem is introduced. Thus, the
combined effects of small subsample sizes and low lesion
frequencies that characterize various age/sex categories will
give rise to unstable, or spurious, lesion frequencies that are
of limited inferential utility.
The problems described above can be overcome by the
use of macrosamples. Here, the age and sex specific
skeletal lesion frequencies that are observed for two or
more groups are combined. Although between group
differences in skeletal lesion frequency will be masked, the
macrosample summary data will more clearly identify the
age and sex specific lesion patterns that are common to all
human groups. The latter permit the skeletal biologist to
infer probable cause of a disorder with greater confidence.
Data which summarize the age and sex specific frequency
of occurrence for porotic hyperostosis in the combined
Libben and Bt-5 macrosample are given in Tables 14 and 15.
TABLE 14. Age-Specific Frequency of Occurrence for Porotic Hyperostosis
in the Combined Sample of Libben and Bt-5 Subadults (0-15 years)
Age
Group
ni
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
Extra-orbital
Lesions
n2
(%)
n3
(%)
n4
(%)
n5
(%)
0.0 - 0.5
99
0
0.0
0
0.0
0
0.0
0
0.0
0.5 - 1.0
55
13
23.6
0
0.0
13
23.6
9
16.4
1.0 - 3.0
85
52
61.2
7
8.2
59
69.4
18
21.2
3.0 - 5.0
36
9
25.0
10
27.8
19
52.8
2
5.6
5.0 -10.0
63
15
23.8
20
31.8
35
55.6
5
7.9
10.0 -15.0
72
19
26.4
12
16.7
31
43.1
1
2.8
Total
410
108
26.3
49
12.0
157
38.3
36
8.8
Symbols: total number of specimens that could be examined for the presence/absence of porotic hyperostosis per age group;
/?„ number of specimens that were scored positive for unremodeled lesions per age group; n3, number of specimens that were
scored positive for remodeled lesions per age group; n4, total number of specimens that were scored positive for remodeled
and/or unremodeled lesions per age group; number of specimens that manifested extra-orbital lesions per age group.
TABLE 15. Age-Specific Frequency of Occurrence Porotic Hyperostosis Among the Combined Libben and Bt-5 Adult Samples
Female Male
Age
Group
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
"i
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
n2
(To)
n3
(To)
>U
(%)
n2
(To)
n3
(To)
»4
(To)
15-19
27
i
25.9
8
29.6
15
55.6
26
i
3.9
4
15.4
5
19.2
20-24
25
i
28.0
4
16.0
11
44.0
30
i
3.3
4
13.3
5
16.6
25-29
27
3
11.1
5
18.5
8
29.6
29
i
3.5
5
17.2
6
20.7
30-34
24
1
4.2
4
16.7
5
20.8
42
0
0.0
7
16.7
7
16.7
35 -39
24
2
8.3
5
20.8
7
29.2
37
1
2.7
3
8.1
4
10.8
40-44
33
2
6.1
8
24.2
10
30.3
19
0
0.0
1
5.3
1
5.3
45-49
21
3
14.3
1
4.8
4
19.1
14
1
7.1
1
7.1
2
7.3
+50
28
3
10.7
3
10.7
6
21.4
1 1
0
0.0
1
9.1
i
9.1
Total
209
28
13.4
38
18.2
66
31.6
209
5
2.4
26
12.1
31
14.9
Symbols: nt, total number of specimens that could be examined for the presence/absence of porotic hyperostosis per age group; n2, number of specimens that that were scored
positive for unremodeled lesions per age group; n3, number of specimens that were scored positive for remodeled lesions per age group; n4, total number of specimens that were
scored positive for remodeled or unremodeled lesions per age group
28
Mensforth
No. 46
B
FIGURE 7. Libben and Bt-5 combined macrosample age specific frequencies of occurrence for (7a) unremodeled and (7b) remodeled
porotic hyperostosis.
The age and sex specific frequency distributions for
unremodeled and remodeled lesions in the macrosample are
illustrated in Figures 7a and 7b. respectively. The ages at
onset, peak incidence, and remission for unremodeled
lesions that characterize the macrosample are similar to
those that were described earlier for each group. Thus, peak
frequencies for unremodeled lesions occurred in young
children one to three years of age, and adult females 15 to 25
years of age at time of death. In addition, adolescents and
adults over 45 years of age showed slight, but insignificant,
increases in the frequency of unremodeled lesions. However,
adult males showed a low incidence of active lesions at all
ages. Thus, age and sex related macrosample patterns for
unremodeled lesions identify those age/sex groups of a
population at highest risk of acquiring iron deficiency.
Moreover, these patterns exhibit a marked concordance with
the hypothetical model of unremodeled porotic hyperostosis
lesion frequencies that would be expected as a result of iron
deficiency alone (see Figure 3a).
The ages at onset, peak incidence, and remission for
remodeled lesion frequencies in the Libben and Bt-5
macrosample also conform to the pattern that would be
expected to occur in association with iron deficiency
anemia. Figure 7b clearly shows that remodeled lesion
frequencies display a lag phase which translates to the right
of each high risk age/sex group. These patterns are
concordant with those established in the hypothetical
model illustrated in Figure 3c. Macrosample data given in
Table 13 also show that young adult females 15 to 30 years
of age have a frequency of unremodeled porotic
hyperostosis which is 13 percent higher than adult females
over 30 years of age at death. A chi square comparison
indicates that this difference is statistically significant
(%2=7.22; p<. 01). Thus, the age-related frequency
distribution for unremodeled porotic hyperostosis in adult
females identifies those individuals at greatest risk of
acquiring iron deficiency anemia during the peak years of
reproductive activity.
Results that have been presented thus far can be
summarized as follows:
1) The vast majority of porotic hyperostotic lesions
that were observed in the Libben and Bt-5 groups
consisted of slight bony involvements that were
primarily restricted to the superior orbital plates.
Extra-orbital lesions, and lesions of moderate
expression, were most commonly seen in
subadults that were six months to three years of
age at time of death in each group.
2) Patterns in the ages at onset, peak incidence, and
remission for unremodeled porotic hyperostosis
were similar in the two groups such that;
a) the highest frequency of unremodeled lesions
occurred in subadults,
b) adult females showed the second highest
frequency of unremodeled lesions at time of
death, and
1991
Paleoepidemiology of Porotic Hyperostosis
29
c) adult males showed the lowest incidence of
unremodeled lesions.
3) Within group comparisons indicated that;
a) subadults in each group experienced a
significantly greater frequency of unremodeled
lesions compared to adult males, and Libben
subadults also displayed a significantly greater
frequency of unremodeled lesions compared to
adult females,
b) adult females in each group exhibited
unremodeled lesion frequencies that were sig-
nificantly greater compared to adult males, and
c) young adult females that were 15 to 30 years
of age showed a greater frequency of
unremodeled lesions compared to females
over 30 years of age in each group. However,
this difference was only significant for the
Libben adult female comparison.
4) Unremodeled and remodeled porotic hyperostosis
frequency data for the Libben and Bt-5
macrosample indicate a marked concordance with
the hypothetical iron deficiency anemia model.
Thus, age and sex specific patterns in the frequency
of occurrence for porotic hyperostosis clearly
identify those age/sex groups at greatest risk of
acquiring iron deficiency anemia in human groups.
5) Between group comparisons showed that:
a) unremodeled lesion frequencies that were
observed in Libben and Bt-5 adult males, and
adult females, were similar in magnitude, and
b) the only significant difference in the frequency
of porotic hyperostosis that characterized the
two groups was confined to subadults. Here,
Libben children displayed a significantly
greater frequency of unremodeled and total
lesions compared to Bt-5 subadults.
Discussion
Level of Non-Specificity and Differential Diagnosis for
Porotic Hyperostosis
With regard to level of non-specificity, it must be
emphasized that disease prevalence differentials enhance
our ability to infer the probable cause of porotic
hyperostosis with a high degree of confidence on a
populational basis. For example, due to their rare
occurrence in small anthropological populations, many of
the conditions which promote erythroid marrow hyperplasia
(see Table 1) could have played little or no singular, or
collective, role in the etiology of the lesion (eg., hereditary
spherocytosis, congenital nonspherocytic anemias, etc). Iron
deficiency anemia and the congenital hemolytic anemias are
the only disorders that are sufficiently prevalent to account
for the frequency with which porotic hyperostosis occurred
in earlier human groups. In addition, the congenital
hemolytic anemias can only be invoked to explain
frequency of the lesion in certain geographically restricted
populations in the Old World. Regardless, the view posited
here is that porotic hyperostosis, as a bioassay of population
fitness, is characterized by a very low level of non-
specificity relative to most other non-specific indicators of
disease and nutritional stress.
The various pathological skeletal changes that occur in
response to the anemias, and which aid in the differential
diagnosis thereof, have been thoroughly investigated by
Moseley (1965, 1966, 1974). His observations, combined
with the age and sex related patterns reported here, served as
the basis for evaluating the pathogenesis of Libben and Bt-5
porotic hyperostosis. The skeletal remains of all Libben and
Bt-5 individuals were therefore surveyed for pathological
changes known to occur in iron deficiency anemia and the
congenital hemolytic anemias (i.e., sickle cell anemia and
thalassemia) (Moseley, 1975). For reasons discussed above,
other conditions were excluded from consideration.
Skeletal changes in the skull that are produced by
erythroid marrow hyperplasia were described earlier. Those
that are observed most frequently affecting the long bones
include (1) widening of the marrow spaces, (2) cortical
thinning, and (3) coarsening of trabecular patterns (Angel,
1967; Moseley, 1974). However, these skeletal changes
occur in response to the hemolytic anemias, iron deficiency
anemia, and are also known to accompany a wide variety
of other disease and nutritional disorders (Jaffe, 1972;
Aegerter and Kirkpatrick, 1975; Greenfield, 1975). These
pathological changes are therefore highly non-specific and
of little diagnostic utility.
Here, it should be emphasized that skeletal changes other
than those produced by marrow hyperplasia are more
important for diagnosis of the hemolytic anemias. For
example, vertebral step deformity and hand-foot syndrome
are the most reliable skeletal criteria for the identification of
sickle cell anemia (Moseley, 1974). Also, children with sickle
cell anemia often exhibit bone infarctions, osteomyelitis, and
periostitis (Jaffe, 1972). However, it is well known that
infants and young children are at greater risk of acquiring
bone infections compared to older individuals (Robbins,
1974). Furthermore, periosteal reactions can be induced by a
wide variety of disease agents (Greenfield, 1975). Therefore,
the latter types of skeletal lesions cannot be regarded as
diagnostic for the hemolytic anemias.
In other circumstances, the extent of bone marrow
proliferation is a more important indicator as to the nature
of the underlying anemic stimulus. Studies concerned with
the biodynamics of marrow response to anemia have shown
that bone marrow proliferation is dependent on the amount
30
Mensforth
No. 46
of iron available to the tissue (Hillman and Henderson,
1969; Hillman, 1970). For example, the level of erythroid
marrow hyperplasia that occurs in iron deficiency anemia is
usually self-limiting and generally approximates only 2 to 3
times the normal rate (Finch, 1970). In severe chronic iron
deficiency anemia bone marrow proliferation may on
occasion reach levels that are 4 to 6 times normal (Giblett
et al. , 1950; Hillman and Henderson, 1969). In contrast,
individuals suffering from hemolytic anemias frequently
exhibit a marrow response that is 5 to 10 times above
normal which then results in excessive hypertrophic bony
changes in the skull (Finch, 1970; Moseley, 1974).
Moseley (1974) has remarked that the skeletal changes
due to marrow hyperplasia which occur in thalassemic
homozygotes are more pronounced than those found in any
other condition. Facial bones are often involved to the
extent that malocclusion and a rodent fades deformity
become manifest. The most reliable post-cranial skeletal
change that has been observed in thalassemic homozygotes
is bulbous expansion of the ribs. This primarily affects the
posterior portions of the ribs and represents an exaggerated
response to subperiosteal bone marrow proliferation
(Moseley, 1974). Thus, the excessive levels of marrow
proliferation which occur in the hemolytic anemias are
determined, in part, by the increased amount of bioavailable
iron that is retrieved from lysed red blood cells (Smith,
1972). None of the skeletal changes that are considered
pathognomonic for the hemolytic anemias were observed in
any Libben or Bt-5 individuals. The overwhelming majority
of porotic hyperostotic lesions that were seen in the two
groups involved only slight degrees of osseous tissue
hypertrophy. These findings suggest that most Libben and
Bt-5 individuals with porotic hyperostosis experienced a
limited level of marrow hyperplasia which is similar to that
reported in acute iron deficiency anemia. This is in contrast
to the pronounced skeletal changes that would be expected
to occur in greater frequency as a result of severe chronic
iron deficiency anemia or the hemolytic anemias.
It is concluded here that (1) the marked age and sex
specific frequency distributions for the lesion, (2) the
absence of skeletal changes that are considered
pathognomonic for the hemolytic anemias, and (3) the low
levels of osseous tissue response that were observed,
support the inference that the majority of porotic
hyperostosis in the Libben and Bt-5 skeletal groups was the
result of acute iron deficiency anemia.
Age and Sex Related Demographic Sensitivity of
Porotic Hyperostosis
Mortality data reported earlier show that no significant
differences characterize between group comparisons of
Libben and Bt-5 adult survivorship. Similarly, no significant
differences were observed for between group comparisons
of Libben and Bt-5 porotic hyperostosis frequency in adults.
These results held true for both sex-combined and sex-
specific intergroup comparisons of mortality and lesion
frequency data. Therefore, the demographic and
paleoepidemiological patterns that were observed for adults
in the two skeletal groups display a marked concordance.
Even though adult survivorship distributions were not
substantially different for comparisons within or between
each group, the average Libben:Bt-5 adult q.v ratio of 1.24
shows a slight trend favoring higher age-specific mortality
rates in Libben adults.5 Similarly, the Libben:Bt-5 adult sex-
combined porotic hyperostosis ratios for the frequency of
unremodeled and total lesions are 1.09 and 1.35,
respectively. Therefore, the minor between group
differences in the frequency of porotic hyperostosis
correspond in direction, though not necessarily in
magnitude, to the minor difference in mortality rates that
characterize Libben and Bt-5 adults. Within group
comparisons of adult male versus female porotic
hyperostosis showed that Libben and Bt-5 females both had
a significantly greater frequency of unremodeled lesions at
time of death. The Libben adult femaleimale porotic
hyperostosis frequency ratios for unremodeled lesions are
6.50 and 4.21, respectively. Comparable sex ratios for total
lesions at Libben and Bt-5 are 2.19 and 1.97, respectively.
The sex-related lesion frequency differences described
above would prompt some skeletal biologists to suggest that
elevated pregnancy and lactation stress would result in
higher levels of morbidity and mortality in adult females.
However, it is unwise to base such conclusions on skeletal
lesion frequency data alone. It was previously stated that no
significant differences characterize patterns of adult male
and female survivorship in each group. Indeed, Libben and
Bt-5 adult females displayed a trend where age-specific
survivorship either equaled, or exceeded, that of males in
each series. Given that factors involved in the risks of
acquiring iron deficiency anemia differ markedly for adult
males and females, the finding that Libben and Bt-5 females
displayed a significantly greater frequency of unremodeled
lesions is not unexpected. However, it is clear that these sex-
related differences in the frequency of porotic hyperostosis
are demographically insensitive. Therefore, while sex-
related differences in the frequency of the skeletal lesion
may be of value for purposes of differential diagnosis, they
do not appear to provide useful information about Libben
and Bt-5 adult sex-specific mortality experience.
Information reported earlier shows that the only
significant differences between Libben and Bt-5 porotic
hyperostosis frequencies and demographic parameters are
confined to subadults. That is, (1) Libben children
displayed a significantly greater frequency of unremodeled
and total porotic hyperostotic lesions at time of death, and
(2) significantly fewer Libben individuals survived to age
1991
Paleoepidemiology of P orotic Hyperostosis
31
15 compared to Bt-5 subadults. These results indicate the
frequency of porotic hyperostosis in subadults and sur-
vivorship at age 15 are inversely related in the two groups.
With regard to demographic sensitivity, the additional
findings are of interest. The Libben:Bt-5 subadult porotic
hyperostosis ratios for the frequency of unremodeled and
total lesions are 2.18 and 1.88, respectively. The average
Libben:Bt-5 mortality ratio for subadults 0 to 15 years of
age is 1 .60. Differences in the frequency of Libben and Bt-
5 subadult porotic hyperostosis therefore correspond in
direction and magnitude to differences in subadult mortality
that was observed for the two groups. Indeed, ratios that
compare Libben and Bt-5 subadult lesion frequencies
exceed the subadult mortality ratio and exhibit demographic
hypersensitivity for individuals in this age category.
For inferential purposes, the relationships described
above were formalized so that subadult porotic
hyperostosis lesion frequencies could be used to predict
survivorship at age 15 ( 1 5 ) in earlier human groups. The
algorithms are presented as follows:
la. y =
y =
x\ =
lb. y
v =
*2 =
.05337(.Y!) + 69.072
1 1 5
frequency of unremodeled porotic
hyperostosis that is observed in
subadults (0-15 years)
-0.4307(x2) + 71.506
•15
frequency of total porotic hyperostotic
lesions that are observed in subadults
(0-15 years).
Given that only two skeletal samples were used to
generate the equations listed above, it is clear that the
relationships posited must be regarded as hypotheses. The
extent to which the models accurately reflect the relationships
between porotic hyperostosis and survivorship at age 15 in
earlier human groups awaits more rigorous tests that employ
a substantially greater number, and temporospatial diversity,
of skeletal groups. Future analyses are required to determine
(a) the precision with which lesion frequencies can accurately
estimate demographic parameters of earlier human groups,
and (b) whether a linear or non-linear model provides the best
measure of a relationship as such. Alternatively, it may be
found that porotic hyperostosis lesion frequencies provide no
useful information for demographic inference whatsoever.
These are matters that must be addressed in future research.
Developmental Factors Which Influence Age-Related
Differences in the Frequency of Occurrence and
Morphological Expression of Porotic Hyperostosis
Results presented here, and elsewhere, document that
children under ten years of age generally display greater
frequencies and degrees of severity for porotic hyperostosis
compared to adults in earlier human groups (Moseley, 1965;
El-Najjar et al ., 1976; Lallo et al. , 1977). The intrinsic
physiological and extrinsic environmental factors which
promote risk of acquiring iron deficiency anemia in
subadults, particularly those that are six months to three
years of age, have been discussed. There are, in addition,
developmental factors that differentially influence the extent
to which subadults and adults are likely to manifest skeletal
changes in response to erythroid marrow hyperplasia.
For example, the liver and spleen play a dominant role in
red blood cell formation during much of the fetal period
(Smith, 1972). However, shortly prior to birth these tissues
quiesce. Thereafter, bone marrow cavities become the
principal sites of hematopoiesis throughout life (Sodeman
and Sodeman, 1974). In the infant virtually all bone marrow
cavities are actively involved in red blood cell formation.
This results in a circumstance where, until four years of age,
a delicate balance exists between the limited bone marrow
space that is available for hematopoiesis, and the bone
marrow space that is required to meet the rapidly growing
child’s increased red blood cell requirements (Smith, 1972;
Sodeman and Sodeman, 1974). By four years of age the
skeleton has achieved sufficient volumetric growth to
provide bone marrow space in excess of hematopoietic
needs. Then, during the second decade of life the bone
marrow cavities in the appendicular skeleton undergo a
slow transition where active red marrow is replaced by
inactive yellow marrow. By twenty years of age active red
marrow in the appendicular skeleton is confined to the
proximal regions of the long bones. In contrast, bones of the
axial skeleton (i.e., skull, vertebrae, sternum, and ribs)
remain active in red blood cell formation throughout life
(Robbins, 1974; Hardesty and Weatherall, 1982).
Therefore, skeletal changes that are due to erythroid
marrow hyperplasia are most likely to occur in anemic
children under four years of age. This is a fundamental
consequence of (a) the competitive relationship between
limited supply and increased demands for active marrow
space, combined with (b) the numerous constitutional and
environmental factors that promote the risk of developing
iron deficiency anemia at this age. Infants and young
children can thus be regarded as hypersensitive to skeletal
changes in the anemias.
For adults and subadults alike, it is reasonable to
presume that the first tissues to respond to an anemic
stimulus will be those that are already active in red blood
cell formation (i.e., bones of the axial skeleton) (Smith,
1972; Sodeman and Sodeman, 1974). In contrast to young
children, however, the adult appendicular skeleton provides
a substantial reserve of yellow marrow that is capable of
reverting to active red marrow in response to a hypoxic
stimulus (Sodeman and Sodeman, 1974). Therefore, the
32
Mensforth
No. 46
extent to which skeletal changes will occur in the adult
cranium as a result of erythroid marrow hyperplasia
depends on ( 1 ) the duration and severity of the underlying
anemic condition, and (2) the rate at which appendicular
bone marrow sites are recruited for compensatory red
blood cell formation.
Given that substantial hematopoietic reserves are readily
available to the adult, it is reasonable to suggest that
skeletal changes due to erythroid marrow hyperplasia will
be uncommon, or extremely limited, in these individuals.
In accord with this relationship are results reported by
Stuart-Macadam (1985) and results obtained herein (i.e.,
that Libben and Bt-5 adults displayed low frequencies of
unremodeled porotic hyperostosis overall, and that such
lesions were minimal bony responses to marrow
proliferation). Thus, adults are best regarded as
hyposensitive to skeletal changes in the anemias.
It is well recognized that porotic hyperostosis of the
superior orbital plates (i.e., cribra orbitalia) is by far the
most common expression of the disorder. In this study
orbital lesions were observed in 97.2 percent (247/254) of
all Libben and Bt-5 individuals that had either remodeled
or unremodeled porotic hyperostosis. Among the small
sample of individuals that had one or more extra-orbital
lesions in the two groups 80.6 percent (29/36) also
displayed orbital lesions at time of death. The specific
reason as to why the superior orbital plates are so
responsive to erythroid marrow hyperplasia has eluded
functional explanation. Nonetheless, this skeletal site
exhibits a marked sensitivity to bone marrow proliferation
and must be considered as a superior index of the skeletal
disorder. Furthermore, porotic hyperostosis affecting the
superior orbital plates cannot be detected by standard
clinical x-ray apparatus unless the skeletal changes are as
pronounced as those shown in Figure I . Therefore,
workers who (1) restrict their definition of porotic
hyperostosis to include only those skeletal lesions that are
seen affecting the frontal, parietal, or occipital bones
proper, and (2) attempt to identify porotic hyperostosis in
the living by radiographic techniques, will grossly
underestimate the frequency with which the pathological
bony response occurred.
The way in which erythroid marrow hyperplasia gives
rise to the morbid appearance of porotic hyperostosis as
seen in the skeleton is relatively easy to comprehend.
Laboratory studies have shown that tissue hypoxia initially
stimulates a small number of hematopoietic stem cells to
enter into an intense phase of proliferation (Till and
McCulloch, 1961; Trentin, 1971). This results in a
circumstance where clonal colonies of undifferentiated
stem cells quickly become established in the bone marrow
(Sodeman and Sodeman, 1974). Then, after a few days,
these differentiate into discrete colonies or nests of
hematopoietically active bone marrow cells.
Hyperplastic activity of discrete bone marrow cell
colonies is the best explanation for the manner in which
porotic hyperostotic lesions acquired their porous
appearance (see Figure lb). The degree to which osseous
tissue hypertrophy accompanied such lesions was probably
dependent on ( 1 ) the level of hyperplasia that was attained
by discrete bone marrow cell colonies, and (2) the extent to
which additional marrow space was required to accom-
modate cells that were actively involved in hematopoiesis.
Factors Involved in the Etiology of Iron Deficiency
Anemia at Libben and Bt-5
Malabsorption syndrome as a consequence of intestinal
cestodiasis is a factor which may have contributed to iron
deficiency anemia in the Libben and Bt-5 groups. Parasitic
tapeworm infestation is usually contracted by ingestion of
raw, or under-cooked, meat or fish, and occurs in
geographically widespread human populations (Robbins,
1974). Libben and Bt-5 faunal analyses described earlier
indicate that both groups relied heavily on seasonally
available freshwater marine resources. This dependence
was particularly marked for the Libben group where an
abundance of freshwater fish were exploited. Thus, it is
conceivable that fish tapeworm infestation (i.e.,
Diphyllobothrium latum), and subsequent intestinal
malabsorption of dietary nutrients, affected some
individuals in both skeletal populations.
However, epidemiological surveys in contemporary
aboriginal societies have shown that microbial respiratory
and gastrointestinal tract infections play a much more
substantial role in precipitating, and exacerbating,
nutritional crises, particularly in infants and young children
(Maynard and Hammes, 1970; Jose and Welch, 1970).
Although parasite loads reach higher levels in anemic and
malnourished children, they generally play no major role in
initiating nutritional deficiency syndromes during the early
years of life (Jose and Welch, 1970). Therefore, it seems
unlikely that intestinal cestodiasis was important in the
etiology of iron deficiency anemia and nutritional stress in
Libben and Bt-5 individuals.
Constitutional factors such as prematurity and low
birthweight also probably played no substantial role in the
pathogenesis of Libben and Bt-5 porotic hyperostosis in
subadults for the following reasons. First, the iron
deficiency anemia of prematurity would be expected to
produce skeletal changes that could be seen in individuals
that were three to six months of age at time of death.
However, no lesions were observed in Libben or Bt-5
infants prior to six months of age. Second, premature
infants in contemporary under-developed societies rarely
survive the neonatal period due to the numerous
physiological handicaps that these individuals experience
1991
Paleoepidemiology of P orotic Hyperostosis
33
at birth (Levine and Gordon, 1942; Maynard and Hammes,
1970; Korones, 1976). Given that premature infants only
constitute 6-12 percent of live births in human groups
(Maynard and Hammes, 1970 and references therein),
combined with their high risk of mortality, it is highly
improbable that a sufficient number of premature infants
would have survived to account for the frequencies with
which porotic hyperostosis occurred in Libben and Bt-5
infants.
Similarly, diet does not appear to be a major factor in
the etiology of Libben and Bt-5 porotic hyperostosis. Floral
and faunal analyses indicate that resources available to the
two groups provided an adequate supply of bioavailable
iron. Furthermore, these resources were low in chelating
agents that are known to inhibit the absorption of dietary
iron. It remains likely, however, that dietary factors were to
some extent involved in promoting the onset of iron
deficiency anemia that occurred during the weaning period
in each group. Many human societies routinely incorporate
high carbohydrate gruels in the weaning diet for reasons
principally related to ease of preparation, mastication, and
assimilation (Gordon et ai, 1963; Scrimshaw and Young,
1976; Farb and Armelagos, 1980). These are usually
manufactured from wild or domesticated cereal grains or
other indigenous vegetable resources. In addition, the
introduction of solid food items in the diet exposes the
weanling to novel pathogens. The latter play a substantial
role in precipitating nutritional crises and growth
retardation due to recurrent episodes of acute and chronic
gastrointestinal tract infections (Gordon et ai, 1963).
Still, there is little evidence to suggest that chronic iron
deficiency anemia due to dietary inadequacy was an
important epidemiological factor in the development of
Libben and Bt-5 porotic hyperostosis. The dramatic
decrease in the frequency of unremodeled porotic
hyperostosis that was seen in subadults over three years of
age, and low levels of osseous tissue response that were
observed in each group, are concordant with this view.
These findings are in contrast to those reported for several
maize dependent New World skeletal groups where the
incidence of porotic hyperostosis remained high throughout
the childhood years, and lesions displayed more severe
degrees of bony involvement (Moseley, 1965; El-Najjar et
al., 1976; Lallo et al., 1977).
In order to more clearly evaluate the extent to which
diet may have played a role in the etiology of Libben and
Bt-5 subadults, lesion frequencies for these individuals
were compared to those that occurred in subadults from
two maize dependent prehistoric groups. The latter
represent subsamples of the skeletal materials that were
employed by El-Najjar (1976) where the relationship
between diet and the frequency of occurrence for porotic
hyperostosis was investigated in several New World
prehistoric groups.
The Anasazi subadults (n= 54) were recovered from
Canyon de Chelly in northeast Arizona and are affiliated
with the Basketmaker II-III cultural period that extended
from 400 to 700 A.D. (El-Najjar et al., 1976). 6 The
Peruvian subadults (n = 70) represent a poorly documented
skeletal series of prehistoric South American Indians that
were recovered in the earlier part of the 20th century (El-
Najjar, 1976). 7 For these individuals age at death was
estimated by assessments of dental maturation and eruption
status in accordance with published standards (Krogman
and Iscan, 1986).
Subadults that were represented in the Anasazi and
Peruvian skeletal samples ranged in age from birth to ten
years. Although both skeletal series have been classified as
maize dependent, the Anasazi group were primarily
horticulturalists that supplemented dietary intake by
foraging (El-Najjar et al., 1976). In contrast, the Peruvian
group practiced intensive agriculture where subsistence
was highly dependent upon the use of domesticated
cultigens (El-Najjar, 1976). For the Anasazi and Peruvian
subadult samples combined, this worker observed a total
frequency of porotic hyperostosis that was 75.8 percent
(94/124). Similarly, frequency data reported by El-Najjar
(1976) yielded a figure of 74.8 percent (80/107) for
subadults in these two groups. Inter-observer error was
approximately one percent. Therefore, the methods used by
El-Najjar and associates (1976) to identify porotic
hyperostosis were found to be highly replicable.
Data enumerating the age-specific frequencies for
unremodeled, remodeled, and extra-orbital porotic
hyperostosis that was observed in Anasazi and Peruvian
subadults (0-10 years) is given in Table 16. Summary data
which compare the total frequency for porotic hyperostosis
in Bt-5, Libben, Anasazi, and Peruvian subadults that
ranged in age from birth to ten years are listed in Table 17.
The age-specific frequency with which unremodeled
porotic hyperostosis occurred in these groups is illustrated
in Figure 8. Results show that the age at onset, peak
incidence, and age at remission for unremodeled lesions
are markedly similar in each group. The low-to-high rank
order frequency for unremodeled, total, and extraorbital
lesions that were observed is Bt-5, Libben, Anasazi, and
Peruvian (Table 17). Chi square values that compare the
total frequencies with which porotic hyperostosis occurred
in each group are listed in Table 18. Results show that (1)
the frequency of total lesions and extra-orbital lesions were
significantly different in each group, and (2) the maize
dependent subadults had a significantly higher frequency
of total lesions and extra-orbital lesions compared to the
non-maize dependent Libben and Bt-5 subadults.
However, there is strong evidence to suggest that the two
maize dependent subadult samples are demo-graphically
34
Mensforth
No. 46
TABLE 16. Age-Specific Frequency of Occurrence for Porotic Hyperostosis in Prehistoric Anasazi and Coastal Peruvian Subadults
(0-10 years)
Anasazi Coastal Peruvian
Age
Group
«/
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
"i
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
n2
(%)
(%)
»4
(%)
n2
(%)
(%)
>U
(%)
0.0 - 0.5
1
0
0.0
0
0.0
0
0.0
2
0
0.0
0
0.0
0
0.0
0.5 - 1.0
4
2
50.0
0
0.0
2
50.0
4
2
50.0
0
0.0
2
50.0
1.0 - 3.0
18
13
72.2
1
5.6
14
77.8
24
19
79.2
4
16.7
23
95.8
3.0 - 5.0
13
5
38.5
2
15.4
7
53.8
13
4
30.8
5
38.5
9
69.2
5.0 -10.0
18
3
16.7
10
55.6
13
72.2
27
8
29.6
16
59.3
24
88.9
Total
54
23
42.6
13
24.1
36
66.7
70
33
47.1
25
35.7
58
82.9
Symbols: tip total number of specimens that could be examned for the presence/absence of porotic hyperostosis per age group; n2i number of specimens that were scored positive for
unremodeled lesions per age group; n2 number of specimens that were scored positive for remodeled lesions per age group; n4 total number of specimens that were scored positive
for remodeled or unremodeled lesions per age group.
TABLE 17. Comparison of the Total Frequency With Which Porotic Hyperostosis Occurred
Among Subadults (0-10 years) From Four Culturally and Temporospatially Diverse Skeletal Groups
Unremodeled Remodeled Total Extra-orbital
Lesions Lesions Lesions Lesions
Skeletal
Group
1l
n2
(%)
(%)
n4
(%)
»5
(%)
Bt-5
88
12
13.6
7
8.0
19
21.6
3
3.4
Libben
250
77
30.8
30
12.0
107
42.8
31
12.4
Anasazi
54
23
42.6
13
24.1
36
66.7
18
33.3
Peruvian
70
33
47.1
25
35.7
58
82.9
46
65.7
Total
462
145
31.4
75
16.2
220
21.2
98
21.2
Symbols: n]t total number of specimens that could be examined for the presence/absence of porotic hyperostosis; n-,, number of
specimens that were scored positive for unremodeled lesions; n3, number of specimens that were scored positive for remodeled
lesions; n4, total number of specimens that were scored positive for remodeled and/or unremodeled lesions; n5 total number of
specimens that exhibited extra-orbital lesions.
unsound as a result of selective recovery practices. For
example, even though the state of preservation is nothing
less than superb for these two groups, each skeletal
collection primarily consists of isolated crania alone. That
is, post-cranial materials were saved for only 22.2 percent
(12/54) of Anasazi subadults and 14.3 percent (10/70) of
Peruvian subadults. In addition, reference to Table 16 shows
that Anasazi and Peruvian infants (0-1 year) are grossly
under-represented in each group. When compared to Libben
and Bt-5, where infants with a low frequency of porotic
hyperostosis are present in substantial numbers, this will
have the net effect of inflating lesion frequencies in the
Anasazi and Peruvian samples when all individuals from
birth to ten years of age are compared. Therefore, the
comparisons presented earlier are biased and require
adjustment before valid conclusions can be reached.
Because infants are markedly under-represented in the
maize dependent samples, these individuals were excluded
from further consideration. Adjusted summary data given in
Table 19 compares the total frequency with which porotic
hyperostosis was observed in Bt-5, Libben, Anasazi, and
Peruvian subadults that ranged in age from one to ten years.
Chi square values that compare these adjusted lesion
frequencies are listed in Table 20. Most between group
differences that were reported earlier are still apparent for
the adjusted data. However, unremodeled and total lesion
frequencies in Libben and Anasazi subadults are now quite
similar. Thus, the extent to which dietary differences may
have resulted in a greater frequency of porotic hyperostosis
in Anasazi versus Libben subadults (i.e., the pattern that
was observed for comparisons of unadjusted lesion
frequencies) is now obscure. Only extraorbital lesion
frequencies remain significantly different for all four
subadult samples that were examined.
1991
Paleoepidemiology of P orotic Hyperostosis
35
FIGURE 8. Age-specific frequencies of unremodeled porotic
hyperostosis that were observed among subadults ( 0-10 years )
from four culturally diverse skeletal sample.
Alternatively, it is suggested that a fundamental
difference in levels of infectious disease load may have been
more important in generating the subadult porotic
hyperostosis frequency distributions that were observed
among Libben subadults. That is, in addition to the
demographic patterns reported here for the Libben and Bt-5,
comparative analyses of tibia long bone growth perform-
ance, and frequency data for periosteal reactions, that were
observed in subadults from these two groups, support the
inference that Libben infants and children experienced
higher levels of morbidity and mortality in response to
elevated levels of infectious disease.
Results published elsewhere showed that Libben
children six months to two years of age experienced a major
period of slowed growth relative to Bt-5 (Mensforth, 1985).
Growth suppression continued until four years of age when
Libben children accrued a maximum 9.94 percent
decrement in tibia growth relative to Bt-5. Furthermore,
Libben subadults, displayed a frequencies of periosteal
reactions that were significantly greater compared to Bt-5
children (%2=7.29; p<.01). In particular, for those ranging in
age from birth to three years in particular, 51% of Libben
and 27% of Bt-5 infants and young children displayed
periosteal reactions (%2=12.24; p<.001). Among individuals
affected in this age group, 79% of Libben and 65% of Bt-5
young children displayed unremodeled, or active, lesions at
time of death. These skeletal lesions were interpreted to
represent common microbial infections that had become
established in bony tissues as a result of hematogenous
seeding (Mensforth, 1986, 1985; Mensforth et al., 1978).
TABLE 18. Chi Square Values That Compare the Frequency of Occurrence for Porotic Hyperostosis
in Subadults (0-10 years) From Four Culturally and Temporospatially Diverse Skeletal Groups
Skeletal Group
Skeletal
Group
Lesion
Status
Libben
Anasazi
Peruvian
Bt-5
Unremodeled Lesions
9Mb
15.1 1°
21.49"
Extra-orbital Lesions
5.82c
23.78"
70.74"
Total Lesions
15.52"
28.65"
58.57"
Libben
Unremodeled Lesions
2.80
6.47c
Extra-orbital Lesions
14.39"
85.07"
Total Lesions
10.15/j
35.13"
Anasazi
Unremodeled Lesions
0.25
Extra-orbital Lesions
12.80"
Total Lesions
4.36"
a: significant at the .001 level of probability
b: significant at the .01 level of probability.
c: significant at the .05 level of probability.
36
Mensforth
No. 46
TABLE 19. Comparison of the Total Frequency With Which Porotic Hyperostosis Occurred
Among Subadults (1-10 years) From Four Culturally and Temporospatially Diverse Skeletal Groups
Skeletal
Group
nl
Unremodeled
Lesions
Remodeled
Lesions
Total
Lesions
Extra-orbital
Lesions
n2
(%)
n3
(%)
n4
(%)
n5
(%)
Bt-5
50
9
18.0
7
14.0
16
32.0
1
2.0
Libben
151
67
44.4
30
19.9
97
64.2
24
15.9
Anasazi
49
21
42.9
13
26.5
34
69.4
16
32.7
Peruvian
64
31
48.4
25
39.1
56
87.5
44
68.8
Total
314
128
40.8
75
23.9
203
64.6
85
27.1
Symbols: nt, total number of specimens that could be examined for the presence/absence of porotic hyperostosis; n2, number of
specimens that were scored positive for unremodeled lesions; n ,, number of specimens that were scored positive for remodeled
lesions; n4, total number of specimens that were scored positive for remodeled and/or unremodeled lesions; n5 total number of
specimens that exhibited extra-orbital lesions.
TABLE 20. Chi Square Values That Compare the Frequency of Occurrence for Porotic Hyperostosis
in Subadults ( 1-10 years) From Four Culturally and Temporospatially Diverse Skeletal Groups
Skeletal Group
Skeletal
Group
Lesion
Status
Libben
Anasazi
Peruvian
Bt-5
Unremodeled Lesions
11.11°
1.24b
1 1 .42°
Extra-orbital Lesions
6.66b
16.35°
52.35°
Total Lesions
15.86°
13.84°
37.16°
Libben
Unremodeled Lesions
0.03
0.30
Extra-orbital Lesions
6.49c
58.07°
Total Lesions
0.43
11.85°
Anasazi
Unremodeled Lesions
0.35
Extra-orbital Lesions
5.62c
Total Lesions
14.52°
a: significant at the .001 level of probability
b: significant at the .01 level of probability.
c: significant at the .05 level of probability.
The extent to which habitation of the Black Swamp
elevated risk of acquiring infectious diseases among the
Libben people remains enigmatic. Prior to being drained, in
recent historic times the relatively unchanged Black Swamp
was dreaded by pioneers. However, the Black Swamp
acquired its reputation because it was a formidable barrier
to travel and settlement (Kaatz, 1955). Thus, at least in the
recent historic past, the Black Swamp has no reputation for
being a source of profound human misery related to disease.
Nonetheless, a more specific scenario could be posited
where we might expect aboriginal inhabitants of the rich
Black Swamp to have experienced high and prolonged
seasonal exposures to mosquito, and other, insect bites.
Among infants and children, these circumstance alone could
result in cumulative blood losses that could precipitate, or
exacerbate, a latent iron deficient state. These, in turn, could
easily elevate risk of microbial invasion. The latter would
occur either by direct introduction of pathogens at the time
when a blood meal is being secured, or indirectly as a
consequence of skin irritation and inflammation. In these
1991
Paleoepidemiology of P orotic Hyperostosis
37
ways, pathogens which normally colonize the skin and
maintain a symbiotic relationship, such as Staphylococcus
aureus, experience greater opportunities for capillary invasion
and hematogenous seeding of the host. Alternatively, food
preparation techniques, or the selective lack thereof, may have
promoted a greater frequency, intensity, and duration of
gastrointestinal tract infections in Libben infants and children.
While neither of the hypotheses presented above, either
alone or in combination, can be accepted or refuted,
skeletal evidence presently available favors the former
relationship. That is, the majority of periosteal reactions
observed in Libben subadults (1) occurred in individuals
under three years of age (51% affected), (2) the highest
frequency characterized infants that died in the first year of
life (57% affected), (3) patterns of skeletal involvement
clearly indicate that the majority of affected individuals
suffered from a systemic blood-borne infectious disease,
and (4) the infectious disease episodes suffered by affected
individuals persisted for periods of time sufficient to elicit
a frank bony response to the underlying stress episode
(Mensforth et al, 1978; Mensforth, 1986).
Although specific agents involved in the etiology of
Libben and Bt-5 porotic hyperostosis and periosteal
reactions remain speculative, the strong synergistic
relationships between iron deficiency anemia and
infectious disease discussed earlier cannot be ignored (i.e.,
impaired immune response, hypoferremia, and
malabsorption syndrome). Comparative population studies
have repeatedly demonstrated that infections, primarily
those of the respiratory and gastrointestinal tracts, have a
negative influence on hemoglobin status and play a
fundamental role in precipitating nutritional crises such as
iron deficiency anemia (Maynard and Hammes, 1970; Jose
and Welch, 1970; Kaplan et al, 1973; Burks et al., 1976).
These effects are most pronounced in children under three
years of age where infectious disease and nutritional
deficiency commonly results in the pneumonia-weanling
diarrhea syndrome (Gordon et al, 1963). The synergistic
consequences of anemia and infection in contemporary
underdeveloped societies are elevated morbidity, mortality,
and early onset of growth retardation (Maynard and
Hammes, 1970; Jose and Welch, 1970; Burks et al., 1976).
The extent to which increased population densities and
degrees of sedentism may have resulted in a greater
prevalence of microbial and parasitic infectious diseases,
nutritional deficiencies, and growth retardation in the
Anasazi and Peruvian skeletal groups cannot be assessed.
This is due to the fact that these samples lack suitable
demographic documentation and are further characterized
by infant under-representation and poor recovery of post-
cranial skeletal materials. However, Kent (1987) has
recently reported a case study where it is cogently argued
that higher population densities, greater sedentism, and
elevated levels of infectious disease may have played a
more substantial role than maize diet in the etiology of
porotic hyperostosis in prehistoric Anasazi populations.
For purposes of demographic inference, algorithms la
and lb were modified so that Libben and Bt-5 porotic
hyperostosis frequencies adjusted for the one to ten year
subadult age period could be used to predict survivorship at
age 15 (115) for the Anasazi and Peruvian groups. These
equations are as follows:
2a. y =
y -
X\ -
2b. y =
y =
X2 ~
-0.3295(xi) + 67.532
ll5
unremodeled porotic hyperostosis
frequency that is observed in subadults
(1-10 years)
0.2702(a2) + 70.246
•l5
total porotic hyperostosis frequency
that is observed in subadults (1-10
years)
These algorithms predict average values for survivorship
at age 15 in the Anasazi and Peruvian groups which
should approximate 52.5 + 1.3 and 49.1 + 3.5,
respectively. The degree to which porotic hyperostosis
lesion frequencies accurately reflect these demographic
projections will await further investigations of these two
groups that employ sound demographic data. Finally,
reference to Tables 19 and 20 show that extraorbital
porotic hyperostotic lesion frequencies were significantly
different for comparisons of all four subadult skeletal
samples. Indeed, it was found that extra-orbital lesions
accumulated at an exponential rate relative to increments
in total lesion frequencies that were observed in band level
foragers and maize dependent subadults. This relationship
can be expressed as follows:
y-e ax - 1
y = extra-orbital lesion frequency predicted
.V = total porotic hyperostosis frequency observed
e = 2.7182818
a = .04721
The low-to-high rank order frequencies for extra-orbital
lesions are Bt-5 (2.0%), Libben (15.9%), Anasazi (32.7%),
and Peruvian (68.8%). These figures suggest that maize
utilization by the Anasazi and Peruvian groups, combined
with higher levels of population density and degree of
sedentism, resulted in a greater duration and severity of
iron deficiency anemia compared to the Libben and Bt-5
skeletal populations. Moreover, it is most probable that
endemic hookworm disease played a substantial role in the
etiology of chronic iron deficiency anemia seen in Peruvian
infants and children.
38
Mensforth
No. 46
With regard to porotic hyperostosis, the view posited
here is that extra-orbital lesion frequencies may provide a
superior skeletal index of the extent to which chronic iron
deficiency anemia was prevalent in earlier human groups.
As such, a useful index as such can be calculated by
dividing the frequency of extra-orbital lesions observed by
the frequency of total lesions observed. For the subadult
skeletal samples described above these ratios are as
follows: Bt-5, 0.06; Libben, 0.25; Anasazi, 0.47; Peruvian,
0.79. The benefit of this measure is its ability to detect the
presence of chronic or protracted episodes of erythroid
marrow hyperplasia regardless of whether total lesion
frequencies are low, moderate, or high in skeletal samples
under investigation.
Summary and Conclusions
The age and sex related frequencies with which porotic
hyperostosis occurred in the Libben Late Woodland (»=580)
and Bt-5 Late Archaic (n= 247) skeletal populations were
examined. The goals of the study were to (1) identify those
factors which were important in the etiology of the skeletal
lesions that were observed in each group, and (2) evaluate
the extent to which porotic hyperostosis serves as a useful
bioassay of disease and nutritional stress in prehistoric
human populations. It was suggested that stress indicators
of greatest paleoepidemiological utility will be those that
display high levels of biological and demographic
sensitivity combined with a low level of nonspecificity.
Qualitative assessments of lesion activity status (i.e.,
remodeled or unremodeled) and patterns of skeletal
involvement (orbital versus extra-orbital; cranial versus
post-cranial, and slight, moderate, and severe degrees of
bony response) were recorded for all Libben and Bt-5
individuals that displayed porotic hyperostosis at time of
death. Unremodeled lesions were used to establish the age-
specific frequency distributions for the lesions and identify
those age/sex groups at greatest risk of acquiring the
disorder. Patterns of skeletal involvement aided in
differential diagnoses and served as a skeletal index of the
degree to which an underlying hypoxic stimulus resulted in
a greater duration and severity of the stress episode.
Results showed that the majority of Libben and Bt-5
porotic hyperostosis consisted of slight bony involvements
that were restricted to the superior orbital plates of affected
individuals. More pronounced degrees of osseous tissue
response were somewhat more common in individuals that
were six months to three years of age at time of death. No
individuals in either group displayed skeletal changes that
are considered diagnostic for hemolytic anemias such as
sickle cell anemia or thalassemia.
Within group comparisons for broad age/sex categories
showed that (a) subadults had a significantly greater
frequency of unremodeled lesions compared to adults, (b)
adult females had a significantly higher frequency of
unremodeled lesions compared to adult males, and (c)
young adult females had a greater incidence of unremodeled
lesions compared to females over 30 years of age. However,
the latter difference was only significant for the Libben
group.
Patterns in the ages at onset, peak incidence, and
remission for unremodeled and remodeled porotic
hyperostosis were similar in the two groups. These data
differed in no substantial way from patterns predicted by the
hypothetical iron deficiency anemia related porotic
hyperostosis model that was presented. Indeed, Libben and
Bt-5 age and sex specific frequency distributions for
unremodeled porotic hyperostosis clearly identified those
age/sex groups at greatest risk of acquiring iron deficiency
anemia as a result of intrinsic physiological and extrinsic
environmental risk factors.
Sex-specific between group comparisons showed that
Libben and Bt-5 adult males, and adult females, displayed
similar frequencies of porotic hyperostosis. The only
substantial difference that was observed between the two
groups was confined to subadults where Libben children
displayed a significantly greater frequency of unremodeled
and total lesions compared to Bt-5 children.
Compared to most other conditions that promote
erythroid marrow hyperplasia, only iron deficiency anemia
and/or the congenital hemolytic anemias (i.e., sickle cell
anemia or thalassemia) are sufficiently prevalent to account
for the frequencies with which porotic hyperostosis has
been observed in earlier human groups. Therefore, it was
concluded that porotic hyperostosis is characterized by a
very low level of non-specificity relative to most other
skeletal indicators of disease and nutritional stress.
Comparisons of Libben and Bt-5 average mortality ratios
and porotic hyperostosis frequency ratios for subadults, and
adults (sex-combined), were directionally concordant. Thus,
porotic hyperostosis exhibits demographic sensitivity as a
bioassay of health conditions for the two groups. The only
exception to this was that sex ratios for adult lesions
frequencies did not conform to the patterns of adult male and
female survivorship that characterized each group. That is,
mortality rates were approximately equal for the sexes in
each group whereas adult female: male lesion frequency
ratios were skewed. Therefore, it was concluded that porotic
hyperostosis is demographically insensitive as an indicator of
differential mortality experience in adult males and females.
Infants and young children from six months to four years
of age were found to be hypersensitive to skeletal changes
in the anemias. This appears to be a primary consequence of
the age-specific competitive relationship between limited
supply and increased demands for active marrow space,
combined with numerous constitutional and environmental
factors that elevate risk of developing iron deficiency
1991
Paleoepidemiology oe P orotic Hyperostosis
39
anemia in very young individuals. In contrast, adults are
regarded as hyposensitive to skeletal changes in the anemias
due to the fact that substantial hematopoietic bone marrow
reserves are available to these individuals. Thus, porotic
hyperostosis exhibits a biological sensitivity that bears a
strong relationship with developmental age.
The epidemiological patterns that were observed for
Libben and Bt-5 porotic hyperostosis support the
interpretation that iron deficiency anemia was the primary, if
not sole, cause of the skeletal lesion in these two groups.
With regard to etiology, however, the following
circumstances must be emphasized. Floral and faunal
analyses indicate that both band level societies exploited a
diverse abundance of local dietary resources. Moreover, no
subadults or adults in either group exhibit skeletal evidence to
suggest that chronic malnutrition owing to dietary inadequacy
was a major factor influencing health status. Therefore, diet
probably played a very minor role in the etiology of porotic
hyperostosis at Libben and Bt-5. Similarly, evidence from
contemporary epidemiological surveys suggests that
parasitism played no significant role in precipitating
nutritional crises in these earlier human groups.
Aside from cultural affiliations and temporospatial
distributions, the principal difference between the Libben
and Bt-5 skeletal samples concerns the fact that levels of
subadult mortality were substantially greater. Flere, it is
suggested that local environmmental circumstances
associated with the habitation and exploitation of the Black
Swamp may have played a fundamental role in elevating
infectious disease loads which resulted in a greater
prevalence of iron deficiency anemia in Libben children.
The strong synergistic relationships between iron deficiency
anemia and infectious disease are known to have their
greatest impact on patterns of infant and child morbidity,
mortality, and growth performance. Moreover, these
relationships are concordant with the findings that Libben
subadults experienced greater frequencies of porotic
hyperostosis, periosteal reactions, and an early onset of long
bone growth retardation compared to Bt-5 children.
Given that subadults exhibit a marked biological and
demographic sensitivity to erythroid marrow hyperplasia (1)
algorithms were presented whereby subadult porotic
hyperostosis frequencies can be used to predict survivorship
at age 15 in earlier human groups, and (2) porotic
hyperostoss frequencies for Libben and Bt-5 subadults were
compared to those observed in two prehistoric maize
dependent skeletal samples. When the latter comparisons
were restricted to the one to ten year age interval, in order to
adjust for infant under-representation in the Anasazi and
Peruvian samples, it was found that Libben and Anasazi
subadults had similar frequencies of unremodeled and total
lesions. These findings suggest that similarities in degree of
sedentism and population density may be responsible for
the similarity in lesion frequencies that were observed.
Alternatively, it is suggested that increased sedentism,
population density, and high levels of endemic hookworm
infestation contributed to the elevated frequency of porotic
hyperostosis that characterizes the sample of Peruvian
subadults.
In contrast, the frequency with which extra-orbital
porotic hyperostosis occurred was significantly different in
all four subadult samples that were examined. It was found
that extra-orbital lesions accumulate at an exponential rate
relative to total lesions, and may therefore provide a
superior skeletal index of the extent to which chronic iron
deficiency anemia occurred in earlier human groups. The
greater prevalence of extra-orbital lesions in the maize
dependent subadult samples may reflect the extent to which
dietary inadequacy promoted a greater duration and severity
of iron deficiency anemia in these groups.
Finally, it is interesting that few investigators concerned
with the pathogenesis of porotic hyperostosis in Old World
skeletal groups have seriously entertained the iron
deficiency anemia hypothesis (for exceptions see Hengen,
1971 and Carlson et al., 1974). This is particularly true for
researchers who have been concerned with the etiology of
the lesion in circum-mediterranean skeletal groups (Angel,
1967; Ascenzi and Salistreri, 1977; and Germana and
Ascenzi, 1980). Given the fact that iron deficiency anemia
is prevalent throughout the world (Witts, 1966), particularly
in agricultural communities (World Health Organization,
1968), it is clear that the iron deficiency and hemolytic
anemia hypotheses cannot reasonably be considered as
mutually exclusive explanations for the occurrence of
porotic hyperostosis in Old World circum-mediterranean
skeletal populations. It is highly improbable that those
human groups subjected to the selection pressures of
falciparum malaria would be uniquely exempt from basic
human, and mammalian, pathophysiological responses to
disease and nutritional stress. Likewise, the iron deficiency
and hemolytic anemia hypotheses cannot reasonably be
invoked to explain the occurrence of the skeletal lesion on a
discrete continental basis as has been suggested (Angel,
1967). Thus, many studies that have investigated the
pathogenesis of porotic hyperostosis in Old World
prehistoric groups are clearly in need of substantial revision.
Acknowledgements
I would like to thank George Amielagos, Richard Meindl,
and several anonymous reviewers who provided valuable
comments and criticisms on earlier versions of this manuscript.
I would also like to thank John Blank, members of the
Cleveland State University Instructional Media Services, and
members of the University of Oklahoma Electron Microscopy
Laboratory who assisted in the preparation of photographs and
illustrations used in this study.
40
Mensforth
No. 46
Notes
1. The Libben site skeletal materials are permanently curated
by the Department of Sociology and Anthropology at Kent
State University, Kent, Ohio.
2. The Carlston Annis Bt-5 site skeletal materials are
permanently curated by the Department of Anthropology at
the University of Kentucky, Lexington.
3. The criterion that was used to identify lesions of minimal
expression conform to those employed by El-Najjar et al.
(1976) where one or more clusters of pores which extended 5
millimeters in diameter or greater were considered as
indicating presence of the lesion. Lesions of moderate
expression typically involved a greater surface area of the
affected bone, displayed a well-developed cribriform mesh,
and usually exhibited clearly discernable osseous tissue
hypertrophy. Lesions of severe expression were interpreted to
represent skeletal changes that were associated with extensive
osseous tissue hypertrophy, marked expansion of diploic
spaces, and commonly involved the frontal, parietal, and
occipital bones in combination with the superior orbital plates.
4. In the present study porotic hyperostotic lesions were
simply classified as orbital or extra-orbital with respect to
cranial bones affected. Other workers have suggested lesion
typologies based on physical appearance of the lesion or the
anatomical disposition of the lesion in the skull. For example,
Nathan and Haas (1966) described orbital porotic hyperostosis
as being porotic, cribrotic, or trabeculated in appearance.
Carlson and associates (1974) have alternatively suggested
that lesions be classified as cribra orbitalia, spongy
hyperostosis, and osteoporotic pitting. These descriptive terms
were not used in the present study because they primarily
reflect degree of osseous tissue response and no further
pathophysiological significance can be attributed to the use of
such jargon. Thus, the terminology employed here
corresponds with that currently advocated by several
researchers who have directed their long term efforts toward
developing a bettern comprehension of the nature and
significance of porotic hyperostosis in earlier human groups
(G. J. Armelagos, personal communication).
5. Age-specific Libben:Bt-5 qx ratios are listed in Table 2.
The average Libben:Bt-5 qx ratios for adults and subadults
were estimated by calculating the mean for all qx ratios that
were observed for each age group. The values for adults (+15
years) and subadults (0-15 years) are 1.24 and 1.60,
respectively.
6. The Anasazi skeletal materials are permanently curated by
the American Museum of Natural History, New York.
7. The Peruvian skeletal materials are permanently curated by
the Chicago Field Museum, Illinois.
References
Aegerter, E. and J.A. Kirkpatrick. 1975. Orthopedic Diseases:
Physiology, Pathology, and Radiology. (4th edition.) W.B.
Saunders Co., Philadelphia.
Agarwal, K., N. Dhar, M.N. Shah, and O.P. Bhardway. 1970.
Roentgenologic changes in iron deficiency anemia. American
Journal of Roentgenology, Radiation Therapy, and Nuclear
Medicine 110:635-637.
Akel, R.N., J.D. Frankish, C.P. Powles, K.R. Tyler, J. M. Watt, H. J.
Weston, and I. A.M. Prior. 1963. Anemia in Maori and
European infants and children on admission to hospital: a
cooperative survey from six New Zealand hospitals. New
Zealand Medical Journal 62:29-33.
Andelman, M.B., and B.R. Sered. 1966. Utilization of dietary iron
by term infants. American Journal of Diseases of Children
111:45-55.
Aksoy, M., N. Camli, and S. Erdem. 1966. Roentgenographic bone
changes in chronic iron deficiency anemia. Blood 27:677-686.
Angel, J.L. 1964. Osteoporosis: thalassemia? American Journal of
Physical Anthropology 22:369-374.
Angel, J.L. 1966. Porotic hyperostosis, anemias, malarias, and
marshes in the prehistoric eastern Mediterranean. Science
153:760-763.
Angel, J.L. 1967. Porotic hyperostosis or osteoporosis symmetrica,
pp. 378-389. In: Diseases in Antiquity. Edited by Sandison and
Brothwell. Thomas Publishing Co., Springfield, Illinois.
Apte, S.V., and L. Iyengar. 1970. Absorption of dietary iron in
pregnancy. American Journal of Clinical Nutrition 23:13-71.
Arbeter, A., L. Echeverri, D. Franco, D. Munson, H. Velez, and J. J.
Vitale. 1971. Nutrition and Infection. Federation Proceedings
30:1421-1428.
Ashworth, A., P.F. Milner, J. C. Waterlow, and W. B. Walker. 1973.
Absorption of iron from maize (Zea mays L.) and soya beans
(Glycine hispida Max.) in Jamaican infants. British Journal of
Nutrition 29:269-278.
Ascenzi, A., and P. Salistreri. 1977. Porotic hyperostosis and the
problem of the origin of thalassemia in Italy. Journal of
Human Evolution 6:595-604.
Badenoch, J., J.R. Evans, and W.C.D. Richards. 1957. The
stomach in hypochromic anemia. British Journal of
Hematology 3:175-185.
Baggs, R.B., and S.A. Miller. 1973. Nutritional iron deficiency as a
determinant of host resistance in the rat. Journal of Nutrition
103:1554-1560.
Baker, D.H. 1964. Roentgen manifestations of Cooley’s anemia.
Annals of the New York Academy of Sciences 1 19:642-661 .
Baker, S.J. 1978. Nutritional anaemia — a major controllable public
health problem. Bulletin of the World Health Organization
55:659-675.
Becker, M.H., N.B. Genieser, S. Piomelli, D. Dove, and R.D.
Mendoza. 1971. Roentgenographic manifestations of pyruvate
kinase deficiency hemolytic anemia. American Journal of
Roentgenology, Radiation Therapy, and Nuclear Medicine
113:491-498.
1991
Paleoepidemiology of P orotic Hyperostosis
41
Beisel, W.R. 1982. Single nutrients and immunity. American
Journal of Clinical Nutrition 35: 417-468.
Betke, K. 1970. Iron deficiency in children, pp. 519-528. In: Iron
Deficiency: Pathogenesis, Clinical Aspects, Therapy. Edited by
Hallberg, Harwerth, and Vanotti. Academic Press, New York.
Betts, N.M., and A. Weidenbenner. 1986. Dietary intakes, iron
status and growth status of Southeastern Asian refugee
children. Nutrition Research 6:509-515.
Bhaskaram, P., J.S. Prasad, and K.A. Krishnamanchari. 1977.
Anemia and immune response. Lancet 1 : 1000.
Bhaskaram, P. and V. Reddy. 1975. Cell-mediated immunity in iron
and vitamin-deficient children. British Medical Journal
(Clinical Research) 3:522.
Bowen, J.E. 1990. The late prehistory of northwestern Ohio. Ph.D.
Dissertation, Ohio State University.
Bradfield, R.B., V.V. Jensen, L. Gonzales, and C. Garrayar. 1968.
Effect of low-level iron and vitamin supplementation on a
tropical anemia. American Journal of Clinical Nutrition 21:57-67.
Brendstrup, P. 1950. Intravenous iron loadings of patients with
infections. Acta Medica Scandinavica 239:201-209.
Bressani, R„ R. Paz, and N.S. Scrimshaw. 1958. Chemical changes
in corn during the preparation of tortillas. Journal of
Agriculture and Food Chemistry’ 6:770-774.
Britton, H.A., J.P. Canby, and C.M. Kohler. 1960. Iron deficiency
anemia producing evidence of marrow hyperplasia in the
calvarium. Pediatrics 25:621-627.
Brubaker, R.R., E.D. Beesley, and M.J. Surgalla. 1965. Pasteurella
pestis: role of pesticin 1 and iron in experimental plague.
Science 149:422-424.
Bullen, J.J., H.J. Rogers, and L. Leigh, 1972. Iron-binding proteins
in milk and resistance to Escherichia coli in infants. British
Medical Journal 1:69-75.
Bullen, J.J., and H.J. Rogers. 1968. Effect of haemoglobin on
experimental infections with Pasteurella septica and
Escherichia coli. Nature 217:86.
Bullen, J.J., L.C. Leigh, and H.J. Rogers. 1968. The effect of iron
compounds on the virulence of Escherichia coli for Guinea-
pigs. Immunology 15:581-588.
Burko, H., H.Z. Mellins, and J. Watson. 1961 . Skull changes in iron
deficiency anemia simulating congenital hemolytic anemia.
American Journal of Roentgenology 86:447-452.
Burks, J.M., M.A. Siimes, W.C. Mentzer, and PR. Dallman. 1976.
Iron deficiency in an Eskimo village. The value of serum ferritin
in assessing iron nutrition before and after a three month period
of iron supplementation. Journal of Pediatrics 88:224-228.
Caffey, J. 1939. The skeletal changes in the chronic hemolytic
anemias (erythroblastic anemia, sickle cell anemia, and chronic
hemolytic icterus). American Journal of Roentgenology
37:293-324.
Callender, S.T., B.J. Mallett, and M.D. Smith. 1957. Absorption of
hemoglobin iron. British Journal of Hematology 3: 1 86- 1 92.
Carlson, D.S., G.J. Armelagos, and D.P. Van Gerven. 1974. Factors
influencing the etiology of cribra orbitalia in prehistoric Nubia.
Journal of Human Evolution 3:405-410.
Carpenter, C. 1987. Infectious diarrheal syndromes, pp. 502-506. In:
Harrison's Principles of Internal Medicine. (11th edition.)
Edited by E. Braunwald et al. McGraw Hill, New York.
Cartwright, G.E., M.A. Lauritsen, P.J. Jones, I.M. Merril, and M.M.
Wintrobe. 1946. The anemia of infection. I. Hypoferremia,
hypercupremia, and alterations in porphyrin metabolism in
patients. Journal of Clinical Investigation 25:65-80.
Cassidy, C.M. 1980. Nutrition and health in agriculturalists and
hunter-gatherers: a case study of two prehistoric populations,
pp. 117-145. In: Nutritional Anthropology: Contemporary
Approaches to Diet and Culture. Edited by R. F. Kandal and G.
H. Pelto. Redgrave, Pleasantville, New York.
Chandra, R.K. 1973. Reduced bactericidal capacity of polymorphs in
iron deficiency. Archives of Diseases of Childhood 48:864-866.
Chandra, R.K. 1985. Trace element regulation of immunity and
infection. Journal of American College of Nutrition 4:5-16.
Chandra, R.K. and B. Au. 1980. Single nutrient deficiency and cell-
mediated immune responses. III. Vitamin A. Nutritional
Research 1:181-1 85.
Chandra, R.K. and A.K. Saraya. 1975. Impaired immuno-
competence associated with iron deficiency anemia. Journal
of Pediatrics 86:899-902.
Charley, P.J., B. Sarkar, C.F. Stitt, and P. Saltman 1963. Chelation of
iron by sugars. Biochemica and Biophysica, Acta 69:3 1 3-32 1 .
Cohen, M.N. and G.J. Armelagos. 1984. Paleopathology at the
Origins of Agriculture. Academic Press, New York.
Conrad, M.E. 1970. Factors affecting iron absorption, pp. 87-120.
In: Iron Deficiency: Pathogenesis, Clinical Aspects, Therapy.
Edited by Hallberg, Harwerth, and Vanotti. Academic Press,
New York.
Cook, J.D., and E.R. Monsen. 1976. Food iron absoiption in human
subjects. III. Comparison of the effect of animal proteins on
nonheme iron absorption. American Journal of Clinical
Nutrition 29:859-867.
Cook, J.D., G.M. Brown, and L.S. Valberg. 1964. The effect of
achylia gastrica on iron absorption. Journal of Clinical
Investigation 43: 1 1 85- 1 1 9 1 .
Cule, J., and I.L. Evans. 1968. Porotic hyperostosis and the
Gelligaer Skull. Journal of Clinical Pathology 2 1 :753-758.
Davidson, L.S.P., H.W. and Fullerton. 1938. Chronic nutritional
hypochromic anemia. Edinburgh Medical Journal 45:1-23.
Davidson, L.S.P., H.W. Fullerton, and R.M. Campbell. 1935.
Nutritional iron deficiency anemia with special reference to
prevalence and age and sex incidence. British Journal of
Medicine (August):389 1-3894.
Davidson, W.M.B., and J.L. Markson. 1955. The gastric mucosa in
iron deficiency anemia. Lancet (2):639-643.
Davis, PS. 1970. The effect of iron-binding proteins on iron
absorption, pp. 129- 134. In: Iron Deficiency: Pathogenesis,
Clinical Aspects, Therapy. Edited by Hallberg, Harwerth, and
Vanotti. Academic Press, New York.
Dawson, J.P, and J.F. Desforges. 1958. Dietary and storage factors
in iron deficiency anemia of infancy. American Journal of
Diseases of Children 96: 169-178.
42
Mensforth
No. 46
Dykstra. O.H., and T. Halberstma. 1940. Polycythemia vera in
childhood (report of a case with changes in the skull).
American Journal of Diseases of Children 60:907-916.
El-Najjar, M.Y. 1976. Maize, malaria, and the anemias in the pre-
Columbian New World. Yearbook of Physical Anthropology
20:329-337.
El-Najjar, M.Y., B. Lozoff, and D.J. Ryan. 1975. The paleo-
epidemiology of porotic hyperostosis in the American
Southwest: radiological and ecological considerations.
American Journal of Roentgenology, Radiation Therapy, and
Nuclear Medicine 125:91 8-924.
El-Najjar, M.Y., and A.L. Robertson. 1976. Spongy bones in
prehistoric America. Science 193:141-143.
El-Najjar, M.Y., D.J. Ryan, C.G. Turner, and B. Lozoff. 1976. The
etiology of porotic hyperostosis among the prehistoric and
historic Anasazi Indians of Southwestern United States.
American Journal of Physical Anthropology 44:477-488.
Eng, L.L. 1958. Chronic iron deficiency anemia with bone
changes resembling Cooley’s anemia. Acta Haematologica
19:263-268.
Fagundes-Netto, U. 1984. Ultrastructural study of alterations in the
small intestinal epithelium of children with acute diarrhea.
Journal of Pediatric Gastroenterology Nutrition 3:5 10-5 15.
Fairgrieve, S.I. (1990) Size of the sella turcica and its relation to iron
deficiency anemia: a prehistoric example. American Journal
of Physical Anthropology 83:111-121.
Farb, P., and G.J. Armelagos. 1980. Consuming Passions: the
Anthropology of Eating. Houghton Mifflin Co., Boston.
Faulk, W.P. 1976. The immunological system in health and
malnutrition. Proceedings of the Nutrition Society 35:253-261.
Ferris, N. and R.G. Mayer 1 990. Rescue Excavations of the Younge
Phase Cherry Lane (AaHp-21) site, Leamington, Essex County,
Ontario. Report prepared by Mayer, Poulton & Associates, Inc.
134 Commissioners Rd., London, Ontario, NGJ 1X8
Finch, C.A. 1968a. Want amid plenty. American Journal of Clinical
Nutrition 21:1138.
Finch, C.A. 1968b. Iron deficiency in the United States. Journal of
the American Medical Association 203:407-412.
Finch, C.A. 1970. Diagnostic value of different methods to detect
iron deficiency, pp. 409-422. In: Iron Deficiency:
Pathogenesis, Clinical Aspects, Therapy. Edited by Hallberg,
Harwerth, and Vanotti. Academic Press, New York.
Fletcher, J. 1971. The effect of iron and transferrin on the killing of
Escherichia coli in fresh serum. Immunology 20:493-500.
Forman, M.R., B.I. Graubard, H.J. Hoffman. R. Beren, E.E. Harley,
and P. Bennett. 1984. The Pima infant feeding study: breast-
feeding and respiratory infections during the first year of life.
International Journal of Epidemiology 13:447-453.
Foy, H., A. Kondi, and W.H. Austin. 1959. The effect of dietary
phytate on fecal absorption of radioactive ferric chloride.
Nature 183:691-692.
Fullerton, H.W. 1937. The iron deficiency anemia of late infancy.
Archives of Diseases of Childhood 12:91-1 10.
Garibaldi, J.A. 1972. Influence of temperature on the biosynthesis of
iron transport compounds by Salmonella typhirium. Journal
of Bacteriology 1 10:262-265.
Germana, F., and A. Ascenzi. 1980. Porotic hyperostosis in Sardinia.
Journal of Human Evolution 9:75-78.
Ghosh, S., S. Daga, D. Kasthuri, R.C. Misra, and H.K. Chuttania.
1972. Gastrointestinal function in iron deficiency states in
children. American Journal of Diseases of Children 123:14-17.
Goodman, A., D. Martin, and G.J. Armelagos. 1984. Indicators of
stress from bones and teeth, pp. 13-49. In: Paleopathology at
the Origins of Agriculture. Edited by M. Cohen and G.J.
Armelagos. Academic Press, New York.
Gordon, J.E.. I. D. Chitkara, and J. B. Wyon. 1963. Weanling
Diarrhea. American Journal of Medical Science 245:345-377.
Gordon, J.E., J.B. Wyon, and W. Ascoli. 1967. The second year
death rate in less developed countries. American Journal of
Medical Science 254: 1 2 1 - 1 44.
Grantham-McGregor, S.M., P. Desai, and P.F. Milner. 1974.
Hematological levels in Jamaican infants. Archives of
Diseases of Childhood 49:525-530.
Greenberg, G.R., H. Ashenbrucker, M. Lauritsen, M.M. Wintrobe.
1947. The anemia of infection. IV. The lack of relationship
between the diversion of iron from the plasma and the origin of
the anemia. Journal of Clinical Investigation 26: 114-1 20.
Greenfield, G.B. 1975. Radiology of Bone Diseases. (2nd edition), J.
B. Lippincott Co., Philadelphia.
Grieger, T.A., and M.J. Kluger. 1978. Fever and survival: the role of
serum iron. Journal of Physiology (London) 279:187-196.
Gross, R.L. and P.M. Newberne. 1980. Role of nutrition in
immunologic function. Physiology Reviews 60:188-302.
Gryboski, J., and W.A. Walker. 1983. Gastrointestinal Problems in
the Infant. (2nd edition.) W. B. Saunders Co., Philadelphia.
Guha, D.K., B.N.S. Walia, B.N. Tandon, and O.P. Ghai. 1968. Small
bowel changes in iron deficiency anemia of childhood.
Archives of Diseases of Childhood 43:239-244.
Guyton, A.C. 1981. Medical Physiology. W.B. Saunders Co.,
Philadelphia.
Hallberg, L., H. G. Harwerth, and A. Vanotti. 1970. Iron
Deficiency: Pathogenesis, Clinical Aspects, Therapy.
Academic Press, New York.
Hallberg, L., and L. Sovell. 1967. Absorption of hemoglobin iron in
man. Acta Medica Scandinavica 181 :335-354.
Halsted, J.A., A.S. Prasad, and M. Nadimi. 1965. Gastrointestinal
function in iron deficiency anemia. Archives of Internal
Medicine 16:253-256.
Hanson, L.A., and J. Winberg. 1972. Breast milk defense against
infection in the newborn. Archives of Diseases of Childhood
47:454-460.
Hardesty, R.M., and D.J. Weatherall. 1982. Blood and Its
Disorders. Blackwell Scientific Publications, Oxford.
Harrison, M.L. 1978. Taphonomy of the Libben site, Ottawa
County, Ohio. Thesis, Kent State University, Kent, Ohio.
Hawksley, J.C., R. Lightwood, and U.M. Bailey. 1934. Iron
deficiency anemia in children: its association with
gastrointestinal disease, achlorhydria and hemorrhage.
Archives of Diseases of Childhood 9:359-372.
Heath, C.W., and A.J. Patek, 1937. The anemia of iron deficiency.
Medicine (Baltimore) 16:267-350.
1991
Paleoepidemiology of P orotic Hyperostosis
43
Hegsted, D.M., C.A. Finch, and T.D. Kinney. 1949. The influence of
diet on iron absorption. II. The inter-relation of iron and
phosphorous. Journal of Experimental Medicine 90: 147- 1 56.
Heinrich, H.C. 1970. Intestinal iron absorption in man — methods of
measurement, dose relationship, diagnostic and therapeutic
applications, pp. 213-296. In: Iron Deficiency: Pathogenesis,
Clinical Aspects, Therapy. Edited by Hallberg, Harwerth, and
Vanotti. Academic Press, New York.
Hengen, O.P. 1971. Cribra orbitalia: pathogenesis and probable
etiology. Homo 22:57-75.
Henschen, F. 1961. Cribra cranii, a skull condition said to be of
racial or geographical nature. Pathologica et Microbiologia
(Basel) 24:724-729.
Herbert, J.R. 1985. Effects of water quality and water quantity on
nutritional status: findings from a South Indian Community.
Bulletin of the World Health Organization 63:143-153.
Herbert, V. 1987. Recommended dietary intakes (RDI) of iron in
humans. American Journal of Clinical Nutrition 45:679-686.
Hercberg, S., M. Chauliac, M. Devanlay, P. Galan, J.L. Pureur, Y
Soutre, A. Houdegbe, Y. Agboton, I. Zohoun, A. Masse-
Raimbault, and H. Dupin. 1986. Evaluation of the iron status
of a rural population in South Benin. Nutrition Research
6:627-634.
Hillman, R.S. 1970. Factors affecting hemoglobin regeneration, pp.
531-536. In: Iron Deficiency: Pathogenesis, Clinical Aspects,
Therapy. Edited by Hallberg, Harwerth, and Vanotti. Academic
Press, New York.
Hillman, R.S., and P. A. Henderson. 1969. Control of marrow
production by level of iron supply. Journal of Clinical
Investigation 48:454-460.
Hodges, D.C. 1987. Health and agricultural intensification in the
prehistoric valley of Oaxaca, Mexico. American Journal of
Physical Anthropology 73:323-332.
Hoffman-Goetz, L., and M.J. Kluger. 1979. Protein deprivation: its
effects on fever and plasma iron during bacterial infection in
rabbits. Journal of Physiology ( London) 295:419-430.
Hoffman-Goetz, L., D. McFarlane, B. Bistrian, and G.L. Blackburn.
1981. Febrile and plasma iron responses in rabbits injected
with endogenous pyrogen from malnourished patients.
American Journal of Clinical Nutrition 34: 1 109-11 16.
Hooton, E.A. 1930. Indians of Pecos Pueblo. Yale University Press,
New Haven, Connecticut.
Hoshishima, K., M. Ito, and S. Hyodo. 1985. The trace dose of
metal(s) and the immunological reaction in mice. Nutritional
Research (Supplement l):740-744.
Howell, N. 1982. Village composition implied by a
paleodemographic life table. American Journal of Physical
Anthropology 59:263-269.
Hunter, C.A. 1960. Iron deficiency anemia in pregnancy. Surgery,
Gynecology, and Obstetrics 110:210-214.
Hrdlicka, A. 1914. Anthropological work in Peru in 1913, with
notes on the pathology of ancient Peruvians. Smithsonian
Miscellaneous Collections 1914, LXI, (18), 1-69.
Hussain, R., and V.N. Patwardhan. 1959. The influence of phytate
on the absoiption of iron. Indian Journal of Medical Research
47:676-682.
Hwang, Y.F., and E.B. Brown. 1965. Effect of desferrioxamine in
iron absorption. Lancet 1 : 1 35- 1 37.
Jacobs, A., and M. Wormwood. 1982. Iron metabolism, iron
deficiency, and overload, pp. 149-197. In: Blood and Its
Disorders. Edited by R.M. Hardesty and D.J. Weatherall.
Blackwell Scientific Publications, Oxford.
Jacobs, A. and D.H.M. Joynson. 1974. Lymphocyte function and
iron-deficiency anemia. Lancet (2):844.
Jaffe, H.L. 1972. Metabolic, Degenerative, and Inflammatory
Diseases of Bones and Joints. Lea & Febiger, Philadelphia.
Jelliffe, D.B., and V. Blackman. 1962. Bahima disease. Possible
“milk anemia” in late childhood. Journal of Pediatrics
61:774-779.
Jose, D.G., and J.S. Welch. 1970. Growth retardation, anemia, and
infection with malabsorption and infestation of the bowel.
Medical Journal of Australia 1 :349-356.
Josephs, H. 1956. Hypochromic microcytic anemia of infancy:
iron depletion as a factor. Pediatrics 18:959-978.
Josephs, H. 1953. Iron metabolism and the hypochromic anemia
of infancy. Medicine (Baltimore) 32:125-213.
Josephs, H. 1936. Anemia of infancy and early childhood.
Medicine (Baltimore) 15:307-451.
Joynson, D.H., A. Jacobs, D.M. Walker, and A.E. Dolby. 1972.
Defect of cell-mediated immunity in patients with iron-
deficiency anemia. Lancet 2:1058-1059.
Kaatz, M.R. 1955. The Black Swamp: a study in historical
geography. Annals of the Association of American
Geographers XLV(l):l-35.
Kaplan, G.J., J.K. Fleshman, T.R. Bender, C. Baum, and PS.
Clark. 1973. Long-term effects of otitis media. A ten-year
cohort study of Alaskan Eskimo children. Pediatrics
52:577-585.
Kaye, D., and E.W. Hooke. 1963. The influence of hemolysis of
blood loss on susceptibility to infection. Journal of
Immunology 91:65-75.
Keenlyside, D.L. 1978. Late prehistory of Point Pelee, Ontario
and environs. Archaeological Survey of Canada, Paper No.
80, National Museums of Canada, Ontario.
Kenney, M.A. 1985. Factors related to iron nutrition of adolescent
females. Nutrition Research 5:157-166.
Kent, S. 1987. The influence of sedentism and aggregation on
porotic hyperostosis and anaemia: a case study. Man
21:605-636.
Kilpatrick, G.S. 1970. Prevalence of anemia in the United
Kingdom, pp. 441-446. In: Iron Deficiency: Pathogenesis,
Clinical Aspects, Therapy. Edited by Hallberg, Harwerth,
and Vanotti. Academic Press, New York.
Kluger, M.J., and B.A. Rothenburg. 1979. Fever and reduced
iron: an immune response to bacterial infection. Science
203:374-376.
Kochan, I. 1978. The role of iron in bacterial infections, with
special considerations of host-tubercle bacillus interaction.
Current Topics in Microbiology and Immunology 60: 1 -30.
Kochan, I. 1977a. The role of serum iron in the regulation of
nutritional immunity. Advances in Chemistry Series 162:55-77.
44
Mensforth
No. 46
Kochan, I. 1977b. Role of siderophores in nutritional immunity and
bacterial parasitism, pp. 251-288. In: Microorganisms and
Minerals. Edited by E. D. Weinberg. Dekker Press, New York.
Korones, S.B. 1976. High-risk Newborn Infants. (2nd edition), C.V.
Mosby Co., St. Louis, Missouri.
Krakker, J.J. 1984. Investigations of a small late prehistoric site in St.
Clair County, Michigan. The Michigan Archaeologist 30:86-123.
Krakker, J.J. 1983. Changing sociocultural systems during the late
prehistoric period in southeast Michigan. Ph.D. Dissertation,
University of Michigan.
Krantman, H.J., S.R. Young, B. J. Ank, C.M. O’Donnell, G.S.
Rachelefsky, and E.R. Stiehm. 1982. Immune function in pure
iron deficiency anemia. American Journal of Diseases of
Children 136:840-844.
Krogman, W.M., M.Y. and Iscan. 1986. The Human Skeleton in
Forensic Medicine. (2nd edition.) C. C. Thomas, publisher,
Springfield, Illinois.
Kuhn, I.N., M. Layrisse, M. Roche, C. Martinez, and R.B. Walker.
1968. Observations on the mechanisms of iron absorption.
American Journal of Clinical Nutrition 21:11 84- 1 1 88.
Kuhns, W.J., C.J. Gubler, G.E. Cartwright, and M.M. Wintrobe.
1950. The anemia of infection. XIV. Response to massive
doses of intravenously administered saccharated oxide of iron.
Journal of Clinical Investigation 29: 1 505- 1513.
Kuvibidila, S.R., B.S. Baliga, and R.M. Suskind. 1983a. The effect
of iron-deficiency anemia on cytolytic activity of mice spleen
and peritoneal cells against allogenic tumor cells. American
Journal of Clinical Nutrition 38:238-244.
Kuvibidila, S.R., B.S, Baliga, and R.M. Suskind. 1983b. Impairment
of blastogenic response of splenic lymphocytes from iron-
deficient mice: in vivo repletion. American Journal of Clinical
Nutrition 37:15-25.
Kuvibidila, S.R., B.S. Baliga, and R.M. Suskind. 1981. Effects of
iron-deficiency anemia on delayed cutaneous hypersensitivity
in mice. American Journal of Clinical Nutrition 34:2635-2640.
Lahey, M.E. 1962. Protein-losing enteropathy. Pediatric Clinics of
North America 9:689-702.
Lallo, J.W., G.J. Armelagos, and R.P Mensforth. 1977. The role of
diet, disease, and physiology in the origin of porotic
hyperostosis. Human Biology 49:47 1-483.
Lanzkowski, P. 1968. Radiological features of iron deficiency
anemia. American Journal of Diseases of Children 1 16:16-29.
Lanzkowski, P„ and D. McKenzie. 1959. Iron deficiency in Cape
Coloured and African children in Cape Town. South African
Medical Journal 33:22-24.
Laskari, A. 1984. Hematologic Manifestations of Childhood
Diseases. Thieme-Stratton, New York.
Layrisse, M., J.D. Cook, C. Martines-Torres, M. Roche, I.N. Kuhn,
R.B. Walker, and C.A. Finch. 1969. Food iron absorption: a
comparison of vegetable and animal foods. Blood 33:430-443.
Layrisse, M., C. Martinez-Torres, and M. Roche. 1968. Effect of
interaction of various foods on iron absorption. American
Journal of Clinical Nutrition 2 1 : 1 1 75- 1 1 83.
Liebman, M 1985. Iron and folate status of an adolescent female
population. Nutrition Research 5:621-625.
Lennox, P.A. 1982. The Bruner-Colasanti site: an early late
woodland component, Essex County, Ontario. Archaeological
Survey of Canada, Paper No. 110, National Museums of
Canada, Ottawa.
Levine, S.Z., and H. H. Gordon. 1942. Physiological handicaps of
the premature infant. I. Their pathogenesis. II. Clinical
applications. American Journal of Diseases of Children
64:274-296, 297-312.
Lovejoy, C.O., R.S. Meindl, R. P. Mensforth, and T. J. Barton. 1985.
Multifactorial determination of skeletal age at death: a method
and blind tests of its accuracy. American Journal of Physical
Anthropology 68:1-14.
Lovejoy, C.O., R.S. Meindl, T.R. Przybeck, T.J. Barton, K.G.
Heiple, and D. Kotting. 1977. Paleodemography of the Libben
site, Ottawa County, Ohio. Science 198:291-293.
Macdougall, L.G., R. Anderson, G.M. McNab, and J. Katz. 1975.
The immune response in iron deficient children: impaired
cellular defense mechanisms with altered humoral components.
Journal of Pediatrics 86:883-843.
Mackay, H.M.M. 1933. Factors causing variation in hemoglobin
level with age in the first year of life. Archives of Diseases of
Childhood 8:251-263.
Mackay, H.M.M. 1931. Nutritional anemia in infancy. With special
reference to iron deficiency. Medical Research Council Special
Reports, Series No. 157:1-125.
Mahoney, A.W., and D.G. Hendricks. 1975. Utilization of dietary
calcium by iron-deficient rats. Nutrition and Metabolism
18:6-15.
Malina, R.M. 1975. Growth and Development: the First Twenty
Years in Man. Burgess Publishing Co., Minneapolis, Minnesota.
Manchandra, S.S., H. Lai, and S. Khanna. 1969. Iron stores in health
and disease. Bone marrow studies in 1134 children in Punjab,
India. Archives of Diseases of Childhood 44:580-584.
Marquardt, W.H.. and P.J. Watson. 1983. The shell mound archaic of
Western Kentucky, pp. 323-339. In Archaic Hunters and
Gatherers in the American Midwest. Edited by Phillips and
Brown. Academic Press, New York.
Marquardt, W.H. 1972. Recent investigations in a Western Kentucky
shellmound. Paper presented at the Annual Meetings of the
Society for American Archaeology, Miami Beach, Florida.
Martin, C.M., J.H. Jandl, and M. Finland. 1963. Enhancement of
acute bacterial infections in rats and mice by iron and their
inhibition by human transferrin. Journal of Infectious
Diseases 112:158-163.
Martinez-Torres, C., and M. Layrisse. 1974. Interest for the study of
dietary absorption and iron fortification. World Review of
Nutrition and Dietetics 19:51-70.
Martinez-Torres, C., and M. Layrisse. 1971. Iron absorption from veal
muscle. American Journal of Clinical Nutrition 24:53 1-539.
Martinez-Torres, C., and M. Layrisse. 1970. Effect of amino acids on
iron absorption from a staple vegetable food. Blood 35:669-682.
Martorell, R. 1980. Interrelationships between diet, infectious
disease, and nutritional status, pp. 81-106. In: Social and
Biological Predictors of Nutritional Status. Physical Growth,
and Neurological Development. Edited by L. S. Green and F.
E. Johnston. Academic Press, New York.
1991
Paleoepidemiology of Porotic Hyperostosis
45
May, J.A. 1979. Shell midden formation processes: a
methodological perspective. Paper presented at the 36th
Annual Meeting of the Southeastern Archaeological
Conference, Atlanta, Georgia, November, 1979.
Maynard, J.E., and L.M. Hammes. 1970. A study of growth,
morbidity and mortality among Eskimo infants in Western
Alaska. Bulletin of the World Health Organization 42:6 1 3-622.
McFarlane, H. 1976. Malnutrition and impaired immune response to
infection. Proceedings of the Nutrition Society 35:263-282.
McFarlane, H., and J. Hamid, 1973. Cell-mediated immune response
in malnutrition. Clinical Experimental Immunology 13:153-164.
McFarlane, H., S. Reddy, K.J. Adcock, H. Adeshina, A.R. Cooke,
and J. Akene. 1970. Immunity, transferrin, and survival in
Kwashiorkor. British Medical Journal (October):268-270.
McMurray, D.N. 1984. Cell-mediated immunity in nutritional
deficiency. Progress in Food and Nutrition Science 8: 193-228.
Mensforth, R.P 1985. Relative tibia long bone growth in the Fibben
and Bt-5 prehistoric skeletal populations. American Journal of
Physical Anthropology 68:87-106.
Mensforth, R.P. 1986. The pathogenesis of periosteal reactions in
earlier human groups: diagnostic, epidemiological, and
demographic considerations. Paper presented at the 55th
Annual Meeting of the American Association of Physical
Anthropologists, Albuquerque, New Mexico, April 1986.
American Journal of Physical Anthropology 69:239-240
Mensforth, R.P. 1990. Paleodemography of the Carlston Annis (Bt-
5) Fate Archaic skeletal population. American Journal of
Physical Anthropology 82:81-99.
Mensforth, R.P, C.O. Lovejoy, J.W. Fallo, and G.J. Armelagos.
1978. The role of constitutional factors, diet, and infectious
disease in the etiology of porotic hyperostosis and periosteal
reactions in prehistoric infants and children. Medical
Anthropology 2: 1-59.
Moore, S. 1929. The bone changes in sickle cell anemia with a note
on similar changes observed in skulls of Ancient Mayan
Indians. Missouri Medicine 26:561-564.
Moore, C.V., H.Q. Bierman, V. Minnich, and W.R. Arrowsmith.
1940. Studies on iron absorption with special reference to the
relationship of iron metabolism to ascorbic acid. American
Association for the Advancement of Science 1 3:34-37.
Morck, T.A., S.R. Lynch, B.S. Skikne, and J.D. Cook. 1981. Iron
availability from infant food supplements. American Journal
of Clinical Nutrition 34:2630-2634.
Moseley, J.E. 1974. Skeletal changes in the anemias. Seminars in
Roentgenology 9: 1 69- 1 84.
Moseley, J.E. 1966. Radiographic studies in hematologic bone
diseases: implications for paleopathology, pp. 121-130. In:
Human Paleopathology. Edited by S, Jarcho. Yale University
Press, New Haven, Connecticut.
Moseley, J.E. 1965. The paleopathologic riddle of “symmetrical
osteoporosis.” American Journal of Roentgenology 95 : 1 35- 1 42.
Muller, H. 1935. Osteoporosis of the cranium in Javanese.
American Journal of Physical Anthropology 20:493.
Naiman, J.L. , F.A. Oski, L. K. Diamond, G.F. Vawter, and H.
Swachman. 1969. The gastrointestinal effects of iron-
deficiency anemia. Pediatrics 33:83-99.
Nathan, H., and N. Haas. 1966. Cribra orbitalia, a bone condition
of the orbit of unknown nature. Israel Journal of Medical
Science 2:171-191.
Nice, C.M., M.L. Daves, and G.H. Wood. 1964. Changes in bone
associated with cyanotic congenital cardiac disease.
American Heart Journal 68:25-3 1 .
Olivares, M., F. Pizzaro, P. Chadud, and A. Stekel. 1986. Age-
related changes in laboratory measures of iron nutrition in
childhood. Nutrition Research 6:1045-1049.
Palazzari, G.L., K. Matyniak, and L.O. Schulz. 1986. Iron status
independent of socioeconomic variables in school-age
children. Nutrition Research 6:1 131-1 137.
Palkovich, A.M. 1987. Endemic disease patterns in
paleopathology: porotic hyperostosis. American Journal of
Physical Anthropology 74:527-537.
Palkovich, A.M. 1985. Agriculture, marginal environments, and
nutritional stress in the prehistoric southwest, pp. 425-438.
In: Paleopathology at the Origins of Agriculture. Edited by
M. Cohen and G.J. Armelagos, Academic Press, New York.
Parry, M.F., R.K. Root, J.A. Metcalf, K.K. Delaney, L.S. Kaplow,
and W.J. Richard. 1981. Myeloperoxidase deficiency:
prevalence and clinical significance. Annals of Internal
Medicine 95:293-301.
Polk, H.C., and A. A. Miles. 1971. Enhancement of bacteria!
infection by ferric iron: kinetics, mechanisms, and surgical
significance. Surgery 70:7 1 -77 .
Pollack, S.. R.M. Kaufman, and W.H. Crosby. 1964. Iron
absorption: effects of sugars and reducing agents. Blood
24:577-581.
Powell, J.W., H.S. Weens, and N.K. Wegner. 1965. The skull
roentgenogram in iron deficiency anemia and in secondary
polycythemia. American Journal of Roentgenology,
Radiation Therapy, and Nuclear Medicine 95: 143-147.
Prasad, J.S. 1979. Leukocyte function in iron-deficiency anemia.
American Journal of Clinical Nutrition 32:550-552.
Pritchard, J.A., and D.E. Scott. 1970. Iron demands during
pregnancy, pp. 173-212. In: Iron Deficiency: Pathogenesis,
Clinical Aspects, Therapy. Edited by Hallberg, Harwerth,
and Vanotti. Academic Press, New York.
Prufer, O.H. and O.C. Shane, III. 1976. The Portage-Sandusky-
Vermilion River Region in Ohio, pp. 283-304 in The Late
Prehistory of the Lake Erie Drainage Basin: A 1972
Symposium Revised. (Edited by D.S. Brose) Scientific
Papers of The Cleveland Museum of Natural History,
Cleveland, Ohio.
Purtillo, D.T., and D.H. Connor. 1975. Fatal infections in protein-
calorie malnourished children with thymolymphatic atrophy.
Archives of Diseases of Childhood 50: 149- 1 53.
Putschar, W.G.J. 1966. Problems in the paleopathology of bone,
pp. 57-65. In: Human Paleopathology. Edited by S. A.
Jarcho. Yale University Press, New Haven, Connecticut.
Rawson, A., and F.Q. Rosenthal. 1960. The mucosa of the
stomach and small intestine in iron deficiency. Lancet
( 1 ):730-731 .
Robbins, S.L. 1974. Pathologic Basis of Disease. W.B. Saunders
Company. Philadelphia.
46
Mensforth
No. 46
Roche, M., and M.E. Perez-Gimenez. 1959. Intestinal loss and
reabsorption of iron in hookworm infection. Journal of
Laboratory and Clinical Medicine 54:49-52.
Root. P.K., and M.S. Cohen. 1981. The microbial mechanism of
human neutrophils and eosinophils. Review of Infectious
Disease 3:565-598.
Ruff, C.B. 1981. A reassessment of demographic estimates for
Pecos Pueblo. American Journal of Physical Anthropology
54:147-151.
Rybo, G. 1970. Menstrual loss of iron, pp. 163-172. In: Iron
Deficiency: Pathogenesis, Clinical Aspects, Therapy. Edited by
Hallberg, Harwerth, and Vanotti. Academic Press, New York.
Saddi, R., and G. Schapira. 1970. Iron requirements during
growth, pp. 183-198. In: Iron Deficiency: Pathogenesis,
Clinical Aspects, Therapy. Edited by Hallberg, Harwerth,
and Vanotti. Academic Press, New York.
Santos, J.I. 1986. Nutritional implications and physiologic
response to pediatric diarrhea. Pediatric Infectious Disease 5
(supplement): 1 52- 1 54.
Sax, B. 1963. Roentgen manifestations of iron deficiency anemia
in the skull of infants and children simulating those seen in
Cooley’s and sickle-cell hemolytic anemia. Germantown
Hospital, Philadelphia Journal 4:72-75.
Schade, A.L. 1963. Significance of serum iron for the growth,
biological characteristics, and metabolism of Staphylococcus
aureus. Biochemische Zeitschrift 338:140-148.
Schutte, D. 1979. Iron deficiency anemia as the cause of porotic
hyperostosis among the PreColumbian Maya. Thesis,
California State University, Fullerton.
Scott, D.E., J.A. Pritchard, A.E. Saltin, and J.M. Humphreys.
1970. Iron deficiency during pregnancy, pp. 491-504. In:
Iron Deficiency: Pathogenesis, Clinical Aspects, Therapy.
Edited by Hallberg, Harwerth, and Vanotti. Academic Press,
New York.
Scrimshaw, N.S. 1964. Ecological factors in nutritional disease.
American Journal of Clinical Nutrition 1-: 1 12-122.
Scrimshaw, N.S., and R.M. Suskind. 1959. Interactions of
nutrition and infection. American Journal of Medical
Science 237:367-403.
Scrimshaw, N.S., and V.R. Young. 1976. The requirements of
human nutrition. Scientific American 235 (September):51-64.
Shah, A.G., and S. Seshadri. 1985. Anemia in relation to dietary
iron deficiency, iron availability and parasitic infestation.
Nutrition Research 5:341-347.
Shahidi, N.T., and L.K. Diamond. 1960. Skull changes in infants
with chronic iron deficiency anemia. New England Journal
of Medicine 262:137-139.
Sharpe, L.M., W.C. Peacock, R. Cooke, and R.S. Harris. 1950.
The effect of phytate and other food factors on iron
absorption. Journal of Nutrition 41 :433-446.
Shaw, R., and W.O. Robertson. 1964. Anemia among hospitalized
infants. Ohio State Medical Journal 60:45-47.
Shearman, D.J. C., I.W. Delamore, and D.L. Gardner. 1966.
Gastric function and structure in iron deficiency. Lancet
( 1 ):845-848.
Sheldon, W. 1936. Anemia with bone changes in the skull. Royal
Society of Medicine, London Proceedings 29:743.
Smith, C.H. 1972. Blood Diseases of Infancy and Childhood. (3rd
edition). C.V. Mosby Company, St. Louis, Missouri.
Sodeman, W.A., and W.A. Sodeman. 1974. Pathologic
Physiology: Mechanisms of Disease. W. B. Saunders
Company, Philadelphia.
Srikantia, S.G., J.S. Prasad, C. Bhaskaram, and K.
Krishnamachari. 1976. Anemia and immune response.
Lancet ( 1 ): 1 307-1 309.
Stewart, A. 1937. Gastric acidity in infants and young children
under normal and pathological conditions, with special
reference to nutritional anemia. British Journal of
Children 's Diseases 34: 1 -27.
Stuart-Macadam, P. 1985. Porotic hyperostosis: representative of
a childhood condition. American Journal of Physical
Anthropology 66:391-398.
Stuart-Macadam, P. 1987a. A radiographic study of porotic
hyperostosis. American Journal of Physical Anthropology
74:511-520.
Stuart-Macadam, P. 1987b. Porotic hyperostosis: new evidence to
support the anemia theory. American Journal of Physical
Anthropology 74:521-526.
Stuart-Macadam, P. 1989. Porotic hyperostosis: relationship
between orbital and vault lesions. American Journal of
Physical Anthropology 80: 1 87-193.
Sturgeon, P. 1956. Iron metabolism: a review with special
consideration of iron requirements during normal infancy.
American Academy of Pediatrics 18:267-300.
Thomas, W.J., H.M. Koenig, A.L. Lightsey, and R. Green. 1977.
Free erythrocyte protoporphyrin: hemoglobin ratios, serum
ferritin and transferrin saturation levels during treatment of
infants with iron deficiency anemia. Blood 49:455-462.
Till, J.E., and E.A. McCulloch. 1961. A direct measurement of
the radiation sensitivity of normal mouse bone marrow cells.
Radiation Research 14:213.
Trentin, J.J. 1971. Determination of bone marrow stem cell
differentiation by stromal hemopoietic inductive
microenvironment. American Journal of Pathology 65:621.
Trucco, J.I., and A.K. Brown. 1967. Neonatal manifestation of
hereditary spherocytosis. American Journal of Diseases of
Childhood 1 13:263-270.
Turnbull, A.F. Cleton, C.A. Finch, L. Thompson, and J. Martin.
1967. Iron Absorption. IV. The absorption of hemoglobin
iron. Journal of Clinical Investigation 41:1897-1907.
Vannotti, A. 1957. The role of the reticuloendothelial system in
iron metabolism, pp. 172-186. In: Physiology of the
Reticuloendothelial System. Blackwell Scientific
Publications, Oxford.
Venkatachalam, P.S. 1968. Iron metabolism and iron deficiency in
India. American Journal of Clinical Nutrition 21:1 156-1 161.
Von Endt, D.W., and D.J. Ortner. 1982. Amino acid analysis of
bone from a possible case of prehistoric iron deficiency
anemia from the American southwest. American Journal of
Physical Anthropology 59:377-385.
1991
Paleoepidemiology of P orotic Hyperostosis
47
Wakefield, E.G., S.C. Dellinger, and J.D. Camp. 1937. A study of
the osseous remains of the “mound builders” of Eastern
Arkansas. Journal of American Medical Science 193:489-495.
Walker, P.L. 1986. Porotic hyperostosis in a marine-dependent
California Indian population. American Journal of Physical
Anthropology 69:345-354.
Walter, T., S. Arrendondo, M. Arevalo, and A. Stekel. 1986.
Effect of iron therapy on phagocytosis and bactericidal
activity in neutrophils of iron-deficient infants. American
Journal of Clinical Nutrition 44:877-882.
Webb, W.S. 1950. The Carlston Annis mound site 5 Butler
County, Kentucky. University of Kentucky Reports in
Anthropology, volume 7, (number 4), Lexington, Kentucky.
Weinberg, E.D. 1966. Roles of metallic ions in host-parasite
interactions. Bacteriological Reviews 30: 1 36- 151.
Weinberg, E.D. 1977. Microorganisms and Minerals. Edited by
E.D. Weinberg. Marcel Dekker, Inc., New York.
Weinberg, E. D. 1974. Iron and susceptibility to infectious
disease. Science 184:952-956.
Weiss, K.M. 1972. On the systematic bias in skeletal sexing.
American Journal of Physical Anthropology 37:239-250.
Weiss, K.M. 1973. Demographic models for anthropology.
Memoirs of the Society for American Archaeology (number
27), American Antiquity 38, (number 2), part II.
Welcker, H. 1888. Ci'ibra orbitalia. Archive fur Anthropologie
17:1-18.
Wen-guang, W.C. Xue-cun, and L. Dong-sheng. 1986.
Hematologic response to iron and ascorbic acid
administration in preschool children with anemia. Nutrition
Research 6:241-248.
Werkman, S.L., L. Shifman, and T. Skelly. 1964. Psychosocial
correlates of iron deficiency anemia in early childhood.
Psychosomatic Medicine 26: 1 25- 1 34.
Williams, A.O., S.B. Lagundoye, and C. L. Johnson. 1975.
Lamellation of the diploe in the skulls of patients with sickle
cell anemia. Archives of Diseases of Childhood 60:948-952.
Williams, H.U. 1929. Human paleopathology: with some original
observations on symmetrical osteoporosis of the skull.
Archives of Pathology 7:839-902.
Wilson, J.F., D.C. Heiner, and M.E. Lahey. 1962. Studies on iron
metabolism. I. Evidence of gastrointestinal dysfunction in
infants with iron deficiency anemia: a preliminary report.
Journal of Pediatrics 60:787-800.
Winters, R. H. 1974. Introduction to the new edition of Indian
Knoll, by W.S. Webb, University of Tennessee Press,
Knoxville.
Witts, L.J. 1966. Anemia as a world health problem. International
Society of Hematology, 11th Congress of the University of
Sidney, Plenary Sessions, 85-102.
Wood-Jones, F. 1910. General pathology (including diseases of
the teeth). In: Report on Human Remains. Edited by G. E.
Smith and F. Wood-Jones. Archaeological Survey of Nubia
report for 1907-1908, volume 2, Cairo.
Woodruff, C.W. 1958. Multiple causes of iron deficiency anemia
in infants. Journal of the American Medical Association
167:715-720.
Woodruff, C.W., and E.P. Bridgeforth. 1953. Relationship
between the hemogram of the infant and that of the mother
during pregnancy. Pediatrics 12:681.
World Health Organization. 1968. Nutritional anemias: report
of a WHO scientific group. WHO Technical Report Series
No. 405.
Wretlind, A. 1970. Food iron supply, pp. 39-70. In: Iron
Deficiency: Pathogenesis, Clinical Aspects, Therapy.
Edited by Hallberg, Harwerth, and Vanotti. Academic
Press, New York.
Yetkin, S., S. Altay, G. Ciliv, and Y. Lalely. 1979. Myelo-
peroxidase activity and bacterial function of PMN in iron
deficiency. Acta Haematologica 61:10-14.
Yusufji, D., V.I. Mathan, and S.J. Baker. 1973. Iron, folate, and
vitamin B12 nutrition in pregnancy: a study of 1000
women from Southern India. World Health Organization
Bulletin 48:15-22.
Zaino, D.E., and E.C. Zaino. 1969. Cribra orbitalia in the
aborigines of Hawaii and Australia. American Journal of
Physical Anthropology 42:91-94.
Zaino, E.C. 1964. Paleontologic thalassemia. Annals of the New
York Academy of Sciences 1 1 9:402-4 1 2.
Zaino, E.C. 1967. Symmetrical osteoporosis, a sign of severe
anemia in prehistoric Pueblo Indians of the Southwest, pp.
40-77. Miscellaneous Papers in Paleopathology /, Museum
of Northern Arizona, Flagstaff.
Zaino, E.C. 1968. Elemental bone iron in the Anasazi Indian.
American Journal of Physical Anthropology 29:433-436.
KIRTLANDIA
The Cleveland Museum of Natural History
August 1991 Number 46:49-7 1
North American
Late Devonian
Cephalopod Aptychi
Calvin J. Frye and Rodney M. Feldmann
Department of Geology
Kent State University
Kent, Ohio 44242
Abstract
Enigmatic, Hat body fossils have been collected from at least 9 localities in dark shales of Late Devonian
age in northeastern Ohio. The fossils have been found mainly in the Cleveland Shale, a black shale interpreted
to represent an anoxic environment, and more rarely in the Chagrin Shale, which was deposited in a
dysaerobic environment. The benthic fauna of these shales is sparse and restricted. These Ohio fossils are
comparable to similar structures found within the Woodford Shale of Oklahoma, a formation of equivalent age
and depositional environment as the Chagrin and Cleveland shales.
The enigmatic fossil remains comprise at least seven authentic species referrable to the genus Sidetes
Giebel, 1849. Five of these species are found in Ohio. Spathiocaris tenuicosta Cooper, 1932 is
morphologically indistinguishable from Sidetes chagrinensis (Ruedemann, 1916), and is, therefore, placed in
synonymy. Similarly, Spathiocaris striatula Cooper, 1932 is the junior subjective synonym of Sidetes lata
(Ruedemann, 1916) and Spathiocaris woodfordi Cooper, 1932 and Spathiocaris plicifera Cooper, 1932 are
junior subjective synonyms of Sidetes newberryi (Whitfield, 1882).
All of the fossils are extremely thin and, typically, flat structures marked with fine, nearly concentric,
corrugations or folds and range from 0.7 cm to 8 cm in length. At various times they have been considered to be
brachiopods, barnacle plates, cephalopod aptychi, or the phyllopod crustaceans Spathiocaris or Aptychopsis.
Scanning electron microscopy reveals no ultrastructure within the fossils. Electron-dispersive X-ray
spectroscopy indicates they contain neither calcium, strontium, nor phosphorus. Brachiopods and arthropods
from the same units do contain phosphorous. Their general morphology and ornamentation is also unlike that
of brachiopods or arthropods, permitting their assignment to the Cephalopoda. They appear to be the aptychi
of ammonites, structures which probably served as the animal’s lower jaw. Their probable preservation as
carbon films remnant of degraded organic material is consistent with what is known of cephalopod aptychi.
The reconstruction of two specimens that had been cracked and Battened during compaction shows the
original form of the structures to have been broadly curved and scooplike. This is consistent with reconstructions
of Mesozoic ammonite jaws, and strengthens the assignment of these fossils with the cephalopods.
50
Frye and Feldmann
No. 46
Introduction
In recent years there have been widely scattered reports
of fossils believed to be cephalopod aptychi from the
Paleozoic of North America (Brady, 1955, but see
Yochelson, 1971; Closs, et al., 1964; Saunders and Spinosa,
1974; Saunders and Richardson, 1979; Thompson, et al.,
1980; Yochelson, 1983; Kues, 1983; Mapes, 1987;
Landman and Davis, 1988; Harper, 1989). Although several
of these reports claim to be one of only a handful of such
findings, many more specimens are available for study. The
collections of The Cleveland Museum of Natural History
contain over a hundred such specimens from the Late
Devonian Chagrin and Cleveland shales.
These fossils are preserved as flat, glossy black,
carbonaceous structures in the dark shales. They are
marked with fine concentric ridges which parallel the
margin or are truncated by it along the periphery. The
ridges resemble those sometimes seen in brachiopods or
bivalves, and the general outline could be suggestive of
some early crustaceans. Similar fossils also have been
identified as fish scales, barnacle plates, or perhaps
gastropod opercula.
In the earliest description of these Devonian fossils in
North America they were identified as a form of phyllocarid
crustacean, Spathiocaris (Clarke, 1882). Woodward (1885),
however, acknowledged that some “phyllocarids” could, in
fact, be goniatite aptychi. Later, Clarke (1902) expressed
doubts, admitting that they could be cephalopod aptychi or
brachiopod fragments instead. This study was begun in an
attempt to determine their affinities.
It was expected that the application of new methods
might yield additional insight into the subject. Examination
of the microstructure of these structures using the scanning
electron microscope might reveal details of their formation
and growth which would permit a more definitive
identification. Determination of the chemical composition
of these structures by x-ray spectroscopy might also
confirm their affinities. Cephalopod shell material is
aragonitic, whereas their mandibles are calcitic
(Lowenstam, et al., 1984). Aragonite frequently contains
strontium as a significant trace element. Inarticulate
brachiopod shells and arthropod carapaces are composed
primarily of calcium phosphate, not calcium carbonate.
New chemical data would not necessarily be definitive,
however, for bivalves are predominately aragonitic.
The principal purposes of this study are to examine and
describe the “spathiocarids” of the Cleveland and Chagrin
shales, investigate their relationships with similar
Devonian taxa, and attempt to provide solid identification
of their nature, if possible. This latter goal was not fully
realized, but two of the most likely alternatives have been
eliminated. It is probable that these fossils should be
referred to the Cephalopoda.
Morphological Terminology
The terms describing the various forms of these fossil
structures are complicated and somewhat confused. Many
of the terms, originally defined as morphological features,
have subsequently been adopted as taxonomic names.
Further, a number of names of taxa have since been
considered as morphological terms. It seems clearest to use
the simple set of terms proposed by Moore and Sylvester-
Bradley (1957a) in the Treatise on Invertebrate
Paleontology. Aptychus (plural; aptychi) is considered to
be a general name for this group of fossil cephalopod
fragments, although it has a restricted meaning as the group
of bivalved forms. Anaptychus is used to refer to the
univalved structures found alone or in association with the
pair of aptychi sensu stricto in younger rocks. For general
discussion, these terms are useful:
"aptychus [broad sense | — All types of calcareous
or corneous structures presumed to serve as
opercula [or mandibles] of ammonoid conchs.
diaptychus [= aptychus sensu stricto] — Aptychus
composed of two discrete valves,
anaptychus — Univalved type of aptychus.” (Terms
modified from Moore and Sylvester-Bradley,
1957a.)
With the multiplicity of interpretations, published
descriptions of aptychi and similar specimens are hard to
reconcile. The life orientation of these fossil fragments is
necessarily different if they are arthropod carapaces
(Clarke, 1882; Ruedemann, 1916), brachiopods (Clarke,
1902), cephalopod mandibles (Lehmann, 1970), or
opercula (Trauth, 1927;Turek, 1978).
The orientation of these structures would be the same
whether they were interpreted as crustacean carapaces or
cephalopod mandibles. In each case, the rostrum or apex
(center of the concentric ornamentation) is anterior; the
hinged region is medial. The morphological terms used
here (Figure 1) are based on those defined by Clarke
(1962) for the description of coleoid cephalopod
mandibles. The anterior angle is here defined as the angle
of the anterior margin of the flattened fossil; convex if the
rostrum is emergent, concave if it is reentrant.
Should it prove correct that aptychopsid plates were
nautiloid opercula (Holland, et al., 1978; Turek, 1978)
whereas aptychi proper served as ammonite mandibles
(Lehmann, 1970), then new terminology would have to be
created for the former. Not only would the function be
different, but the two structures, otherwise similar in
appearance, would have had opposite orientations in the
living animals.
Previous Work
In 1847, C.G. Giebel reported the finding of some
enigmatic fossil molds in the Cretaceous sandstones around
1991
Late Devonian Cephalopod Aptychi
51
Anterior
Posterior
2
Interior
FIGURE 1. Stylized diagrams of an early cephalopod lower
mandible, showing reference attitude and morphological terms.
Terminology after Clarke (1962). 1, Ventral view, extended and
flattened. 2, Exterior lateral ( oblique ) view.
Salzburg. He described the genus Sidetes Giebel, 1849, two
years later, concluding that these structures were the
aptychi of Sepia Linnaeus, 1758 (Giebel, 1849). The
specimens he observed were semicircular, and marked with
fine concentric lines.
Later, John M. Clarke (1882) described several odd
fossils from Naples, New York, found in Givetian and
Frasnian age shales of the Hamilton and Genesee
formations. He collected thirty specimens over the course
of several years, but remained doubtful as to their
biological affinities. The fossils ranged in size from a few
millimeters in breadth to as much as 90 mm, were flattened
elliptical bodies marked with concentric lines or ridges,
and bore a wedge-shaped cleft. He concluded they were not
brachiopods as they were too large and did not display any
trace of a corresponding ventral valve.
Clarke's descriptions were based on what he thought
were incomplete specimens, consisting of isolated
carapaces. He noted their general similarity to the Silurian
arthropod genus Discinocaris Woodford 1856, and
suggested these new specimens belonged properly with the
phyllopod crustaceans, and that the discovery of a
complete specimen would come with time. He assigned
them to two crustacean genera, Spathiocaris Clarke 1882,
and Lisgocaris Clarke 1882.
At the same time, Whitfield (1882) found similar
fossils, which he referred to the crustacean genus
Plumulites Barrande 1872. These specimens were all
recovered from the Cleveland Shale in Erie County, Ohio.
Over the course of several decades of work with the
British Museum, Henry Woodward had occasion to describe
many small fossils of similar type. He proposed (Woodward,
1865; and see Woodward, 1885a) that some of these “ink-
flecks” were chiton plates, while others were isolated
barnacle plates, for which he proposed the genus Turrilepas
Woodward, 1865. He subsequently (Woodward, 1882)
referred similar specimens from the Devonian of Biidesheim,
Germany to the new phyllopod genera Cardiocaris
Woodward, 1882, and Pholadocaris Woodward, 1882, and a
specimen from the Silurian of Wales to the phyllopod genus
Aptychopsis Barrande, 1872. Later, he (Woodward, 1885)
agreed that some such fossils may be cephalopod aptychi, but
felt that others were certainly phyllopod carapaces.
Ruedemann’s discovery, in 1901, of a very large
brachiopod in the “Hudson river shales” (sic] of New York
prompted Clarke to summarize 25 years of collecting
Spathiocaris and similar fossils (Clarke, 1902). He
observed that none had yet been discovered with
abdominal fragments, and a specimen of Spathiocaris had
been found in the body chamber of the Devonian goniatite
Manticoceras intumescens in Germany (Kayser, 1882). He
had earlier illustrated a similar occurrence of a
“phyllocarid” (Dipterocaris Clarke, 1883) within a
goniatite from the Naples (Portage) shales of New York
(Hall, 1888, pi. 35). His conclusions at this time were that
Spathiocaris and similar forms were probably cephalopod
aptychi, but that Discinocaris was perhaps a brachiopod.
He gave this interpretation for the latter as its occurrence in
the Silurian preceeded the appearance of the ammonites.
At about the same time, other dark shales were being
studied. Girty’s monograph (1909) on the Caney Shale
(Devonian-Mississippian) of Oklahoma included a new
genus, ldiotheca, which Girty hesitantly described as a
conulariid. He briefly stated other possibilities, including
its interpretation as a cephalopod aptychus. He ruled out
the possibility of its being an inarticulate brachiopod.
Later, Ruedemann (1916) described four new species of
Spathiocaris from New York and northeastern Ohio. He
suggested several reasons to consider these fossils as
belonging to the Cephalopoda, observing that the method
of growth seen in Spathiocaris and related genera is
similar to that seen in aptychi and not like that of
brachiopods or arthropods. He supposed that the horny
anaptychus would logically preceed calcareous diaptychi
52
Frye and Feldmann
No. 46
in the evolution of these structures, so the presumed lack
of diaptychi from the Paleozoic does not mean that
spathiocarids could not have been anaptychi. Finally, he
suggested that similar structures would, “also have existed
in the Ordovician and Silurian cephalopods and thus
account for those earlier anaptychi considered as
Discinocarina [sic]” (Ruedemann, 1916, p. 102).
Six new species of Spathiocaris were described by C.L.
Cooper (1932) from the Woodford Formation of
Oklahoma, including the redescription of Girty’s Idiotheca
specimen as the new species Spathiocaris woodfordi. The
Woodford is an interbedded black shale/chert unit of Late
Devonian to Early Mississippian age. Many of Cooper’s
descriptions are similar to species from the east. He
described the fauna strictly as crustacean.
Ruedemann (1934) expanded upon the idea that
Spathiocaris was a cephalopod aptychus, citing as evidence
Matern’s (1931) discovery of Spathiocaris koeneni Clarke,
1884 within the body chamber of Crickites holzapfeli
Wedekind, 1913, a European Devonian goniatite. Ruedemann
said it was unlikely that this represents the preservation of a
phyllocarid preying upon a goniatite. The suggestion was
made by Matern, and echoed by Ruedemann, that the
anaptychi were separated from the conchs as the cephalopods
decomposed while still buoyed by gases in the conch.
In his description of the New Albany Shale of Indiana,
Campbell (1946) mentioned several thin beds within this
Devonian black shale as “ Spathiocaris beds.” He
suggested these horizons, where these fossils were locally
abundant, might be useful in stratigraphic correlation.
Unfortunately, such occurrences are too rare to be helpful
(Lineback. 1970; Hasenmueller and Leininger, 1987).
Spathiocaris has also been identified in drill cores from
western Canada (Copeland and Boulton, 1960), along with
a phyllopod (crustacean) telson. This last has been removed
from association with Spathiocaris and redescribed as
Montecaris (Pratt, 1987).
Materials
Specimens in this study were collected by many different
individuals from 1925 to 1989. Most were collected by RA.
Bungart or F. Thompson incidental to collection of Cleveland
Shale fish material for The Cleveland Museum of Natural
History (CMNH). All specimens studied were borrowed from
The Cleveland Museum of Natural History, except for three
specimens of Aptychopsis Barrande, 1872 which were kindly
loaned by the Palaeontological Institute of Lund, Sweden
(LO), Whitfield’s type specimens provided by the American
Museum of Natural History (AMNH), and Cooper’s type
specimens, which were borrowed from the National Museum
of Natural History (USNM).
All the fossils studied from Ohio were preserved as
carbonized films, flattened and compacted into the shale.
Rarely was there a good interface between the specimen
and the matrix, so preparation was held to a minimum. In a
few cases folded specimens were separated from the rock
along their outer surfaces. Most were prepared for
photographic illustration by coating with finely particulate
ammonium chloride. A few fragments were removed and
coated with a thin film of gold for examination with the
scanning electron microscope, but uncoated specimens
were also examined by this technique with good results,
probably due to their carbon content.
The USNM specimens, from the Woodford Shale of
Oklahoma, are preserved as three-dimensional ellipsoidal
packages of thin, sheetlike fossil material within
phosphatic concretions from the shale matrix.
Stratigraphy and Localities
The fossils in this study have come from two units within
the Late Devonian of northeastern Ohio; the dark colored
Chagrin Shale and the overlying black Cleveland Shale
(Fanrmenian). These shales are exposed along the southern
shore of Lake Erie for 150 kilometers and extensively along
many of the streams draining into the lake. The Devonian
shale outcrop belt in northeast Ohio is illustrated in Figure 2,
along with the known sites from which “spathiocarids” have
been collected. These shale units represent prograding distal
deposition of fine-grained sediments from the Catskill Delta
to the east during a time of marine transgression in the
Appalachian Basin (Lewis, 1988).
Prosser first used the name “Chagrin Shale” to describe
the unit of interbedded gray shales and siltstones which
Newberry called the “Erie Shale,” as the latter term was
preoccupied (Prosser, 1912, pp. 14-15). The unit is a wedge-
shaped body thickening eastward into western Pennsylvania,
where it is correlative with the Riceville Shale. It is
underlain by the Huron Shale, another black shale. The
Chagrin thins westward and pinches out east of the Huron
River (Lewis, 1988). The Chagrin Shale consists primarily
of greenish-gray or bluish-gray clayey shales interbedded
with discontinuous siltstones. The shales frequently are
bioturbated and often contain other traces of benthic life
(Barron and Ettensohn, 1981; Hannibal and Feldmann,
1983; Schwimmer, 1988; Schwimmer and Feldmann, 1990).
They represent gradual deposition within a dysaerobic low-
energy environment (Barron and Ettensohn, 1981). The
siltstones probably represent episodic storm events, washing
coarser deltaic sediments westward in the basin (Hannibal
and Feldmann, 1983).
The Cleveland Shale, named by Newberry in 1870, thins
both eastward and westward from its maximum thickness
west of Cleveland (Lewis, 1988). It consists primarily of
black, laminated, fissile shales containing more organic
matter and quartz and less clay (illite) than the gray shales
of the Chagrin (Broadhead, et al., 1982). The lack of an
1991
Late Devonian Cephalopod Aptychi
53
FIGURE 2. Outcrop map of Devonian shales in northeastern
Ohio. Numbered localities refer to sites where “ spathiocarids ”
have been collected.
active benthos and the enrichment in sulfides and organics
indicates that the Cleveland Shale was deposited in deeper,
anoxic conditions (Barron and Ettensohn, 1981). While it
lies over the Chagrin Shale, there is a general east-west
transition between the two units as the gray Chagrin grades
westward into the black Cleveland Shale (Prosser, 1912;
Szmuc, 1970a). The Cleveland Shale is overlain in turn by
the Bedford Formation. The contact is sharp but
conformable, marked by a thin bed of pyritized brachiopods
and vertebrate fragments (Szmuc, 1970b).
Localities
The known localities at which these fossils have been
found are listed here, in order from west to east. The
numbers correspond to those in Figure 2.
1) Chance Creek — An easterly tributary of the Vermilion
River in Forain County. Kipton 7.5 Minute Quadrangle,
Brownhelm Township, T6N, R19W, 41°21'40''N,
82° 18 '00 "W. Exposures occur on Chance Creek, 400 m
south of the intersection of Vermilion Road and Gifford
Road. The Cleveland Shale in this region is approximately
15 m thick (Fewis, 1988). Specimens CMNH 3744, 3746,
and 3747 were collected by William J. Hlavin from a zone
0.5 to 1.5 m below the Cleveland-Bedford formational
contact. “The base of the invertebrate zone is characterized
by a thin bone bed which contains water-worn, disarticulated
elements and teeth of fossil fish” (Hlavin, 1976).
2) Cahoon Cliffs — Cliffs along the Fake Erie shore in Bay
Village, Ohio. North Olmstead 7.5 Minute Quadrangle, T7N,
R15W, 41°29'15"N, 81°55'30"W. The 10 m cliff east of the
mouth of Cahoon Creek is an excellent exposure of the
Chagrin Shale (Prosser, 1912). CMNH 6620 was collected
as float along the beach, 100 meters east of the creek mouth.
3) Little Cedar Point — A bluff at the confluence of the East
and West branches of the Rocky River. North Olmstead 7.5
Minute Quadrangle, T6N, R15W, 41°24'40"N, 81°53’20"W.
The Cleveland Shale in this area forms steep, high cliffs
along the river. The thickness of the unit appears to be 30 m,
with the upper third somewhat more resistant than the rest
(Prosser, 1912). Numerous concretions, cone-in-cone
structures, and pyrite nodules have been found in the
Cleveland Shale in this area. Forty-four specimens were
located in the collections of The Cleveland Museum of
Natural History in association with labels which read, “100
yds. E of ford at base of Fittle Cedar Point, 6/24/51” and
“5/29/51, in a landslide.” Although it is not clear which
specimens, if any, are rightfully associated with these labels,
the locality has been productive of both vertebrate and
invertebrate material.
4) Abram’s Creek — A southerly tributary of the Rocky
River. Fakewood 7.5 Minute Quadrangle, T6N, R14W,
41 °25 '05 "N, 8 1 °52’00"W. CMNH 8312 was collected
where Abram’s Creek meets Rocky River, not far
downstream from Fittle Cedar Point.
5) Big Creek Localities — Along Big Creek from
Brookside Park to the Big Creek Metropark and beyond,
the upper 30 m of the Chagrin and at least 15 m of the
Cleveland are exposed. The Cleveland appears at the top of
the cliffs on the south side of the creek just above
Brookside Park; about three km upstream it reaches the
stream bed south of West Park Cemetery, near the western
edge of the Cleveland South Quadrangle (Prosser, 1912).
Focalities in the upper reaches of Big Creek include four
along the northwest branch of the creek, and a region west
of Finndale and north of Memphis Road. The fossils were
collected as float and occasionally in situ in the creek bed.
Fakewood 7.5 Minute Quadrangle:
5a) 30 m east of W. 140th St., T7N, R14W,
41°26'15"N, 81°47'30''W. CMNH 8317.
5b) First bend below W. 130th St., T7N, R13W,
41°26'15”N, 81°46'45"W. CMNH 8304.
5c) Above W. 128th St., T7N, R13W, 41°27'00"N,
81°46'45"W. CMNH 8303, CMNH 8315, CMNH
8316, CMNH 8318.
5d) At W. I 17th St. T7N, R13W, 41°27T5"N,
8 1 °46 '00 " W. CMNH 3745.
5e) Region of the Metropark north of Memphis Road,
T7N, R13W, 4 1 °23 ’45 "N to 41°26'30"N,
54
Frye and Feldmann
No. 46
8 1 °45 '15 "W to 8 1 °45 '30 "W. CMNH 7942,
CMNH 8159, CMNH 8306, CMNH 8309,
CMNH 8310, CMNH 831 1, CMNH 8314.
6) Brookside Park — Located along Big Creek between the
Big Creek Metropark and the creek’s terminus at the
Cuyahoga River. Trace fossils have been collected as float
along the base of outcrops south of Big Creek in this area
(Hannibal and Feldmann, 1983). Cleveland South 7.5
Minute Quadrangle, T7N, R13W, 41°26'30''N to
41°27'30"N, 8 1 °42'15 ”W to 81°44'00"W. CMNH 8305,
CMNH 8307, CMNH 8313.
7) Skinner's Run — On the border of Brooklyn Heights,
Parma, and Seven Hills, Ohio. The Chagrin and the
Cleveland are exposed along this tributary of the Cuyahoga
River. At their contact is the Skinner’s Run pyrite bed
(Hlavin, 1976), a pyritized lag deposit enriched in fossils.
Specimens were collected from the lower portion of the
Cleveland Shale, somewhat above the pyrite bed. South
central */9th of the Cleveland South 7.5 Minute Quadrangle,
T6N, R12W, 41°24'30"N, 81°40'30"W.
8) Euclid Creek - — East of Cleveland, in Euclid. The
Chagrin Shale is exposed along Euclid Creek’s west branch
in the Euclid Creek Metropark. East Cleveland 7.5 Minute
Quadrangle, T8N, R11W, 41°33'00''N, 81°31'45"W.
CMNH 6576 was collected in situ north of a small gully
extending from the park road, 1.8 km (1.1 mi) south along
the park road from its intersection with Highland Road.
9) Mill Creek — Camp Stigwandish, Lake County, Ohio.
The Chagrin Shale is exposed in the cliffs along this
tributary of the Grand River. Thompson 7.5 Minute
Quadrangle, T11N, R6W, 41°44'15"N, 81°02'00”W. CMNH
7948, collected as float from a steep cliff on the west side of
the stream, 87 m upstream from the Doty Road bridge.
Systematic Paleontology
General
Aptychi present a particular problem to systematists.
They are distinctive enough to be useful in stratigraphy,
particularly where they are locally abundant enough to
constitute “aptychus beds” (Campbell, 1946; Trauth, 1930).
It is useful to be able to distinguish the forms by name, and
indeed, many names were applied to these fossils by earlier
workers before their nature was understood. However, the
variability resulting from preservation and compaction has
led to the erection of more species than was perhaps
warranted (Turek, 1978).
As cephalopod conchs were discovered with associated
aptychi, taxonomic conflicts arose; aptychus names often
had priority over those well-known for the conch. Article
23 of the Code of Zoological Nomenclature (International
Commission on Zoological Nomenclature, 1985) would, in
strict interpretation, have the earlier established name
become that of the taxon, although section (b) of that
article indicates that if this were to cause instability or
confusion, an author can refer the case to the Commission
for an individual ruling. In either case, one of the two
names would have to be suppressed.
There is a complex heirarchy of ammonite taxonomy,
based entirely on characters exhibited by the conchs.
Aptychi do not possess sufficient morphological variation
to permit diagnosis at the generic or specific levels. Thus in
some cases, single “species” of aptychi have been found to
belong to two or more genera of ammonites as
distinguished by conchs.
A simple solution to this problem was proposed by
Arkell (1954; 1957a). He favored the suppression of all
names based solely on the aptychus of an ammonite. This
proposal does assure the stability of ammonite nomen-
clature, at the expense of abandoning names useful for
identifying aptychi as discrete entities separate from the
remainder of the animal, as is often the case.
A more sweeping suggestion was made by Moore and
Sylvester-Bradley (1957b) that a separate, parallel system of
nomenclature be established for “parataxa;” names based on
aptychi, individual conodonts, and isolated holothurian
elements. In particular these names would compete with
whole-animal names for the purposes of homonymy but not
for priority. This proposal was fully supported by Arkell
(1957b) as an extension of his original intent.
This parataxon proposal has provoked much debate,
such that nearly thirty years later the question of parataxa
has again been “put aside... for further in-depth study and
future consideration.” (International Commission on
Zoological Nomenclature, 1985, p. xii). Until this question
is resolved, the assignment of specimens to specific taxa is
necessarily a cautious endeavour. Arkell (1957a) suggests
using Trauth’s (1927, 1928, 1930, 1931, 1935, 1936)
system of nomenclature as form-genera only, while others
simply refer to aptychi or anaptychi in general terms if
association with specific cephalopods cannot be proved
(Lehmann, 1971, 1981; Harper, 1989). Trauth’s genera,
however, are in many cases junior synonyms of older taxa.
In this work, the taxa described for Devonian specimens
will be considered appropriate, reserving “anaptychus” as a
morphological term only. The “genus” Anaptychus Oppel,
1856, is an erroneous citation, as Oppel used the word
merely as a morphological term describing the aptychus
seen in Ammonites planorbis (citation of Oppel, 1856, in
Moore and Sylvester-Bradley, 1957b). Anaptychus
Stimpson, 1860 (Crustacea), and Anaptychus
Schlumberger, 1868 (Cephalopoda), are junior subjective
synonyms of Sidetes Giebel, 1847.
1991
Late Devonian Cephalopod Aptychi
55
Class Cephalopoda Leach, 1817
Genus Sidetes Giebel, 1847
Anaptychus Schlumberger, 1868; non anaptychus Oppel,
1856, morphological term; non Anaptychus Stimpson,
1860, Crustacea.
Pholadocaris Woodward, 1882
Ca nlioca ris W c )o D wa r [ ) , 1882
?Ellipsocaris Woodward, 1882
Lisgocaris Clarke, 1882
Spathiocaris Clarke, 1882
Idiotheca Girty, 1 909
Palanaptychus Trauth, 1927
Neoanaptychus Nagao, 1931
Type species
Sidetes stricitus Giebel, 1 849.
Diagnosis
Semielliptic carbonaceous structure, weakly convex.
Ornamentation of fine concentric lines about a medial
anterior (by definition) rostrum, parallel to posterolateral
margin, intersecting anterior margin at nearly right angles.
Description
Structure semielliptic, length 22 mm, weakly convex.
Broad, width twice length. Anterior margin straight. Posterior
margin smoothly curved. Ornamentation concentric with
posterior margin, perpendicular with anterior margin, finely
spaced at about 15/cm. Composition carbonaceous, probably
conchiolinous, with no evidence of calcareous component.
Locality of type species
Unknown, “from hard sandstone banks near Salzburg”
(Giebel, 1849). Age is Late Cretaceous (Senonian).
Type
Location unknown.
Remarks
Five species of Sidetes, described below, are recognised
from the Ohio Shale. All are preserved as carbonaceous
films, compressed and flattened to varying degrees. None
appears to be accompanied by a calcitic or aragonitic
component. All have a roughly semielliptic outline, and bear
concentric ornamentation which terminates at the anterior
margin in a manner unlike that of brachiopods or bivalves.
The specimens described by Cooper (1932) have been
reexamined, as they were collected from the time-equivalent
Woodford Shale of Oklahoma, a unit similar in character and
depositional setting to the Cleveland Shale (Cardott and
Lambert, 1985). The Woodford specimens are preserved in a
different manner than the Ohio forms, however. They do not
exhibit the extreme flattening common with the Ohio
specimens, but are to varying degrees three-dimensional,
with a significant mass of matrix material preserved within
the interior of the structure, almost as a “steinkern.” In
general, they seem to have been preserved within calcareous
or phosphatic concretions, a common alternative mode of
preservation for [Upper-] Paleozoic aptychi (Mapes, 1987).
All are similar to the Ohio specimens, with the addition of
one species, S. gouldi, not recognised in the Ohio material.
Clarke’s (1882) type species of Spathiocaris was
Spathiocaris emersoni , originally described from New York.
His original specimens have not been discovered. We were
able to examine two specimens from Virginia (Butts, 1942),
deposited at the National Museum of Natural History.
In his discussion of Spathiocaris lata , Ruedemann
described transverse frontal grooves extending halfway to
the anterolateral angles. These grooves caused him to,
“recall those of the aptychus of some ammonites”
(Ruedemann, 1916, p. 95). Such grooves are not seen in any
specimen in this study. It is possible that they may represent
in some manner a reflexed portion of the anterior margin
analogous to the short inner lamella seen in anaptychus-type
ammonite jaws from the Mesozoic (Lehmann, 1979; Kanie,
1982; Tanabe, 1983). Such structures may well have been
obliterated in highly compressed material as is common in
the Ohio Shale. Exterior molds, of course, would not reveal
the inner lamellae. Finally, the Woodford Shale (Cooper,
1932) material has not been prepared to reveal the interiors
of the convexly folded specimens, so such structures may
well be preserved within the matrix which remains.
The species are differentiated on the basis of the ratio of
breadth to length, the angle of the anterior margin, the
general outline of the structure in its extended, flattened
form, and, to a lesser degree, the nature of ornamentation.
To this end, the following key is provided as an aid in
species identification:
KEY TO DEVONIAN APTYCHI
REFERRABLE TO SIDETES
la) Width greater than length 2
lb) Width equal to or less than length 3
2a) Width twice length; ornamentation finely
spaced, 1 6/cm 5. newberryi
2b) Width about 46 length; ornamentation coarse,
about 8/cm S. gouldi
3a) Width half to 4/5 length 4
3b) Width about equal to length 5
4a) Anterior margin acutely convex;
outline elliptical S. ulrichi
4b) Anterior margin straight or broadly
concave; outline triangular S. chagrinensis
4c) Anterior margin acutely concave,
notched S. emersoni
56
Frye and Feldmann
No. 46
FIGURE 3. Sidetes chagrinensis ( Ruedemann, 1916). 1, Neotype , CMNH 3745, from Big Creek in Cleveland. 2, CMNH 3746, from Chance
Creek in Lorain Co. 3, USNM 112031 , from the Arbuckle Mountains of Oklahoma. (Spathiocaris tenuicosta Cooper, 1932). Scale is one
centimeter.
5a) Outline sub-pentagonal, lateral margins
nearly parallel, meeting anterior margin
at distinct angle 5. lutheri
5b) Outline semielliptical, greatest width near
anterior margin, narrowing posteriorly
Anterior margin broadly concave S. lata
Sidetes chagrinensis (Ruedemann, 1916)
Figures 3.1 - 3.3
Spathiocaris chagrinensis Ruedemann, 1916, p. 95.
Spathiocaris tenuicosta Cooper, 1932, p. 350.
Diagnosis
Structure elongate semielliptical, length greater than
width. Posterior margin strongly rounded and narrow;
greatest width anterior near rostrum. Lateral margins but
slightly curved, extending obliquely forward; anterior
margin straight or broadly concave. Concentric ridges
closely arranged, not very prominent. Fine longitudinal lines
along hood radiate from rostrum to posterior tip.
Description
Fossil elongate semielliptical, appearing as a rounded
isosceles triangle with base at anterior margin. Length about
one quarter greater than width, ranging from 25 to 38 mm.
Posterior margin strongly rounded and narrow, lateral
margins slightly sinuously curved, extending obliquely
forward to anterior margin, which is straight or slightly
obtuse and concave. Concentric ridges closely arranged (14-
28/cm), not very prominent. Fine longitudinal lines along
hood radiate from rostrum to posterior tip, diverging slightly.
Type
Ruedemann’s holotype was collected from the Chagrin
Shale at Chippewa Creek in Brecksville, Ohio. It was in the
Western Reserve University collection, parts of which have
been transferred to The Cleveland Museum of Natural
History. This specimen has not been located, however. For
this reason, we designate CMNH 3745 as the neotype to
serve in place of the missing holotype.
Material
Examined in this study were two specimens collected in
June or July of 1965 by William Hlavin from the Cleveland
Shale. CMNH 3745 is from Big Creek near W. 117th St. in
Cleveland, and CMNH 3746 is from Chance Creek, Lorain
County. Also studied were two specimens from the
Woodford Shale assignable to this species, namely USNM
1 12031, Cooper’s (1932) holotype of S. tenuicosta , and one
of the paratypes of S. gouldi , USNM 112035, both from the
Arbuckle Mountains of Oklahoma.
Remarks
Sidetes newberryi, S. lutheri , and S. lata are each broader
than S. chagrinensis , with width/length ratios approximately
1991
Late Devonian Cephalopod Aptychi
57
one or greater. Sidetes ulrichi and S. emersoni , while also
narrow, differ significantly in that the anterior margins are
strongly curved rather than straight. Few specimens from
the Chagrin Shale are well-preserved, and none of those
now available represent this species. Even though the
specimens in this study are from a different unit than the
primary types, in all other respects they appear to conform
to Ruedemann’s description and illustrations of the species.
Cooper’s (1932) species, Spathiocaris tenuicosta ,
differs from S. chagrinensis only in the manner of its
preservation. It is folded along the median into a groove
and ridge rather than a simple crease. In all other respects,
it resembles the other specimens described here, and is
properly assigned to S. chagrinensis.
Sidetes emersoni (Clarke, 1882)
Figures 4. 1 , 4.2
Spathiocaris emersoni Clarke, 1882, p. 477.
Diagnosis
Structure semielliptical, length greater than width.
Posterior margin rounded to subtriangular. Narrow, greatest
width at anterior wing angle. Anterior margin deeply
concave, notched. Concentric ridges closely arranged,
well-marked.
Description
Fossils are semielliptical to subtriangular, width 2/3 to 4/5
length. Greatest width measured between wing angles,
anterior of rostrum. Length ranges from 1 8 to 40 mm,
width from 12 to 32 mm. Posterolateral margins straight to
broadly curved. Ornamentation fine, about 20/cm,
concentric with posterolateral margin, bending toward
median at anterior margin. Anterior angle deeply concave,
angle near 120°.
Types
Clarke’s type specimens were from the Portage shales in
Naples, Ontario Co., New York. They have apparently been
lost. Two specimens collected by Butts (1942) and now in
the National Museum of Natural History, USNM 97992-a
and -b, are designated as “hypotypes.”
Material
The specimens studied were the hypotypes, from
Millboro, 1.6 km south of Shawver Mill, Virginia.
Remarks
Sidetes emersoni is narrower than S. newberryi or S.
gouldi , while the deep anterior angle on S. emersoni
differentiates it from S. ulrichi , S. chagrinensis , and S. lata.
In outline, S. emersoni does not possess the sharp angle
2
FIGURE 4. Sidetes emersoni (Clarke. 1882). Hypotypes from
Millboro, Virginia. 1, USNM 97992-a. 2, USNM 97992-b. Arrow
points to fragment of opposite wing presetted on elevated portion
of the matrix. Scale is one centimeter.
where posterior and lateral margins meet, as does S.
lutlteri , being instead nearly triangular.
It is interesting that none of the Ohio nor Oklahoma
specimens can be assigned to this, the type species for the
genus Spathiocaris. Only the one specimen of S. lutheri
may possess an anterior margin as concave as S. emersoni.
All other specimens studied have anterior margins which
are less concave or even convex.
Sidetes gouldi (Cooper, 1 932)
Figures 5.1, 5.2
Spathiocaris gouldi Cooper, 1932, p. 349.
Diagnosis
Structure large, semielliptical, broad; width about 1.5
times length. Anterior angle straight or broadly concave.
Ornamentation concentric with posterolateral margin,
terminating anteriorly with inward bend toward median;
spacing coarse, about 8/cm.
Description
Fossils are large, length about 40 mm; outline
semielliptical; broad, width about 65 mm (1.5 times
length). Anterior angle appears nearly straight.
Ornamentation consists of ridges concentric with
posterolateral margin, spaced about 8/cm. Anterior portions
of ridges curve inward toward median.
58
Frye and Feldmann
No. 46
Figure 5. Sidetes gouldi
f Cooper , 1932), from the
Arbuckle Mountains. 1, USNM
112023, holotype, from Sycamore
Creek, OK. 2, USNM 112034.
Scale is one centimeter.
Types
Cooper’s (1932) type material includes the holotype,
USNM 112023, and a paratype, USNM 112034. USNM
112035, another paratype, is smaller and is ornamented
with much finer ridges than the other two specimens. It is
also much narrower in outline than the others, and properly
should be referred to S. chagrinensis. The two type
specimens definitely referrable to this species are from the
Arbuckle Mountains in Oklahoma; the holotype was
collected from Sycamore Creek.
Material
Specimens studied were USNM 112023 and 112034.
None of the Ohio specimens of appropriate breadth exhibit
ornamentation so coarse as to permit their assignment to
this species.
Remarks
Sidetes gouldi and S. newberryi are the broadest species
studied, all others being much narrower. Sidetes gouldi is
slightly less broad than S. newberryi , which approaches a
width/length ratio of 2 to 1. Sidetes gouldi is further
distinguished from S. newberryi in that the ornamentation
is half as finely spaced as that of the latter.
The two specimens have been removed from their
surrounding concretions, and the anterior margins are poorly
preserved, making determination of the anterior angle difficult.
There is no indication, however, that it was significantly
concave, but rather it appears to be nearly straight. While the
spacing of the ornamentation near the rostrum frequently is
finer than elsewhere on specimens of Sidetes, even this finer
region is more coarsely ornamented than the ridges of S.
newberryi , allowing easy distinction of the two species.
Sidetes lata (Ruedemann, 1916)
Figures 6.1 - 6.7
?Cardiocaris lata Woodward, 1882, p. 388.
Spathiocaris lata Ruedemann, 1916, p. 94.
Spathiocaris striatula Cooper, 1932, p. 351.
Diagnosis
Structure semielliptical, length about equal to greatest
width, which is near anterior margin. Anterior margin
broadly concave. Ornamentation fine, concentric with
posterolateral margin.
Description
Fossil semielliptical, posterolateral margin nearly
circular or slightly flattened posteriorly. Greatest width,
near anterior margin, approximately equal to length.
Anterior angle broadly concave. Ornamentation fine, about
24/cm, concentric with posterolateral margin, bending
toward median at anterior margin.
1991
Late Devonian Ceehaeopod Aptychi
59
FIGURE 6. Sidetes lata (Ruedemann, 1916). 1, CMNH 83 1 3a. from Big Creek at Brookside Park. Cleveland. 2, CMNH 8307a. also from
Big Creek. 3, USNM 112038. syntype of Spathiocaris striatula Cooper. 1932. 4, L1SNM 112032. S. striatula syntype. 5. USNM 112028 6,
USNM 112029 7, USNM 112041. Scale is one centimeter.
60
Frye and Feldmann
No. 46
FIGURE 7. Sidetes lutheri ( Clarke , 1882). 1, CMNH 8302. Scale is one centimeter. 2, USNM 264093, the holotype. Scale is one millimeter
Type
Ruedemann’s holotype was collected in 1895 from the
“Chemung beds” near Avoca, Steuben Co., New York. It is
now in the New York State Museum, NYSM 9860.
Material
Specimens studied which can be assigned to this species
include CMNH 8307a and CMNH 8313a from Big Creek
at Brookside Park. Woodford Shale specimens which can
be assigned to this species are the type specimens of S.
striatula , USNM 1 12038 and 112032, and three specimens
formerly identified as S. williamsi , USNM 112037, 1 12029,
and 112041. These last specimens do not conform with
Ruedemann’s description (1916) of Spathiocaris williamsi ,
being symmetrical and more narrow.
Remarks
Some specimens show superimposed, concentric
undulations which are flatter and not as pronounced as those
seen in S. ulrichi. Sidetes lutheri is distinctly different in
outline from S. lata , while the other taxa are either decidedly
narrower or significantly broader. Ruedemann described
short carbonaceous lines radiating from a semicircular area
at the anterior angle. These may be artifacts of preservation,
as they have not appeared in any other specimen described
from the North American Paleozoic.
Spathiocaris striatula was distinguished from the other
Woodford Shale species by striations “radiating from the apex
to the lateral and posterior margins of the shell” (Cooper,
1932). This feature is variously affected by preservation.
Ruedemann (1916) noted imperfectly preserved striations in
several species. Such striations are not seen in any of the Ohio
specimens, but are present in several of the Oklahoma fossils,
which are generally better preserved. The proportions of S.
striatula and the ornamentation are identical with Sidetes lata ,
to which it should be referred.
Sidetes lutheri (Clarke, 1 882)
Figures 7.1, 7.2
Lisgocaris lutheri Clarke, 1882, p. 478.
Pholadocaris lutheri Ruedemann, 1916, p. 94
Diagnosis
Outline sub-pentagonal, lateral edges parallel, meeting
posterior margins at sharp angles.
1991
Late Devonian Cephalopod Aptychi
61
FIGURE 8. Sidetes newberryi ( Whitfield , 1882). Type specimens, from Erie County, Ohio. The remaining specimen in the type series is too
degraded to reveal detail, and is not illustrated here. 1, AMNH CU 7452G. 2, AMNH CU 6686G. 3, AMNH CU 551 3G. 4, AMNH CU
6063G. Scale is one centimeter.
Description
Fossil large, length greater than 47 mm, width 43 mm.
Outline sub-pentagonal; lateral margins parallel and
meeting posterior margins at a sharp angle. Surface
ornamented with concentric ridges parallel to margins,
spaced about 16/cm. Ridges turn inward toward rostrum
near anterior margin, which is not preserved.
Type
The holotype, USNM 264093, is in the collection of the
U.S. National Museum. It was collected from near the base of
the Hamilton Formation in Mile’s Gully, Hopewell, New York.
Material
In addition to the holotype, one Ohio specimen was
examined. CMNH 8302 was collected in 1938 by P.A.
Bungart from an unidentifiable locality near Linndale in the
Berea quadrangle map. No stratigraphic information was
recorded.
Remarks
The unique outline and approximately equal length and
width serve to distinguish S. lutheri from all other taxa in
this study. Clarke's (1882) description of Lisgocaris lutheri
was based on a single, very small specimen measuring only
two by three millimeters. The distinctive configuration of
the margins and concentric ornamentation, however, is
maintained in the much larger CMNH specimen. Clarke
described the species in reverse orientation to that given
above, with an “abdominal” cleft beginning centrally and
widening to the “posterior” margin. This anterior region is
not preserved in the CMNH specimen, so the size and
shape of such a cleft cannot be determined. There is
evidence that the lateral margins extended anterior of the
rostrum a slight distance, but whether the anterior angle
was obtuse or acute remains unknown.
Sidetes newberryi (Whitfield, 1882)
Figures 8. 1-8.4, 9. 1-9.7
Plumulites newberryi Whitfield, 1882, p. 217.
Turrilepas newberryi Hall, 1888, p. 219-220.
Idiotheca rugosa Girty, 1909, p. 40.
Spathiocaris cushingi Ruedemann, 1916, p. 96.
Spathiocaris woodfordi Cooper, 1932, p. 351.
Spathiocaris plicifera Cooper, 1932, p. 350.
Diagnosis
Structure large, semielliptical, broad; width about twice
length. Anterior angle straight or broadly concave.
Ornamentation concentric with posterolateral margin,
terminating anteriorly with inward bend toward the
rostrum; spacing fine, about 16/cm.
Description
Structure large, length from 10 to 43 mm, semielliptical
in outline, folded upon itself along the median line, forming
a curved hinge. Broad, width 1.5-2 times the length.
Anterior angle straight or very obtusely concave; anterior
wings extend beyond hinge about one-fifth of total length.
Entire structure extremely thin and flattened; in some
specimens lateral margin is interrupted by one or more
fissures, perpendicular to margin. Ornamentation of ridges
concentric with posterolateral margin, spaced about 16/cm.
Ridges continuous, terminating at anterior margin;
anteriormost portion bent inward toward rostrum.
Types
Sidetes newberryi is the widest species herein studied.
Only S. gottldi approaches it in width, but has much coarser
ornamentation. All other species are rather narrower.
Ruedemann (1916) described the new species Spathiocaris
62
Frye and Feldmann
No. 46
Figure 9. Sidetes newberryi ( Whitfield , 1882). Specimens from the Cleveland and Woodford shales. 1, CMNH 8311, and 2, CMNH 8318,
from Big Creek, arrows showing fissures extending inward from posterior lateral margin. 3, CMNH 8320. 4, CMNH 8327a, with gypsum
encrustation. 5, USNM 112030, holotype of Spathiocaris plicifera Cooper, 1932. 6, USNM 112040, and 7, USNM 112033, types of
Spathiocaris woodfordi Cooper, 1932. Scale is one centimeter.
1991
Late Devonian Cephalopod Aptychi
63
Figure 10. Sidetes ulrichi (Cooper, 1932). 1, USNM 112036, the holotype. 2, CMNH 3744, from Chance Creek. 3, CMNH 8305b, also
from Big Creek. 4, CMNH 8328a. 5, CMNH 8328b. Scale is one centimeter.
cushingi based upon two specimens of Turrilepas
newberryi collected from the Cleveland Shale. These
specimens were part of the Western Reserve University
collection, parts of which ar now at The Cleveland
Museum of Natural History. These two specimens have
not been located, however.
Whitfield’s specimens of Plumulites ( Turrilepas )
newberryi are at the American Museum of Natural
History (AMNH), and include the syntypes AMNH
CU 7452G, AMNH CU 6685G, AMNH CU 6063G,
AMNH CU 6686G, and AMNH CU 5513G. They
were collected from the Cleveland Shale near
Sheffield and Birmingham, in Erie County, Ohio.
These agree in all details with Ruedemann’s
description and the specimens studied here, and do
not exhibit the median carina nor multiple imbricate
plates which might place them with the cirripedes.
This species belongs with the aptychi.
64
Frye and Feldmann
No. 46
FIGURE 11. Scanning electron micrographs of the cross-sections of the unmineralized wing area of cephalopod jaws. No internal
structure is discernable in either specimen. 1, Sidetes sp. CMNH 8317. a presumed Devonian anaptychus. 2, Nautilus pompilius, a modern
nautiloid. Scales are in microns as indicated. The Devonian specimen has been considerably compressed.
Material
Material studied includes Whitfield’s five specimens of
Turrilepas, CMNH 8311 and CMNH 8318, from Big
Creek, collected as float material, and CMNH 8312, from
Abram’s Creek. CMNH 8320 and CMNH 8327a, also
studied, have no reliable collection data recorded.
Ruedemann’s types were collected by Professor H.P.
Cushing from the Cleveland Shale along Cahoon Creek,
not far from Cahoon Cliffs. One specimen, CMNH 8668,
has been found within the old Western Reserve University
collection, but it does not appear to be one of Ruedemann’s
type specimens. Woodford Shale material assigned to this
species are Cooper’s (1932) types of S. woodfordi , USNM
112033 and 112040, and the holotype of S. plicifera ,
USNM 112030.
Remarks
Ruedemann’s description of Spathiocaris cushingi
included mention of a second set of concentric lines
centered upon the wing angles of the valve. This set “is but
faintly shown’’ in the second, larger specimen he described,
and may be invisible depending on preservation. Its
absence should not rule out the assignment of the CMNH
specimens to this species. Girty (1909, p. 40) illustrated a
new genus and species of Pteropoda, Idiotheca ritgosa ,
from the “Woodford chert” [sic] at the base of the Caney
Shale (Devonian-Mississippian) in Oklahoma. He was
uncertain of its affinities, granting that it “may possibly be
an aptychus... which occur so abundantly at a little higher
horizon...” (Girty, 1909). Cooper (1932) redescribed the
same specimen, USNM 112044, along with another.
USNM 112033, as the new phyllocarid species
Spathiocaris woodfordi. His Spathiocaris plicifera ,
holotype USNM 112030, differs from the rest only in
preservation, bearing secondary corrugated folds. These
Woodford Shale specimens are indistinguishable from
Sidetes newberryi , and belong within this taxon.
Ruedemann (1916) also described another species from
the Cleveland Shale along Mill Creek in Newburg, Ohio
(Cleveland), Spathiocaris williamsi, which may represent
another folded specimen with a hinge more strongly
curved than S. newberryi. He described S. williamsi as
being asymmetrical and having the apex or rostrum
displaced to one side of the “median line.” This
asymmetry is suspect. If we consider the lateral margin of
his specimens nearest the rostrum to be the folded hinge
line of a compressed specimen, the half which remains
visible strongly resembles S. newberryi. Possible
differentiating features might be coarser ornamentation
(about 10/cm) and a superimposed concentric furrowing
with a spacing of about 3-4 mm. Insofar as Ruedemann’s
types have not been located for study, it seems prudent to
consider the two taxa as separate.
Sidetes ulrichi (Cooper, 1932).
Figures 10.1 - 10.5
Spathiocaris ulrichi Cooper, 1932, p. 352.
Diagnosis
Structure sub-elliptical, narrower at rostrum. Anterior
margin acutely convex. Broadly convex, highest point
1991
Late Devonian Cephalopod Aptychi
65
0.0 2.0 4.0 6.0 8.0 10.0
Energy [KeV]
2 Concauicaris sp., Cleveland Shale
0.0 2.0 4.0 6.0 8.0 10.0
Energy [KeV]
3 CMNH 8317, Cleveland Shale
0.0 2.0 4.0 6.0 8.0 10.0
Energy [KeV]
FIGURE 12. X-ray emission spectra revealing relative
composition of some Cleveland Shale specimens. 1, a lingulid
brachiopod. 2, Concavicaris, a crustacean, CMNH 3740. 3,
CMNH 8317, a presumed Devonian anaptychus.
posterior of rostrum. Superimposed upon fine, concentric
ornamentation are broad undulations parallel to
ornamentation.
0.0 2.0 4.0 6.0 8.0 10.0
Energy [KeV]
0.0 2.0 4.0 6.0 8.0 10.0
Energy [KeV]
FIGURE 13. X-ray emission spectra revealing relative
composition of 1, an unidentified cephalopod from the Cleveland
Shale, CMNH 8705 2, the wing or collar region of the jaw of
modern Nautilus.
Description
Outline nearly elliptical, narrower at rostrum, length 10
to 60 mm, width about two-thirds of length. Surface
corrugated in broad concentric undulations subparallel with
posterior margin and intersecting lateral margins, spaced
about 3 mm apart in a 30 mm specimen. Ornamentation
similarly oriented, much more finely spaced, 30-60/cm.
Type
Cooper’s holotype of Spathiocaris ulrichi was loaned to
him by Dr. George H. Girty of the U.S. Geological Survey.
It was collected by E.O. Ulrich from the Woodford
Formation (Late Devonian) near Dougherty, Oklahoma. It
is now at the U.S. National Museum, USNM 112036. The
type is 60 mm long and 44 mm wide, somewhat larger than
those from the Ohio Shale.
66
Frye and Feldmann
No. 46
Material
Specimens studied include the holotype; CMNH 8303
and CMNH 8305b, from Big Creek; CMNH 3744 from
Chance Creek; and CMNH 8328a and CMNH 8328c, from
an unknown locality. All the Ohio specimens are from the
Cleveland Shale.
Remarks
Sidetes ulrichi is the single species examined which
bears a convex anterior margin. All others have straight or
indented margins anteriorly. The dimensions of the
specimens studied range from a width-to-length ratio of 0.6
to 0.8; all but the most compacted and flattened exhibit the
corrugations superimposed upon the finer concentric
ridges. The flattest are assigned to this species primarily on
outline and fineness of ornamentation.
Microstructure and Composition
Several of the better preserved specimens, interpreted to
be aptychi, were prepared for examination with the scanning
electron microscope. The calcareous portions of cephalopod
aptychi have a distinctive internal microstructure (Lehmann,
1981). This structure, if identifiable in the fossil specimens,
would confirm their identification. The microstructure of
authentic arthropod cuticle, and brachiopod and bivalve
shells was examined for comparison with the study
specimens. Fish scales were eliminated from consideration
as those in The Cleveland Museum of Natural History
collection from the Late Devonian have morphologies
distinctly different from these aptychi (M. Williams, CMNH,
personal communication).
An International Scientific Instruments Model SX-40A
SEM was used, with an attached Princeton Gamma Tech
System 4 Plus energy dispersive x-ray spectrometer. It is
routine practice in electron microscopy to coat the surface
of the specimen with a conductive material to drain the
accumulation of electric charge built up by electron
bombardment.
Several of the first specimens examined were coated
with gold to a thickness of approximately 500 angstroms,
using ISI’s P-Sl diode sputter coater. This procedure
interfered with the use of the x-ray spectrometer, however.
The K-alpha emission line for phosphorous has an energy
of 2.014 KeV, while gold has an M-alpha emission line at
2. 1 23 KeV, too close to the phosphorous line to be resolved
(Goldstein, et al., 1981). A commonly used alternative is
carbon coating, since the emission spectrum of carbon is
entirely absorbed by the beryllium window in the detector
apparatus. In the absence of a carbon coater, uncoated
specimens were examined. This was successful, perhaps
due to the high (4.60 ± 1 .09 wt.%) average organic carbon
content of the Cleveland Shale (Broadhead, et al., 1982),
and the carbonaceous nature of the specimens themselves.
Examination of several aptychus specimens revealed no
discernable structure remaining within the thin
carbonaceous film (Figure 11.1). All thicker regions
examined were indistinguishable from the shale matrix and
appeared to be molds. It seems that this method of
investigation is of little value with material reduced to a
carbonaceous film under anaerobic or dysaerobic
preservational regimes.
The instrument was used to search for elements that
might allow identification of the material composition of the
aptychi. Qualitative energy-dispersive X-ray (EDX) spectra
were obtained from aptychi and similar appearing fossils and
parts from living organisms. The X-ray energy range from 0
to 20 thousand electron volts was scanned, allowing for the
detection of nearly all the elements. Only those with atomic
number less than berylium were undetectable, as the detector
apparatus blocks X-rays from these elements. From the
spectra obtained, the relative concentrations of calcium,
strontium, and potassium within the samples were used to
test for the calcitic, aragonitic, or phosphatic nature of the
fossils. Other elements discovered were identified and
indicated in the figures. The spectra were compared to test
the usefulness of the method in resolving this question.
Inarticulate brachiopods and certain arthropods have
long been considered the most likely alternative taxa to
which aptychus-like structures might be assigned (Clarke,
1902). Therefore, specimens of a lingulid brachiopod
(CMNH uncataloged) and the crustacean Concavicaris
(CMNH 3740) from the Cleveland Shale were prepared for
SEM and EDX examination. Analyses of these specimens
and a representative anaptychus, CMNH 8317, are shown
in Figure 12. While phosphorous is clearly present in both
the phosphatic brachiopod and the crustacean, it is just as
clearly absent from the anaptychus. Also conspicuously
absent from the anaptychus is calcium or strontium, the
latter being a common marker impurity used to identify
aragonite (Crick, et al., 1987). It may be impossible to
positively identify originally calcareous material from
these units, however. Cephalopods tested from the
Cleveland Shale, for example the one shown in Figure
13.1, were found to be significantly replaced with pyrite,
which can be expected in the metal-rich, anaerobic
conditions of deposition (Baird and Brett, 1986). Strontium
in particular, present originally in trace amounts, may be
undetectable in these altered specimens. The lack of pyrite
replacement, common to many shelly fossils in the
Cleveland, may suggest that the anaptychi were entirely
chitinous in nature, with no mineralized portions.
Modem Nautilus mandibles contain phosphorous, but in
trace quantities. While the exposed, oral portions of the
mandibles are mineralized, the muscle insertion areas are
often only lightly calcified if at all, consisting largely of a
chitin/protein complex, only thinly coated with aragonite.
1991
Late Devonian Cephalopod Aptychi
67
FIGURE 14. Aptychopsis Barrande, 1872. 1, LO 5270 and 2, LO 5268, from a Silurian shale quarry’ in southern Sweden (Stridsberg,
1984). Specimens from the Department of Historical Geology and Palaeontology, University of Lund, Sweden. 3, Reconstruction of
opercular position of aptychopsid plates in an ortlwconic nautiloid, after Turek, 1978.
This aragonitic layer contains small deposits of brushite, a
phosphatic mineral (Lowenstam, et al., 1984; Lowenstam
and Weiner, 1989). Total phosphorous content of this
posterior region is on the order of 0.30%. This posterior
region of the Nautilus mandible also reveals little internal
microstructure beyond subtle layering, even with SEM
examination (Figure 11.2). What structure is seen seems to
be an artifact of breakage. The thinness of the jaw specimen
allowed for electron beam penetration through it and into
the aluminum mounting stub when in the SEM. The X-ray
analysis in Figure 13.2 reveals a strong aluminum peak for
this reason. The other elements present in the analysis are
common trace elements in sea water, and appear to have
been incorporated into the structure in significant amounts.
Perhaps diagenetic alteration resulted in the depletion of
chlorine and enrichment in iron seen in the Cleveland Shale
aptychi. The phosphate minerals were shown by
Lowenstam, et al. (1984) to be limited to the carbonate
layers, which are not present in the Devonian material.
Thus, the x-ray analysis presents evidence denying an
arthropod or brachiopod affinity for these fossils. While
there are other possible origins for carbonaceous fossil
fragments, two of the most likely alternatives based upon
the morphology of the fossils are eliminated from
consideration. The most parsimonious interpretation is that
these are indeed cephalopod aptychi, as suspected by Clarke
( 1902), Girty (1909), and particularly Ruedemann (1916).
Aspects of Functional Morphology
Historically, when cephalopods were first found with
aptychi in place, the approximate match between their
outline and the aperture of the conch suggested they served
as opercula (Woodward, 1885b; Clarke, 1902; Trauth,
1927). This correspondence is by no means exact, however
(Lehmann, 1972, p. 42). Other functions postulated for
these structures were as covers for the nidamentary glands
or as cartilaginous plates for funnel muscle attachment.
Ruedemann (1916, p. 102) suggested the latter, “would
naturally also have existed in the Ordovician and Silurian
cephalopods...” in attempting to explain the Discinocarida.
This prescient speculation was proven sound by the
discovery of specimens of Aptychopsis Barrande, 1872 in
situ in the apertures of orthoconic nautiloids from the
Silurian of central Bohemia (Turek, 1978) and southern
68
Frye and Feldmann
No. 46
FIGURE 15. Hypothesized original shapes of aptychi. 1,
Reconstruction of a Cleveland Shale anaptychus. 2,
Reconstruction of the mandibles of Psiloceras, Hyatt 1867 ( from
Lehmann 1971.)
Sweden (Holland, et al., 1978). Aptychopsid plates form a
neat, flat circular structure almost precisely fitting the
aperture of the nautiloid (Stridsberg, 1984). The three plates
involved would be difficult to fold into a concave structure
(Figure 14.1 ). It would seem that these are indeed opercula.
The structures known as aptychi in Mesozoic ammonites
seem to have served a different function. Again, discoveries
of aptychi in situ in cephalopod body chambers provided
evidence of their function. These were found to be curved
structures, sometimes found associated with an element
resembling the upper jaw of Nautilus (Lehmann, 1971, 1972,
1978, 1981; Tanabe, 1983; Tanabe and Fukuda, 1987). It has
become accepted by many that these aptychi are the lower
jaws of ammonites (Lehmann, 1981; Morton, 1981; Mapes,
1987) . One difficulty with this interpretation is the sheer size
of the structure, relative to both the upper jaw components
and the diameter of the body chamber. In many cases, the
aptychus approximates the cross-sectional area of the body
chamber. Lehmann’s work, cited above, involved serial
sectioning of cephalopods, revealing the three-dimensional
relationship of the body chamber and its contents, and seems
to be the most accurate means of determining the nature of
these structures, at least for the Mesozoic ammonites studied.
There are no rigid calcareous plates found with the
Devonian fossils, so their preservation as flattened bodies
does not necessarily reflect their original morphology. Two
specimens of Sidetes newberryi, CMNH 8311 and CMNH
8318, exhibit short cracks radiating inward from the
posterolateral margin (Figures 9.1, 9.2). These cracks are
precisely what would be expected in a convex structure
which has been flattened. If this is the origin of the cracks,
then closing the gaps should approximate the original form
of the structure. As these specimens were folded in half
along the medial line, photographs of them were prepared
in both normal and reversed orientations. These were then
photocopied and enlarged. The resulting reproductions
were then cut out and the mirror images were attached
along the midline. This caused the paper models to become
slightly convex. The cracks along the margin were then cut
away and the edges rejoined. This caused the models to
assume a broad, scooped shape, similar to that seen in
Lehmann's (1970) reconstruction of the jaws of Psiloceras
from the Jurassic of Germany (Figure 15). Many of the
specimens preserved as flattened bodies are folded along
the median. In some cases they can be separated from the
matrix and studied from both sides, or often the remains of
the upper half are preserved along the margins of the extant
lower portion (see the arrow in Figure 4.2, S. emersoni).
These reconstructions of ammonite jaws (Lehmann,
1970; 1975), suggest to some a scooping, shoveling
application rather than a true biting action seen in modem
Nautilus (Lehmann, 1972), and in other Mesozoic forms
(Nixon, 1988). This conclusion is based on the large size of
the lower jaws (particularly anaptychi), both relative to the
size of the upper jaw and to the total size of the conch. The
lack of calcified rostra, or conchorhynchs, to serve as biting
surfaces further serves as a basis of interpretation. Calcified
beaks have been reported from the Permian (Closs, 1967)
and perhaps from the Mississippian (Landman and Davis,
1988) , but no older specimens have been found. Where
associated upper jaws are of comparable scale, however, it
would seem that, even lacking calcified beaks, the jaws
could be capable of strong biting action similar to modem
coleoids, which lack such calcified surfaces (Tanabe, et al..
1991
Late Devonian Cephalopod Aptychi
69
1980). The shape of the lower jaw is similar for both
applications. Still, the curved shape revealed in the Ohio
Shale specimens by these methods, if genuine, fits known
jaw structures far better than hypothesized opercula or
Aptychopsis. We must conclude that these forms of Sidetes
from the Late Devonian represent cephalopod jaws which
have been preserved separately from the animal’s conch.
Summary and Conclusion
Approximately 120 specimens of Hat, carbonaceous
body fossils have been collected from the Cleveland and
Chagrin shales of northeastern Ohio. The fossils have been
found predominantly in the black Cleveland Shale, which
probably represents an anoxic basinal environment
(Broadhead, et ah, 1982). A few have been collected from
the underlying Chagrin Shale, a gray-green unit deposited
in dysaerobic conditions (Barron and Ettensohn, 1981;
Schwimmer, 1988, Schwimmer and Feldmann, 1990). In
addition, specimens housed at the National Museum of
Natural History which were collected by Cooper in 1932
have been reevaluated.
These fossils have been identified as representing seven
species of Sidetes Giebel, 1847, and interpreted to be
cephalopod jaw elements. This identification is based upon
the general outline of the specimens, and particularly the
pattern of ornamentation they exhibit. This concentric
pattern of fine ridges is unlike that seen in arthropods,
gastropods, bivalves, or brachiopods. Some of the
specimens are preserved unfolded and flattened, while
others are folded in half along the median line.
Further evidence for this interpretation is furnished by
energy dispersive X-ray spectrometry. These fossils show
no trace of having once contained phosphorous within the
structures studied. It is unlikely that alteration after burial
would have so completely removed the element, as both
arthropods and inarticulate brachiopods from the same unit
have remained phosphatic. On this basis, we can conclude
that the Sidetes specimens in this study are neither
arthropods nor brachiopods.
Two specimens showed signs of compaction damage.
When restored to their presumed original shape, they
resemble reconstructions of Mesozoic ammonite jaws. It
appears that these fossils served the same function in
some Devonian cephalopods. By contrast, the Silurian
Aptychopsis may have been a nautiloid operculum rather
than a jaw element.
Acknowledgements
Ellis Yochelson, of the Department of Paleobiology, U.S.
National Museum of Natural History, Washington, D.C.,
helped launch this study and offered helpful comments on
the morphology of aptychi. His review of a previous version
of this manuscript was quite helpful. While our conclusions
differ in some particulars, his comments were most
constructive. Roy Mapes, Ohio University, also helped to
considerably improve the manuscript. The Palaeontological
Institute of the University of Lund loaned the Aptychopsis
specimens for comparison. Jann Thompson of the National
Museum of Natural History located one of Clarke’s
holotypes and those from Cooper's Oklahoma work, and
was of much assistance in our examination of them. Melvin
Hinkley provided information and a loan of Whitfield’s type
specimens from the American Museum of Natural History,
New York. Mary Baum and Wendy Weitzner, of The
Cleveland Museum of Natural History, helped uncover the
more obscure references, and Shya Chitaley and JoAnn
Cobum, CMNH, searched their collections for a few stray
aptychi. Joe Hannibal, CMNH, offered tireless assistance
and made possible the loan of the CMNH specimens
studied. Thanks are also due to Richard E. Carroll of
Michigan State University for collecting one of the
specimens in situ and donating it to The Cleveland Museum
of Natural History for our study and to Donald F. Palmer
and Alan H. Coogan who read an earlier draft of the
manuscript. Contribution 455. Department of Geology, Kent
State University, Kent, Ohio 44242.
References
Arkell. W.J. 1954. Proposed declaration that a generic or specific
name based solely upon the "aptychus" of an ammonite be
excluded from availability under the Regies. Bulletin of
Zoologic Nomenclature 9(9):266-269.
Arkell. W.J. 1957a. Aptychi, pp. L437-L441. In: Treatise on
Invertebrate Paleontology , Volume L, Mollusca 4,
Cephalopoda, Ammonoidea. Edited by R.C. Moore.
Geological Society of America and University of Kansas
Press, Lawrence.
Arkell, W.J. 1957b. The Moore/Syvester-Bradley "parataxa"
proposals. Bulletin of Zoologic Nomenclature 15:78.
Baird, G.C., and C.E. Brett. 1986. Erosion on an anaerobic
seafloor: significance of reworked pyrite deposits from the
Devonian of New York State Palaeogeography,
Palaeoclimatology, Palaeoecology 57:157-193.
Barron, L.S., and F.R. Ettensohn. 1981. Paleoecology of the
Devonian-Mississippian Black-shale Sequence in Eastern
Kentucky With an Atlas of Some Common Fossils. U.S.
Department of Energy, Morgantown Energy Technology
Center, DOE/METC/ 12040- 151, 75 pp.
Brady, L.F. 1955. Possible nautiloid mandibles from the Permian
of Arizona. Journal of Paleontology 29: 1 02- 1 04.
Broadhead. R.F., R.C. Kepferle, and PE. Potter. 1982. Stratigraphic
and sedimentologic controls of gas in shale — example from
the Upper Devonian of Northern Ohio. American Association
of Petroleum Geologists Bulletin 66:10-27.
Butts, C. 1942 (1941). Geology of the Appalachian Valley in
Virginia, Pt. 2, Fossil plates and descriptions. Bulletin 52,
Virginia Geologic Survey. 27 1 pp.
70
Frye and Feldmann
No. 46
Campbell, G. 1946. New Albany Shale. Geological Society of
America Bulletin 57:829-890.
Clarke, J.M. 1882. New phyllopod crustaceans from the Devonian
of Western New York. American Journal of Science , 3rd
Series 23:477-478.
Clarke, J. 1902. A new genus of Paleozoic brachiopods, Eunoa ,
with some considerations therefrom on the organic bodies
known as Discinocaris, Spathiocaris, and Cardiocaris.
Bulletin of the New York State Museum 52:606-615.
Clarke, M.R. 1962. The identification of cephalopod “beaks” and
the relationship between beak size and total body weight.
Bulletin of the British Museum (Natural History ), Zoology
8:419-480.
Closs, D., M. Gordon, Jr., and E.L. Yochelson. 1964. Cornaptychi
from the Permian of Utah. Journal of Paleontology 38:899-
903.
Cooper, C.L. 1932. A crustacean fauna from the Woodford
Formation of Oklahoma. Journal of Paleontology 6:346-352.
Copeland, M.J., and T.E. Bolton. I960. Canadian fossil
Arthropoda, Eurypterida, Phyllocarida, and Decapoda.
Geological Survey of Canada Bulletin 60:35-38.
Crick, R.E., K.O. Mann, and J.A. Chamberlain, Jr. 1987. Skeletal
chemistry of Nautilus and its taxonomic significance.
Biochemical Systematic :s and Ecology 15:461-474.
Giebel, C.G. 1849. Briefliche Mittheilungen an Herrn Beyrich.
Zeitschrift der Deutschen Geologischen Gesellschaft
1 ( 2):99- 1 00.
Girty, G.H. 1909. The Fauna of the Caney Shale of Oklahoma.
Bulletin of the United States Geological Survey 377 . 106 pp.
Goldstein, .1.1. , D.E. Newbury, P. Echlin, D.C. Joy, C. Fiori, and E.
Lifshin. 1981. Scanning Electron Microscopy and X-ray
Microanalysis. Plenum Press, New York, 673 pp.
Hannibal, J.T., and R.M. Feldmann. 1983. Arthropod trace fossils,
interpreted as echinocarid escape burrows, from the Chagrin
Shale (Late Devonian) of Ohio. Journal of Paleontology
57(4):705-7 16.
Harper, J.A. 1989. Occurrence of a cephalopod aptychus in the
Ames Limestone (Virgilian) of western Pennsylvania.
Journal of Paleontology 63( 1 ): 1 1 9- 1 2 1 .
Hasenmueller, N.R., and R.K. Leininger. 1987. Oil-Shale
Prospects for the New Albany Shale in Indiana. Special
Report of the Indiana Geological Survey, Dept, of Natural
Resources, Bloomington, 40:31 pp.
Hlavin, W.J. 1976. Biostratigraphy of the Late Devonian Black
Shales on the Cratonal Margin of the Applachian
Geosyncline. Unpublished Ph.D. Dissertation, Boston
University. 194 pp.
Holland, B., S. Stridsberg, and J. Bergstrom. 1978. Confirmation
of the reconstruction of Aptychopsis. Lethaia 1 1 : 144.
Internationa] Commission on Zoological Nomenclature. 1985.
International Code of Zoological Nomenclature , Third
Edition. LIniversity of California Press, Berkeley. 338 pp.
Kanie, Y. 1982. Cretaceous tetragonitid ammonite jaws: A
comparison with modern Nautilus jaws. Transactions and
Proceedings of the Palaeontological Society of Japan. N.S.
125:239-258. pis. 39, 40.
Kayser, E. 1882. Protokoll der Decembersitzung der Deutschen
Geologischen Gesellschaft, Vortrag iiber Anaptychen. Zeitschrift
der Deutschen Geologischen Gesellschaft 34:818-819.
Kues, B.S. 1983. Cephalopod aptychi from Los Moyos
Limestone, Madera Group (M. Pennsylvanian), near
Albuquerque, New Mexico. New Mexico Geology 5:78-80.
Landman, N.H., and R.A. Davis. 1988. Jaw and crop preserved in
an orthoconic nautiloid cephalopod from the Bear Gulch
Limestone (Mississippian, Montana). Memoir of the New
Mexico Bureau of Mines and Mineral Resources, 44:103-107.
Lehmann, U. 1970. Lias Anaptychen als Kieferelemente
(Ammonoidea). Paldontologische Zeitschrift 44:25-31.
Lehmann, U. 1971. New aspects in ammonite biology.
Proceedings of the North American Paleontological
Convention 1:1251-1269.
Lehmann, U. 1972. Aptychen als Kieferelemente der Ammoniten.
Paldontologische Zeitschrift 46:34-48.
Lehmann, U. 1975. Uber Nahrung und Ernahrungsweise von
Ammoniten. Paldontologische Zeitschrift 49: 187-195.
Lehmann, U. 1978. Uber den Kieferapparat von Ammoniten der
Gattung Parkinsonia. Mitteilungen aus dem Geologisch-
Paldontologischen Institut der Universitat Hamburg 48:79-
84, pi. 9.
Lehmann, U. 1979. The jaws and radula of the Jurassic ammonite
Dactylioceras. Palaeontology 22( 1 ):265-27 1 , pi. 27.
Lehmann, U. 1981. Ammonite jaw apparatus and soft parts. In:
The Ammonoidea. Edited by M.R. House and J.R. Senior.
Systematics Association Special Volume. Academic Press,
New York 18:275-287.
Lewis, T.L. 1988. Late Devonian and Early Mississippian Distal
Basin-Margin Sedimentation of Northern Ohio. Ohio
Journal of Science 88(1 ):23-39.
Lineback, J.A. 1970. Stratigraphy of the New Albany Shale in
Indiana. Bulletin of the Indiana Geological Survey, Dept, of
Natural Resources, Bloomington, 44:67 pp.
Lowenstam, H.A., W. Traub, and S. Weiner. 1984. Nautilus hard
parts: a study of the mineral and organic constituents.
Paleobiology 10(2):268-279.
Mapes, R.H. 1987. Upper Paleozoic cephalopod mandibles:
Frequency of occurrence, modes of preservation, and
paleoecological implications. Journal of Paleontology
61:521-538."
Matern, H. 1931. Oberdevonische Anaptychen in situ und iiber
die Erhaltung von Chitin-Substanzen. Senckenbergiana
13:160-167.
Moore, R.C., and PC. Sylvester-Bradley. 1957a. Taxonomy and
nomenclature of aptychi, pp. L465-L471. In: Treatise on
Invertebrate Paleontology , Volume L, Mollusca 4,
Cephalopoda, Ammonoidea. Edited by R.C. Moore. Geological
Society of America and University of Kansas Press, Lawrence.
Moore, R.C., and P.C. Sylvester-Bradley. 1957b. Second
supplemental application: Application for a ruling by the
International Commission directing that the classification
and nomenclature of ammonoid aptychi (Class Cephalopoda)
be in terms of “parataxa.” Bulletin of Zoological
Nomenclature 15:35-70.
1991
Late Devonian Cephalopod Aptychi
71
Morton, N. 1981 Aptychi: the myth of the ammonite operculum.
Lethaia 14:57-61.
Nixon, M. 1988. The buccal mass of fossil and recent
Cephalopoda, pp. 103-121. In: The Mollusca , Volume 12:
Paleontology and Neontology of Cephalopods. Edited by
K.M. Wilber. Academic Press, San Diego.
Pratt, B.R. 1987. The phyllocarid crustacean Montecaris from the
Devonian of Western Canada. Canadian Journal of Earth
Sciences 24:1267-1269.
Prosser, C.S. 1912. The Devonian and Mississippian formations
of Northeastern Ohio. Bulletin of the Ohio Geological
Survey, Fourth Series, Columbus, Ohio. 15:574 pp.
Ruedemann, R. 1916. Account of some new or little-known
species of fossils, mostly from the Paleozoic of New York.
Bulletin, New York State Museum, 1 89:7- 1 1 2.
Ruedemann, R. 1934. Paleozoic Plankton of North America.
Geological Society of America Memoir, 2:57-6 1 .
Saunders, W.B., and E.S. Richardson, Jr. 1979. Middle
Pennsylvannian (Desmoinesean) cephalopoda of the Mazon
Creek fauna. Northeastern Illinois, pp. 333-359. In: M.H.
Nitecki (ed. ). Mazon Creek Fossils. Academic Press, New
York.
Saunders, W.B., and C. Spinosa. 1974. Unusual fossil cephalopod
jaws from Nevada ( abst .). Abstracts with Programs,
Cordilleran Section Annual Meeting, Geological Society of
America. 6(3):248.
Schwimmer, B.A. 1988. Stratigraphic Distribution of
Brachiopods and Pelecypods in the Upper Devonian
(Famennian) Chagrin Shale in the Cuyahoga River Valley,
Northeast Ohio. Unpublished Masters Thesis, Kent State
University. 134 pp.
Schwimmer, B.A., and R. M. Feldmann. 1990. Stratigraphic
Distribution of Brachiopods and Bivalves in the Upper
Devonian (Famennian) Chagrin Shale in the Cuyahoga River
Valley, Northeast Ohio. Kirtlandia 45:7-31.
Stridsberg, S. 1984. Aptychopsid plates: Jaw elements or
protective operculum? Lethaia 17:93-98.
Szmuc, E.J. 1970a. The Devonian System, pp. 9-21. In: Guide to
the Geology of Northeastern Ohio. Edited by PO. Banks and
R.M. Feldmann. Northern Ohio Geological Society.
Szmuc. E.J. 1970b. The Mississippian System, pp. 23-67. In: Guide
to the Geology of Northeastern Ohio. Edited by PO. Banks
and R.M. Feldmann. Northern Ohio Geological Society.
Tanabe, K. 1983. The jaw apparatuses of Cretaceous
desmoceratid ammonites. Palaeontology 26(3):677-686.
Tanabe, K., H. Hirano, and Y. Kanie. 1980. The jaw apparatus of
Scalarites mihoensis, a Late Cretaceous ammonite, pp. 1 59-
165, pi. 20. Professor Saburo Kanno Memorial Volume.
Tanabe, K., and Y. Fukuda. 1987. The jaw apparatus of the
Cretaceous ammonite Reesidites. Lethaia 20:41-48.
Thompson, E.H., E.L. Yochelson, and R.H. Flower. 1980.
Aptychi from the Pennsylvanian of West Texas. Journal of
Paleontology 54:903-909.
Trauth, F. 1927. Aptychenstudien I: Uber die Aptychen im
Allgemeinen. Annalen des Naturhistorische Museums in
Wien 41:171-259, 1 pi.
Trauth, F. 1928. Aptychenstudien II: die Aptychen der
Oberkreide. Annalen des Naturhistorische Museums in Wien
42:121-193, pis. 2-4.
Trauth. F. 1930. Aptychenstudien 1 1 1 - V : Nachtrag zu den
“Aptychen im Allgemeinen;” Nachtrag zu den “Aptychen
der Oberkreide;” Die Aptychen des Dogger. Annalen des
Naturhistorischen Museums in Wien 44:329-41 1, pis. 3-5.
Trauth, F. 1931. Aptychenstudien VI- VII: Zweiter Nachtrag zu
den “Aptychen im Allgemeinen;” die Aptychen des Malm
und Unterkreide. Annalen des Naturhistorischen Museums in
Wien 45: 17-136, I pi.
Trauth, F. 1935. Die Aptychen des Palaozoikums. Jahrbuch
Preussische Geologische Landesanstalt 55( 1 934 ):44-83, pi.
1-12.
Trauth, F. 1936. Aptychenstudien VIII: die Laevillamellaptychi
des Oberjura und der Unterkreide. Annalen des
Naturhistorischen Museums in Wien 47: 127- 145, pi. 3.
Turek, V. 1978. Biological and stratigraphical significance of the
Silurian nautiloid Aptychopsis. Lethaia 1 1 : 127- 1 38.
Whitfield, R.P 1882. Species from the Huron shales. Annals of
the New York Academy of Sciences 2:217.
Woodward, H. 1865. On the discovery of a new genus of
Cirripedia in the Wenlock Limestone and Shale of Dudley.
Quarterly Journal of the Geological Society of London,
21(8 0:486-489, pi. 14.
Woodward, H. 1882. On a series of phyllopod crustacean shields
from the Upper Devonian of the Eifel. The Geological
Magazine, new series. Decade 2, 9:385-390, pi. 9.
Woodward, H. 1885a. On a new species of Helminthochiton from
the Upper Bala (Silurian) of Girvan, Ayrshire. The
Geological Magazine, new series. Decade 3, 2:352-355.
Woodward, H. 1885b. On some Paleozoic phyllopod-shields, and
on Nebalia and its allies. The Geological Magazine, new
series. Decade 3, 2:345-352, pi. 9.
Yochelson, E.L. 1971. The Permian nautiloid mandible
Rhynchoteuthis kaibabensis reinterpreted as a
polyplacophoran. Journal of Paleontology 45:130-133.
Yochelson, E.L. 1983. A Devonian aptychus (Cephalopoda) from
Alabama. Journal of Paleontology 57: 1 24- 1 27.
■NATURAL HISTORY'
Published by
The Cleveland Museum of Natural History
1 Wade Oval Drive, University Circle
Cleveland, Ohio 44106-1767
Production by
Mary Szelagowski and Wendy Ryan
'ELAND, OHIO
NUMBER 47
KIRTLANDIA
Archaeology
Late Woodland Fortifications in Northern Ohio:
The Greenwood Village Site 3
Stephanie J. Belovich and David S. Brose
•NATURAL HISTORY*
KIRTLANDIA
The Scientific Publication of The Cleveland Museum of Natural History
David S. Brose and
Joseph T. Hannibal, Editors
Brief History and Purpose
Kirtlandia, a publication of The Cleveland Museum of Natural History, is named in honor of Jared
Potter Kirtland, a noted nineteenth-century naturalist who lived in the Cleveland, Ohio area. It began
publication in 1967 and is a continuation of the earlier series Scientific Publications volumes 1 to 10
(1928-1950), and new series volumes I to 4 (1962-1965). Supported by the Kirtlandia Society of The
Cleveland Museum of Natural History, Kirtlandia is devoted to the publication of scientific papers in
the various fields of inquiry within the Museum’s sphere of interest: Cultural and Physical
Anthropology; Archaeology; Botany; Geology; Paleobotany; Invertebrate and Vertebrate
Paleontology; Systematic Ecology; and Invertebrate and Vertebrate Zoology. Issues will vary from
single monographs to collections of short papers, review articles, and brief research notes.
Kirtlandia is abstracted in Biological Abstracts and indexed in Bibliography and Index of Geology and
Zoological Record.
Associate Editors
James K. Bissell, The Cleveland Museum of Natural History
Bruce Latimer, The Cleveland Museum of Natural History
Martin Rosenberg, Case Western Reserve University
Sonja Teraguehi, The Cleveland Museum of Natural History
Editorial Advisory Board
Rodney Feldmann, Kent State University
Michael C. Hansen, Ohio Geological Survey
Richard Meindl. Kent State University
G. Michael Pratt, Heidelberg University
David H. Stansbery, Ohio State University
Frederick H. Utech, Carnegie Museum of Natural History
Ed Voss, University of Michigan
Andrew M. White, John Carroll University
Kirtlandia No. 47
ISSN 0075-6245
© 1992 by The Cleveland Museum of Natural History
Cleveland, Ohio
Copies of Kirtlandia, and many issues of the Scientific Publications series of
The Cleveland Museum of Natural History, are available for sale.
Write to: Library, The Cleveland Museum of Natural History, 1 Wade Oval Drive,
Cleveland, Ohio 44106-1767 for a current price list.
KIRTLANDIA
The Cleveland Museum of Natural History
September 1992 Number 47:3-23
Late Woodland Fortifications in Northern Ohio:
The Greenwood Village Site
Stephanie J. Belovich and David S. Brose
Department of Archaeology
The Cleveland Museum of Natural History
l Wade Oval Drive, University Circle
Cleveland, Ohio 44106-1767
Abstract
Since the late nineteenth century all of the archaeological sites located on steep bluffs
overlooking the major rivers of northern Ohio have been considered fortified villages. The
belief that such sites were defensive was given support by the results of some excavations
which also demonstrated a Late Woodland temporal placement. Recent excavation at one
such site, the Greenwood Village site, demonstrated that its still visible earthworks are not
defensive but used construction techniques similar to those at Southern Ohio Middle
Woodland ceremonial earthworks. Carbon- 14 and thermoluminescence dates place the
earthworks’ construction and ceremonial use of the Greenwood Village site plateau between
A.D. 460 and A.D. 1040. Critical review of several similar hilltop enclosures shows many
are neither defensive works, nor are they all from late prehistoric periods.
Introduction
Since initial documentation, the presence of prehistoric
villages and accompanying fortifications have come to be
the hallmark of the Whittlesey Tradition in northern Ohio.
Recorded by Colonel Charles Whittlesey in the nineteenth
century, these sites, high atop steeply sided bluffs
overlooking major river valleys and their tributary streams,
are characterized by the presence of earthen walls and
ditches. During the years since Whittlesey’s survey many
of these sites have been destroyed through urban
expansion, vandalism or both.
Excavation on some Whittlesey Tradition sites
(Greenman, 1935a, 1935b, 1937; Morgan and Ellis, 1943;
Murphy, 1971a) demonstrated similarities in topographic
location and ceramic assemblage, but failed to recover data
sufficient to discuss spatial organization, subsistence
patterning, or resource procurement and scheduling. With
little evidence these sites were assumed to represent large
fortified villages whose inhabitants focused on maize
agriculture. It was not until 1969, when Brose conducted
extensive excavations at the Whittlesey Tradition South
Park site (33Cu8), that a detailed picture of the tradition
emerged. Three stratigraphically distinct occupations
dating from A.D. 1000 to A.D. 1640 were identified at
South Park (Brose, 1973). Evidence for increasing
reliance upon maize-bean-squash agriculture and
4
Belov ich and Brose
No. 47
eventual year-round village occupation was found.
Information on domestic structures, site patterning, ditch
and earthen fortifications and seasonal scheduling was also
obtained (Brose, 1992). In essence Brose confirmed the
validity of the characterization of the Whittlesey Tradition.
Before Brose's detailed contribution to understanding the
Whittlesey Tradition, the archaeological community had
begun to take for granted the defining characteristics of that
tradition. Essentially, all of the sites located by Colonel
Whittlesey, and for that matter any northern Ohio site on a
steeply sided bluff with ditches, embankments or other
earthworks which could be considered fortifications, were
uncritically assigned to the Whittlesey Tradition. The issue
we address in this paper then, is whether such assignments
are valid. In so doing we will report on recent excavations con-
ducted by Belovich and Brose (1983) and Belovich (T985a,
1985b) at the Greenwood Village site (33Su92: Whittlesey
Fort No. 5). We also will discuss the development of Middle
Woodland earthworks into Late Woodland fortifications.
The Environment
The Greenwood Village site is situated on the western
edge of the glaciated Appalachian Plateau Province. This
area was covered by ice during several Pleistocene stages,
most recently during the Wisconsinan (Belovich, 1985b;
White and Totten, 1982). After the formation and draining
of a series of pro-glacial lakes, the middle-lower Cuyahoga
River flowed north into Lake Erie. The subsequent
formation of numerous tributary streams caused the eastern
and southern highlands along the river valley to become
extensively dissected by small seasonal streams (Belovich,
1985b; Brose et al., 1981; Williams, 1949). Archaeological
sites located within the Cuyahoga River Valley are often
encountered on the summits of these steeply sided
plateaux. Severe erosion has undoubtedly reduced site sizes
(Belovich, 1985b).
The soils of the uplands are derived from underlying
sandstones and shales, or are of glaciofluvia! origin
(Williams, 1949; U.S. Department of Agriculture, 1971).
The narrow floodplain of the Cuyahoga River was built by
successive layering of flood-water sediments. These soils
are highly fertile, and the floodplain has always been
considered prime agricultural land (Brose et ah, 1981, p. 9).
Since 8,000 B.C. the Cuyahoga River Valley has been
covered by a nearly continuous deciduous forest canopy.
Within this mixed mesophytic forest, several specific floral
associations can be identified (Braun, 1950; Gordon, 1969;
Williams, 1949). Along the uplands farthest from the river,
soils are dry and oak-hickory (and at one time, chestnut)
associations predominated. On the slopes, lower elevations
and moister areas, beech-maple associations are found.
Elm-ash forests inhabit the bottomlands, while com-
munities of hemlock and pines thrive along the cool, damp
f.ortr/rtnt
o/ //tf tt/frirnf rt/rf/ ttwr/'A
o/ /Ar ft//* ft// />,/ ra//rt/
f if// Z> OZf/ff
.. " S
s
/>;, >ttn t.
f /nr ftt,
D
// r////f/&tf
\‘r,/ //// ,/
+ o
\r«/
• +
/// ///// -' n'f /■•/
,V".i -f
D
/Jrrr/sf f//t
\ + ,
f {
Sorf/t/it/t/
V □
V. ,iv
'/■ +
< 0
V
X»/f.
□ , Y"/Z
+•
/ffffU/f f
tifff/t
□ /httr/stfZf
'± y»&
Yf/ ft '
O if « vy <
0 ' yoy
Yf/f/A
\ I/tini/i/on
0M
Pttrtoffr d
timihgtu. d
■ , firm
l/r/w/a
f's////r//r/// a// ,v .
- - /'V//V// monarf. «•
f f ! ///• r r // / forts
CO f't/r/tr sc
O o //v v r /r > > r t/r/f/r s
S rt/ tr ,)//// //' S/tr ///tf/
FIGURE 1. Map of Whittlesey forts located in northeast Ohio.
Taken from Ancient Earth Forts of the Cuyahoga Valley, Ohio
1871.
ravines of small streams (Williams, 1949).
Detailed reconstructions of past faunal communities
within the region are presently impossible. Early
ethnohistoric accounts provide only general descriptions
with little specific information. The soils of the uplands,
the rich alluvial bottoms, and the temperate climate would
have facilitated even technologically limited horticulture,
while the mosaic forest patterns would have yielded a large
variety of floral and faunal resources more than adequate to
meet the subsistence needs of the aboriginal populations.
1992
Late Woodland Fortifications in Northern Ohio
5
Historical Background: 1847-1982
The Greenwood Village site (33Su92) was recorded in
1847 by Colonel Charles Whittlesey, then Ohio's Surveyor
General (Whittlesey, 1850, 1867, 1871). This site was the
fifth of eleven “forts” Whittlesey located and reported
along both banks of the Cuyahoga River (Figure 1 ). All but
one of these “forts” sit atop steeply sided and relatively
isolated plateaux. Each site also has at least one earthen
embankment and ditch crossing the single, usually narrow,
level access to the site. At Greenwood Village (Fort No. 5)
and Fort No.’s 1, 3 and 11, more than one ditch/
embankment were noted. Gateways and/or mounds were
also recorded.
Noting the topographic location, Whittlesey not only
referred to these sites as “forts” but described each with
reference to its potential defense against assault. The
persistence of opinion that these sites were, in fact,
defensive has been primarily based on their topographic
location and single, narrow level access. Previous
excavations at several of these sites (Tuttle Hill [Fort 3]:
Greenman, 1937; and South Park: Brose, 1973, 1992) did
reveal fortified late prehistoric villages, giving apparent
confirmation to this idea. Greenwood Village (Fort No. 5)
with its location and extensive earthworks was thought to
be another late prehistoric fortified village.
First mentioned in Descriptions of Ancient Works in
Ohio (1850) Whittlesey called Greenwood Village a
“fortification,” and described three wells, two mounds, one
gateway, and five ditch/embankment lines. The relevant
portions of the text read as follows:
The engineers who selected the site of this
fortification, understood very well the art of turning
natural advantages to good account. Why they did
not embrace in their plan the whole of the level space
within the crest of the bluff, is not easily explained,
unless we presume that their numbers were few, and
not sufficient to defend the whole. On all sides, the
gullies are from eighty to one hundred and ten feet
deep, worn, by running water, into the blue and
yellow hard pan that here forms the bluffs of the
valley of the Cuyahoga River. The earth is as steep as
it will stand; and, in fact, is subject to slides, that lie
in terraces, resembling platforms, made by art.
Before the ground was cultivated, the ditches are said
by Milton Arthur, Esq., the owner of the land, to
have been so deep that a man standing in them could
not look over the wall.
... At the north end of the ditch of the inner wall,
at the neck, there was a narrow space left as a
passage into the work, but none in the outer wall.
There are low mounds at m,m. The approach is along
a sharp ridge called a “hog’s back,” merely broad
FIGURE 2. 1850 Whittlesey map of Fort No. 5: the Greenwood
Village site (33Si<92). Taken from Descriptions of Ancient Works
in Ohio 1850.
enough for a single road track, for the distance of
thirty rods, and the sides are as steep as any part of
the bluffs adjacent . . .
It is not very evident why a few rods of ground
were cut off by lines at the south-west angle, nor why
part of the ditch was made on the inside on the north
and west.
It is very remarkable that, while all the works in
northern Ohio are of a military character, there are no
evidences of attacks by a foe, or of the destruction or
overthrow of any of them . . .
(Whittlesey, 1850, p. 17-18)
6
Belovich and Brose
No. 47
FIGURE 3. 1871 Whittlesey map of Fort No. 5: the Greenwood Village site (33Su92). Taken from Ancient Earth Forts of the Cuyahoga
Valley, Ohio 1871.
The accompanying published map (Figure 2) was
evidently drawn by Whittlesey himself. It is inaccurate in
some aspects: (1) The narrow “hog’s back” does not run
due east as shown, but is oriented east by northeast. 2) The
long axis of the plateau lies in a northwesterly direction. 3)
The Cuyahoga River and Ohio Erie Canal lie more to the
northwest of the site). However, it is the most accurate of
all the maps he published of the site (Belovich 1985b).
Unfortunately the inaccuracies of the 1850 map reappear in
Whittlesey’s later maps.
Nevertheless, this map indicates that Whittlesey
observed two mounds along the northern edge of the site. It
is also clear that four of the embankments had exterior
ditches; one was without a ditch (no doubt due to its
placement along the bluff edge), and an interior ditch,
associated with an embankment which cuts the plateau in
half indicated to Whittlesey (1867) “. . . a state of siege.”
Whittlesey carefully depicted the “. . . narrow space left as a
passage into the work . . .” (Whittlesey, 1850, p. 17-18) as a
small western extension perpendicular to the innermost
eastern embankment which crossed the narrow hogback
leading to the site. One gateway was also present.
Subsequent maps, published by Whittlesey in 1867 and
1871 (Figure 3), were accompanied by essentially similar
descriptions. However the maps themselves had changed
(Belovich, 1985b).
On the 1867 map both mounds grew somewhat larger
while the earthwork extension became smaller. By 1871
the western earthwork extension had disappeared
completely from the map. In addition, not only had the
Ohio Erie Canal and the Cuyahoga River changed
orientation, but the north arrow had joined them. The shape
of the plateau, and the distances between the embankments
also changed. The wells vanished but pits appeared, and
the mounds not only moved but increased from two to
three. These revisions seem so extensive that it is clear that
Whittlesey not only never revisited the site, but he even
failed to revisit his own notes apparently drawing the later
maps from memory, if he drew them at all.
In his 1871 publication Whittlesey brought together all
the data he collected during his survey on aboriginal
occupation sites in the Cuyahoga Valley. Due to the ditches
and embankments at many of these sites, Whittlesey called
them “forts” and assigned to them a numerical order. At
this time the Greenwood Village site was called “Fort No.
5” (Belovich, 1985b, p. 17).
Information for the site (apparently gleaned from
Whittlesey’s reports) was eventually placed in the official
state files. At this time the site was given two names. It was
called “Arthur Fort,” after the landowner mentioned by
Whittlesey in the 1850 and 1867 publications. The second
name, “Whittlesey Fort No. 5,” was taken from
Whittlesey’s 1871 publication. Finally the site was
assigned the number 33Sul0.
Not until 121 years later, in 1971, was the Greenwood
Village site (Fort No. 5) revisited. It was investigated as
1992
Late Woodland Fortifications in Northern Ohio
7
(33Su92), Columbia Road Village (33Su87), and Stanford Knoll
(33Su99) sites.
part of a systematic environmentally stratified survey of
northeast Ohio (Brose, 1976a, 1976b). One of the sampling
quadrants chosen for testing comprised three plateaux
situated along the western bluffs of the Cuyahoga River, on
properties then owned by the Greenwood Village
Development Corporation (Brose, 1976a; Belovich and
Brose, 1983; Belovich, 1985a). The plateaux were labeled
according to cardinal directions and it was atop Greenwood
Village West that testing was performed (Brose, 1976a;
Belovich and Brose, 1983). Three 5 x 5 ft squares were
excavated but only one yielded cultural material. This unit,
number three, was located “about 2 m south of the northern
edge of the narrow plateau neck; to the north of the path”
(Belovich and Brose, 1983, p. 15). Cultural material
consisted of burned and unburned bone belonging to deer,
bear, rodents, birds and fish, and fragments of shell.
Worked, polished and incised bone was recovered as were
several sandstone abraders, bifacial and unifacial tools, and
a Madison projectile point fragment. Shell and grit
FIGURE 5. 1983/1984 Map of the Greenwood Village site
(33Su92) with test unit locations indicated.
tempered ceramics included the types Fairport Harbor
Cordmarked, Fairport Filleted, South Park Notched, and
Tuttle Hill Notched. A few Wellsberg Simple-Stamped
sherds were also recovered. These materials suggested that
the site was a small agricultural village occupied from fall
through spring, ca. A.D. 1300-A.D. 1400 (Brose, 1976a),
relatively late in the Whittlesey Tradition (Brose, 1973).
It wasn't until 1979, when Brose and Belovich were
performing a survey of the newly formed Cuyahoga Valley
National Recreation Area for the National Park Service,
that there was occasion to compare the 1971 field notes
with the Whittlesey maps. At that time it was concluded
that the plateau Brose had labeled Greenwood Village West
was the same site Whittlesey had called Fort No. 5 (Brose
et al., 1981). Field crews were sent to test the plateau. At
that time dense vegetation prevented observation of any
earthworks. Of the 10 shovel tests excavated, only three
yielded artifacts, none of which were diagnostic.
Nevertheless based upon 1971 data and field observations
in 1979, it was determined that large portions of the site
remained intact (Brose et al., 1981). The site was registered
with the state and given a late prehistoric Whittlesey
temporal placement, with a suggested date of A.D. 1250-
1400.
In 1982, interests again focused on the Greenwood
Village site. Further archaeological research had revealed
that the site had two different USNM numbers and two
Belovich and Brose
No. 47
different names. Eventually it was decided that the
Greenwood Village site (33Su92), would become the
official designation for the prehistoric village first
discovered by Col. Charles Whittlesey in 1847 (c.f.
Belovich. 1985b, p. 29-30). At this time it was realized that
the site, unique in its existence, preservation, and the
presence of visible ditches and earthen embankments, was
the only Whittlesey Fort remaining along this portion of
the Cuyahoga River Valley. Cognizant of the fact that 1979
testing failed to recover diagnostic artifacts and that the site
was now suffering from erosion, the National Park Service
permitted two years of phased test excavations at the site.
The Greenwood Village Site
The Greenwood Village site (33Su92), in Northfield
Township, extends over an entire plateau overlooking the
east side of the Cuyahoga River Valley, in Summit County,
Ohio (Figures 4 and 5). On all sides of the site there is
evidence of erosion. It is likely that the plateau, as well as
the site, was much larger when it was occupied by
prehistoric groups.
Fimited reconnaissance in the spring of 1983 (Belovich
and Brose, 1983; Belovich, 1985b) revealed that the only
clearly identifiable ditch/embankment line was the inner-
most one along the southern point of the plateau (Figure 2
and Figure 6). At the western extremity of this
embankment another embankment joined it to form an “L”
extending northwestward along the west side of the
plateau. While the short portion of the “L” appeared to be
present at the time of reconnaissance (Figure 6), its
northwestward extent was not visible. All other
ditch/embankment lines identified by Whittlesey had
disappeared due to erosion or farming (Belovich, 1985b).
While no mounds were observed, several undulations
across the eastern narrow hogback were considered to be
the possible ditch/embankment lines recorded by
Whittlesey (Figure 2). These lines, as well as the interior and
exterior of the earthwork, were tested (c.f. Belovich, 1985b).
Portions of the “highway” reported by Whittlesey to run
through the site and down the western slope to the Canal
were present, though the path had varied and was much
narrowed (Belovich, 1985b, p. 33). It has been established
that this path was once an 1838 roadway (Brose et al.,
1981). This path was later used as a jogging and
motorcycle trail. It has now been converted by the National
Park Service into a pedestrian nature trail.
To test the site function and chronology suggested by
Brose (1976a) and the accuracy of the reports by
Whittlesey (1850, 1867, 1871), four areas of the site were
identified for intensive investigation: 1 ) the eastern, narrow
neck entrance to the site; 2) the area “inside the lines
(where] the ground was much richer than without them”
(Whittlesey, 1871, p. 13); 3) the inner ditch/embankment
FIGURE 6. Southern ditch/embankment line, view east, March,
1992. (Note arrow indicating day pack on embankment against
tree, and tree stump in ditch ).
line whose placement resulted in “. . . reducing the fortified
area to about one-half the space . . .” (Whittlesey, 1867, p.
38); and 4) the combined areas outside the enclosure,
where Whittlesey noted that the ground was not as “rich”
as within the enclosure (Whittlesey, 1871, p. 13), and the
mounds Whittlesey indicated along the northeastern end of
the plateau (Belovich, 1985b).
The Features
Fifteen test units were excavated during the 1983 field
season (Figure 5). After the first season it became clear that
the Greenwood Village assemblage being recovered from
these sub-surface excavations was not related to the
Whittlesey period, but rather to earlier assemblages
(Belovich and Brose, 1982; Brose, 1983; Brose and Scarry,
1976). In light of this, Belovich (1985b, p. 4-5)
hypothesized that not all hilltop enclosures were defensive
works, nor did they all date to the late prehistoric
Whittlesey period. The nine test units excavated during the
1984 field season (Figure 5) were placed to obtain
additional data on the construction, function and age of the
earthen ditches and embankments, and to recover samples
suitable for chronometric dating techniques to clearly place
all structural and non-structural features in time (Belovich,
1985a, 1985b).
Excavations outside the western-most earthwork yielded
little cultural material. Soil development was minimal with
glaciolacustrine gravels at or near the surface (Belovich,
1985b). This confirmed Whittlesey’s observation that the area
outside the enclosure was not as “rich” as that within. Inside
the enclosure, the excavation of 18 test units (40m2)
revealed 15 cultural features. Most were shallow, basin-
shaped, rock-filled, fire pits (Belovich, 1985a, 1985b).
Feature 6, located within the enclosure, was a unique
pottery deposit containing fragments of a single, Fairport
Harbor Cordmarked var. Willoughby vessel (Figure 7).
1992
Late Woodland Fortifications in Northern Ohio
9
\ ■ M
FIGURE 7. Reconstructed Fairport Harbor Cordmarked var.
Willoughby pottery vessel from Feature 6. a, reconstructed vessel;
b, cross section of vessel rim showing large grit-tempering.
FIGURE 8. Feature 10, east profile.
Feature 6 was encountered 36 cm below the surface and
extended to a depth of 58 cm. First identified when
excavators observed pottery sherds standing on end, the
majority of recovered sherds were piled atop one another in
a manner consistent with the interpretation that the vessel
collapsed upon itself. The placement of the sherds in this
tight cluster suggested that some had been thrown or swept
onto the main collapsed pile. Slightly over 200 pottery
sherds were recovered from Feature 6 but not all of them
could be fitted onto the reconstructed vessel. It is likely that
before the vessel was buried some of its constituent sherds
may have been scattered. Feature 6 represents a vessel
possibly broken while sitting in a shallow pit or depression
on the ground’s surface. Five pieces of lithic debitage, one
worked piece of shale, some minute bone flecks and a
small charcoal sample were also recovered from Feature 6.
Feature 10, a meter east of Feature 6, was about 2 m in
diameter and nearly 1 m deep. Areas of dense charcoal
concentration and a discontinuous ring of burned orange-
red soil outlined the feature. The innermost areas were
composed of dark yellowish-brown silts and hard-packed
clayey silts (Belovich, 1985b, p. 55). Six layers were
identified within this great fire pit which yielded lithics,
burned soils, dense charcoal deposits, and nearly 300 kg
(657 lb) of firecracked rock (Figure 8). Samples of pottery
10 YR 4/4 SANDY-SILT WITH PEBBLES
B- 10 YR 3/3
H- 10 YR 2/2
C- 10 YR 3/2 SANDY-SILT WITH FOR
D- 10 YR 2/1 FOR AND CHARCOAL
F- 5 YR 3/4 SILTY- CLAY WITH PEBBLES
0- 10 YR 5/6 SANDY- SILT
HUMUS
M ROCK
Q GLACIAL TILL
0 TL SAMPLE
A C“ 14 SAMPLE
POST MOLD
20cm N
10
Belovich and Brose
No. 47
for thermoluminesence dating were recovered from the
third layer, and 17 carbon 14 samples mostly from the
fourth and fifth layers were collected from feature fill
(Belovich, 1985b). Feature 10 was devoid of any plant
food remains and yielded only exceedingly fragmentary
pieces of bone. Similar features were recorded at the Bugai
site (Halsey, 1976) and also at the early Late Woodland,
Lichliter site near Dayton, Ohio (Allman, 1957). Within the
enclosure no evidence for long-term occupation was found
(Belovich. 1985a). There were no domestic structures,
storage pits, midden or village debris encountered. Even in
features, floral and faunal remains were rare or absent. This
is in marked contrast to the faunal evidence gathered by
Brose in 1971 from areas further east.
Excavations begun in 1983 at one of the undulations
crossing the eastern, narrow hogback entrance revealed it
to be the innermost of the two lines of ditch and
embankments Whittlesey recorded in that area (Figures 2
and 5). The aboriginal ditch had been excavated 30 cm into
the glacial gravels. These excavated gravels were then
mixed with sandy-silts to create a 15-20 cm-thick bed upon
which was placed a 20-25 cm-thick layer of shale,
sandstone, and igneous rock which varied in size from
small pebbles to large cobbles and flagstones (Belovich,
1985b). This prepared rock pavement (Feature 3),
encountered some 75 cm below the surface, formed a
revetment or foundation for the earthen embankment
above. Cobbles and flagstones were noted along the eastern
portion of the embankment, some at increasing depths
within the ditch. This suggests that they were used to face
the eastern, outer slope of the embankment and retard
erosion. A refilled post hole measuring 12.5 cm in diameter
was observed within Feature 3 (Belovich, 1985a). This post
hole may have been part of some superstructure
constructed in front of the earthen embankment to prevent
erosion of the embankment into the adjacent ditch. No
clear evidence of a palisade was observed.
The prepared rock pavement appeared to have been less
than 2.5 m wide, roughly corresponding to the width of the
base of the innermost embankment. If not for a truly
serendipitous event this may have marked the end of our
earthwork investigations. One excavated test unit, located
on the earthen embankment to avoid the roots of several
large trees, failed to find the western end of the prepared
rock pavement. Two other 1 m x 1 m units excavated
adjacent to the west wall of that unit also showed the
pavement continuing. At that point it was decided that
excavation would proceed in a “westerly direction until we
encountered the western termination of Feature 3 or the
Cuyahoga River; whichever came first” (Belovich, 1985b,
p. 66). Feature 3 ended at a distance of 9.1 m, nearly 30 ft
(Figure 9). This prepared rock pavement was constructed
in the same manner as that found beneath the embankment.
It lay upon a sandy-silt gravel matrix, and maintained a
thickness of 20-25 cm until it tapered rather dramatically at
its western margin (c.f. Belovich, 1985b, Figures 12, 13
and Plate 6). Like the embankment pavement, this
extension was constructed of cobbles and tabular stones
laid flat. The stones were from 5-30 cm long and 8-20 cm
thick (Belovich, 1985b). Most of the rocks were shale or
sandstone, no doubt gathered from the stream beds below.
Careful reexamination of Whittlesey’s 1850 map and notes
indicated that at the innermost wall “. . . there was a narrow
space left as a passage into the work, but none in the outer
wall.” This short western extension to the embankment
(Figure 2) was undoubtedly the western extension Belovich
1992
Late Woodland Fortifications in Northern Ohio
11
HUMUS
SILTS
ROCK PAVEMENT
SILTY-SAND AND GRAVEL MATRIX
GLACSO-LACUSTRINE GRAVELS
NORTH
FIGURE 10. Hypothesized reconstruction of Feature 3 ditch and embankment.
had uncovered. Its graphic representation on the 1850 map,
however, contrasts sharply with its length as documented
by Belovich. At the time Whittlesey recorded the site, a dirt
road leading to the Cuyahoga River crossed the
ditch/embankment line at this spot. This road, as well as
farming on the plateau, had probably reduced or covered
much of the original western extension which was only
revealed by these excavations.
An hypothesized reconstruction of this extension and
prepared rock foundation under the embankment, with a
rock facing found within the ditch to retard erosion
(Belovich, 1985a, 1985b), is presented in Figure 10. Based
on these data it was concluded that . . the ditch and
embankment constructions [were] non-defensive
earthworks, and the western extension represent[ed] a
pedestrian ramp . . .” leading into the main area on the
plateau (Belovich, 1985b, p. 180).
The few post molds that were identified at the site were
located within the enclosure. All of them have been
assigned to the Woodland occupation. Three post molds
were located within Feature 10, the great fire pit, and one
adjacent to it. Three additional post molds were identified
about a meter to the southeast of Feature 10. None of these
post molds yielded cultural material. Stratigraphically and
horizontally they all are associated with Feature 10 and
may have functioned as a windbreak and/or a drying rack
for Feature 10 (Belovich, 1985).
One post mold was identified within eastern profile of
Feature 3, the prepared rock pavement. This was the only
post mold identified in association with Feature 3 and it
may have been part of some type of superstructure
constructed in front of the earthen embankment to afford
some protection against erosion of the embankment in to
the adjacent ditch (Belovich, 1985, p. 42).
Five other post molds were located in the central part of
the plateau near Features 15 and 16. Based upon limited
ceramic evidence these two fire pits and their associated
post molds have been assigned to the early Late Woodland
occupation at the site (Belovich, 1985). No cultural
material was recovered from the post molds, but it is
thought that they may have served as windbreaks, spits, or
drying racks for these fire pits.
The Artifacts
The ceramics from the 1983 and 1984 excavation of
Greenwood Village site are predominantly coarse grit-
tempered wares (97%) although limestone tempering is
present (Belovich, 1985b, Table 7). Vessels are simple sub-
conoidal in shape, with weak shoulders, straight necks, and
flat to slightly rounded lips (Figure 7 and Figure ll:a-b).
The average thickness for rim sherds was 0.74 cm
(measured 2 cm below the lip), while the average thickness
for body sherds was 0.84 cm (Belovich, 1985b). The
ceramics are cordmarked or fabric-marked (Figure ll:i).
Cordmarking is medium to coarse and usually vertical
although other orientations are known (Figure 1 1 :c-d and
h). Decoration is rare and when present consists of two to
four thin, weakly incised, discontinuous lines over the
cordmarking at the neck (Figure 1 l:e-g) (Belovich, 1985b).
Similar ceramics found in early Late Woodland sites in
Ohio, Indiana, Michigan, New York, western Pennsylvania,
and Kentucky have been variously called Fairport Harbor
12
Belovich and Brose
No. 47
1992
Late Woodland Fortifications in Northern Ohio
13
Cordmarked (Brose, 1983, 1985), Cuyahoga Cordmarked
(Brose and Scarry, 1976), Jack’s Reef Corded (Ritchie,
1965), Wayne Cordmarked (Fitting, 1964), Mixter
Cordmarked (Shane, 1967), Watson and Mahoning
Cordmarked (Mayer-Oakes, 1955), Newtown Cordmarked
(Oehler, 1950; McMichael, 1984), or Peters Cordmarked
(Prufer, 1967).
While displaying similarities to ceramics from southern
Ohio and western Pennsylvania, Greenwood Village
ceramics may be more closely related to other types in
northeast Ohio (Belovich, 1985b). Some grit-tempered
ceramics (including the reconstructed vessel from Feature
6) appear to be examples of the undecorated grit-tempered
type Faiiport Harbor Cordmarked var. Willoughby (Brose,
1983, 1985). This type is characterized by an unmodified,
sub-conoidal vessel form with flat, slightly everted rims.
Surface treatment consists of cordmarking to the lip.
Fairport Harbor Cordmarked may date to as early as A.D.
900 (Brose, 1985). Greenwood Village ceramics also
possess traits ascribed to the type Cuyahoga Cordmarked
(Brose and Scarry, 1976; Brose, 1985). This type has
globular to semi-conoidal vessels with straight to
moderately outcurved rims, and “massive grit-tempering”
(Brose, 1985, p. 52; Brose and Scarry, 1976). Like the type
Fairport Harbor Cordmarked, Cuyahoga Cordmarked
vessels are usually vertically cordmarked to the lip.
Cuyahoga Cordmarked is distinguished from Fairport
Harbor Cordmarked by the presence of interior
cordmarking on the former (Brose and Scarry, 1976; Brose,
1985). The type Cuyahoga Cordmarked has been dated to
between A.D. 900 and A.D. 1200 although it may be an
early Late Woodland type which might occur to as early as
A.D. 600 (Brose, 1985; Brose and Scarry, 1976, p. 185).
Greenwood Village rims display weakly incised
horizontal lines (Figure ll:e-g), which can easily be seen as
antecedent to the types McFate Incised (Mayer-Oakes, 1955;
Murphy, 1971b; Johnson, 1976) and Reeve Horizontal
(Belovich, 1985b; Fitting, 1964; Murphy, 1971b; Johnson,
1976). Some Greenwood Village sherds were similar to
some ceramics recovered from the Mixter site (Shane,
1967a; Prufer and Shane, 1976), the Lyman site (Muiphy,
1971c), the Fairport Harbor site (Morgan and Ellis, 1943;
Muiphy, 1971a), and to the East Wall site, the Reeves site,
and some ceramics from the earliest levels at the South Park
site (Belovich, 1985b; Brose, 1973, 1985, 1992). All of these
FIGURE 11. Selected grit-tempered rim and body sherds from the
Greenwood Village site (33Su92). a, straight cordmarked rim
sherd ; b, two slightly everted cordmarked rim sherds fitted to one
body sherd; c-d, horizontal cordmarked interior and vertical
cordmarked exterior of a single body sherd ; e-g,five cordmarked
rim sherds and one body sherd with two to four horizontal
incisions; h, overlapping cordmarked body sherd; i, four fabric-
marked body sherds.
sites date between A.D. 800 and A.D. 1200.
Comparisons can also be made to earlier time periods
and other geographic areas. Watson and Mahoning
Cordmarked ceramics from the Upper Ohio Valley (Mayer-
Oakes, 1955; Maslowski, 1973; Hemmings, 1984) have
granite or limestone tempering and straight necks with
flattened or slightly rounded lips. They have vertical
exterior cordmarking to the lip, horizontal interior
cordmarking, and even fabric-marking (Mayer-Oakes,
1955, p. 191-195). Greenwood Village ceramic assemblage
displays all of these characteristics (Belovich, 1985b).
Watson and Mahoning ceramics are considered to be
Middle Woodland to early Late Woodland utilitarian wares.
The Greenwood Village ceramics also show similarities
to the cordmarked grit-tempered wares assignable to the
Newtown Focus of southwestern Ohio, southeastern
Indiana and portions of Kentucky (Oehler, 1950, 1992;
Seeman, 1980; McMichael, 1984). Newtown Cordmarked
ceramics have vertical to slightly flaring rims with
vertically oriented cordmarked exteriors. Some of these
sherds display horizontally cordmarked interiors (Oehler,
1950; McMichael, 1984).
Like the variety of Peters Cordmarked identified at
Chesser Cave in southeastern Ohio (Prufer, 1967, p. 11-
12), Greenwood ceramics have straight necks and
flattened lips. They are grit and limestone tempered,
cordmarked, and, as Prufer stated, exhibit “shoddily and
weakly incised thin lines applied to the neck" (Prufer,
1967, p. 12). Prufer noted the similarity between Peters
Phase ceramics and ceramics from sites grouped together
in the Cole Complex (Baby and Potter, 1965). Potter
(1968, p. 62) suggested a placement of A.D. 800 to A.D.
1300 for the Cole Complex but, as Belovich noted, “Just
as Prufer related Peters to Cole, so Potter related Cole to
the intrusive Mound Culture of south and central Ohio,
suggesting both were the same” (Belovich, 1985b. p. 165-
166). Using chronology and artifact styles Halsey
disagrees and suggested . . the Cole Complex post-
dates the Mills Phase [his name for Intrusive Mound],
although it could possibly be partially contemporaneous”
(Halsey, 1976, p. 525). Halsey then fixed the Mills Phase
somewhere between A.D. 500 and A.D. 1100 (Halsey,
1976, p. 441-446, p. 519-526).
Seven samples for radiometric analysis, four
thermoluminescence samples and three carbon- 14 samples
(Table 1 ) were sent to three different laboratories; DICARB
Radioisotope, Beta Analytic, and Alpha Analytic, by two
different submitters; The National Park Service and The
Cleveland Museum of Natural History (Belovich, 1985b).
Five of these samples were collected from layers within
Feature 10, the great fire pit; one from near Feature 10; and
one sample from the prepared rock pavement (Feature 3).
Unfortunately, for technical reasons, the TL sample
14
Belovich and Brose
No. 47
TABLE 1. Radiocarbon and Thermoluminescence Dates for Greenwood Village site (33Su92).
Provenience
Material
Date In
Years B.P.
Corrected*
Date In
Years A.D.
Corrected*
Range Date
In Years A.D.
Laboratory
A0 19-53
TU #10 and 20
Fea. #10, Sec. B
77cm below surface
sample # 1
pottery
grit-tempered
cordmarked
body sherd
11801 180
A.D. 770
A.D. 590-
A.D. 950
ALPHA- 1484
A0 19-54
TU #19 and 20
Fea. #10, Sec. B
82cm below surface
sample #1
pottery
grit-tempered
cordmarked
body sherd
12001210
A.D. 750
A.D. 540-
A.D. 960*
ALPHA-1485
A0 19- 16
TU #19 and 20
Fea. #10, Sec. C
82cm below surface
sample #1
charcoal
12601 80
A.D. 750
A.D. 650-
A.D. 850*
BETA- 107 15
A0 19-22
TU # 1 0 and 20
Fea. #10, Sec. D
100cm below surface
sample #5
charcoal
1500160
A.D. 525*
A.D. 460-
A.D. 590
BETA- 107 16
A0 19-24
TU #19 and 20
Fea. #10, Sec. D
101cm below surface
sample #7
charcoal
1020170
A.D. 985*
A.D. 930-
A.D. 1040*
DICARB-3072
A0 10-23
TU #10, N 1/2
west side of unit.
Level 3: 26-36cm
limestone-tempered
rim sherd (once
thought to be
shell-tempered)
11101 170
A.D. 840
A.D. 670-
A.D. 1010
ALPHA- 1486
A016-36
TU #16, Fea. #3
Level 9: 1 10- 130cm
1 1cm N, 167cm W,
122cm below surface
cordmarked
grit-tempered
body sherd
9.29cm x 636cm
1 . 1cm
No date obtainable, no fine grain,
fraction present
ALPHA- 1487
(ALPHA-1487) from the pavement was undatable.
Nevertheless, the cordmarked grit-tempered ceramics
from the pavement, identical to ceramics dated
elsewhere on the site, lend confidence to assigning the
earthworks’ construction to the early Late Woodland
Period. Based upon all the dates run, the major period of
occupation and earthwork construction at the Greenwood
Village site occurred some time between A.D. 460 and
A.D. 1040, with the most likely period being between
A.D. 600 and A.D. 800. With the dates from the nearby
Stanford Knoll (Lee, 1986) and Columbia Road Village
(Belovich and Brose, 1982) sites, we have identified the
temporal placement of coarse, grit- and limestone-
tempered, cordmarked ceramics within the Cuyahoga
River Valley. All radiometric information for these three
northeastern Ohio sites is displayed in graphic form in
Figure 12.
The lithic assemblage from Greenwood Village also
contains elements common to early Late Woodland sites:
shale discs and knives, and a projectile point cluster
variously called Chesser Notched, Lowe Flared Base, or
what Lee (1986) called Anthony Side Notched (Figures
1992
Late Woodland Fortifications in Northern Ohio
15
GREENWOOD
STANFORD KNOLL VILLAGE
33Su99 33Su92
COLUMBIA ROAD
VILLAGE
33Su87
A. D. 1 250
1 000
750
500
250
A.D. 1
250
500
750
1 000
B. C. 1 250
□ □
i
i
FIGURE 12. Radiocarbon dates for the Greenwood Village (33Su92), Columbia Road Village (33Su87), and Stanford Knoll (33Su99) sites.
13 and 14) (Belovich, 1985b). The shale discs and
knives, important identifying characteristics of early
Late Woodland lithic assemblages, have been referred to
as slate discs by Halsey (1976) and others (Allman
1957, 1961). This is clearly incorrect since shale, not
slate, occurs in Ohio and slate is almost non-existent in
lower Michigan as well. Small, crudely flaked points,
with diamond or humpbacked cross sections generally
referred to as “fishspears” (Prufer, 1967; Converse,
1984) were also common (Figure 13). Pitted stones, a
net sinker, hammerstones, a small shale hoe, as well as
other groundstone tool fragments were recovered.
Finally a modest amount of firecracked rock totalling
395.47 kg (871.86 lb) was collected and analyzed. The
lithic assemblage needs little more comment than
reiterating that its constituent “fishspears,”
Chesser/Anthony Side Notched/Lowe Flared projectile
points and shale knives and discs consistently occur at
early Late Woodland sites across Ohio (Schatz, 1957;
Allman, 1957, 1961; Converse, 1963, 1984; Prufer,
1965, 1967, 1981; Halsey, 1976; Oplinger, 1981;
Ormerod, 1983).
An Archaic component at the site is represented by a
LeCroy bifurcate and a crudely stemmed projectile point
(Belovich, 1985). Both were surface finds. No features
could be assigned to the Archaic period.
Comparative Discussion
The 1983-1984 Greenwood Village site excavations
failed to yield any evidence for long-term village
occupation. Almost all features excavated were smali
firepits and there were no domestic structures, storage pits
or deep midden zones of any kind (Belovich, 1983, 1985a,
1985b). Even such large, deep, fire pits, as Feature 10 are
essentially devoid of plant food or faunal remains. Like
the other small firepits on the site. Feature 10 represents a
singular utilization event (Belovich, 1985b).
The Greenwood Village site shares traits in common
with other early Late Woodland sites. The Bugai site from
Saginaw County, Michigan, belongs to the Wayne
Mortuary Complex. The site is dated between A.D. 500
and A.D. 1100 (Halsey, 1976, p. 445, 473). At the Bugai
site a basin-shaped fire pit five feet in diameter and about
a foot deep (Halsey, 1976, p. 475, 480-481) yielded fired
ocher, charcoal, and two chipped discs of what Halsey
(1976, p. 506) called slate. Directly below this fire pit
were two bundle burials, one associated with another
“slate" disc. Additional “slate" discs were recovered from
other features at the Bugai site. While even a pottery
concentration was excavated at Bugai (Halsey, 1976, p.
483), there were no earthworks.
The Lichliter site is a late Middle Woodland or early
Late Woodland site near Dayton. Ohio (Allman, 1957).
16
Belovich and Brose
No. 47
1992
Late Woodland Fortifications in Northern Ohio
17
Four houses were located and the one completely
excavated circular house measured 48 ft in diameter. At its
north end was a nearly square fire pit about 4 ft on a side
and almost 2 ft deep. Allman observed heavy con-
centrations of ash and charcoal, noting that “. . . the sides
were quite deeply burnt . . .” (Allman, 1957, p. 60). Faunal
remains were absent but Allman did recover many stones,
. . some over 6 inches in size . . .” and one pottery sherd.
Allman concluded . . that this might have been a
ceremonial fireplace . . .” (Allman, 1957, p. 60). This
feature is similar to Feature 10 at Greenwood Village.
Allman also recorded that over 55 whole and broken
“slate” discs were recovered from the Lichliter site. From
his published photographs, the projectile points from the
site are similar to Hopewell, Chesser, and Anthony Side
Notched points (Allman, 1957, p. 62).
Rock pavements and foundations like those at
Greenwood Village are not unique. The Pollock Works,
first recorded by Squier and Davis, were recently
investigated by Riordan (1982). The site is composed of an
earthen wall, cut by three gateways, stretching across a
plateau situated 30 ft above Massies Creek in Greene
County, Ohio (Riordan, 1982). Outside and some distance
from these earthen constructions, are three semicircular
earthen walls separated by three more gateways. Riordan’s
excavations uncovered three limestone rock pavements at
each gateway which he believes functioned to protect the
earthworks from erosion and may also have served as
walkways to and from the plateau’s interior. Initial
radiometric dates placed the Pollock Works between 230
and 400 B.C. suggesting a late Adena affiliation (Riordan,
1982, p. 15-16). Additional radiocarbon dates obtained
from wood charcoal recovered from on and beneath the
Pollock earthworks clearly indicate that the site’s major
period of construction was during the first and second
centuries A.D. (Riordan, 1986). Riordan considers the
earlier dates aberrant (Riordan, 1986).
The Fort Ancient site also has Middle Woodland earthen
embankments and rock pavements (Essenpreis and
Moseley, 1984). As at Greenwood Village, embankments
outline the plateau, and ditches are found inside the
embankment. Essenpreis and Moseley demonstrated that
stone was used extensively; as facing for the outer
embankment slopes, short walls to retard erosion down the
FIGURE 13. Selected projectile points and scrapers from the
Greenwood Village site (33Su92). a-c, Anthony Side Notched
projectile points; d, hafted scraper made from an Anthony Side
Notched projectile point; e, corner-notched projectile point
similar to Lowe Flared Base type; f, thick, reworked projectile
point similar to Lowe Flared Base or Chesser Notched types; g-o,
projectile points with diamond-shaped or humpbacked cross
sections similar to "fishspears”; p-q, Chesser Notched projectile
points; r, hafted end scraper; s, thumbnail scraper.
plateau slopes, and as cores or foundations for some of the
embankments. Rock pavements, noted as early as 1 940 by
Morgan, seemed to have served as roadways between
various structures within the enclosure (Morgan, 1946,
1970; Morgan and Thomas, 1950; Essenpreis and Moseley
1984) . The numerous gateways, absence of structures
preventing travel through the gateways, and the place-
ment of ditches inside the embankments all suggest a
non-defensive function. Miami Fort, Fort Glenford, Fort
Hill, Windsor Fort, Turner, Spruce Hill, Big Creek and
Indian Fort Mountain have all yielded evidence of earthen
embankments and rock pavements (Fischer, 1974, p. 87).
Fort Hill ( 3 3 C u 1 ) was first surveyed by Charles
Whittlesey prior to 1888. The site is located on a plateau
200 ft above the confluence of the East and West Branches
of the Rocky River. At the eastern tip of this plateau
Whittlesey identified three parallel 15 ft-wide walls with
1 1 ft-wide ditches along their western faces (Whittlesey,
1888). At the time of his survey the maximum distance
from the top of an embankment to base of a ditch was 4 ft
(Whittlesey, 1888). Eight days of shovel testing in late
1985 failed to recover any evidence for a late prehistoric
Whittlesey period village occupation (Lee and Belovich,
1985) . There was no midden deposit of any kind, and
shovel testing across the entire plateau recovered only two
flint flakes (one utilized). Test excavations of the most
visible of the three embankments (all have been severely
disturbed by erosion, plowing or dirt bike traffic) indicated
that the feature was clearly of cultural origin. Artifacts
recovered from the embankment consisted of one flint
flake and one quartz flake. Despite the absence of
diagnostic ceramics, the similarities of Fort Hill to
Greenwood Village (steeply sided plateau [twice the
elevation of Greenwood Village], complete with a series of
ditches and embankments, and exhibiting no midden
deposit and very few artifacts) led Belovich to consider this
as evidence for an early Late Woodland placement for Fort
Hill (Belovich, 1985b).
Windsor Fort (33Ab3) is located on a peninsular plateau
90 ft above the west bank of Phelps Creek in Ashtabula
County. The western, landward side of the plateau is
crossed by two parallel ditches and embankments, about
150 ft long. The ditches lie adjacent to the west side of
each embankment. The deepest, outermost ditch gives its
associated embankment an apparent 7 ft height; however,
when viewed from the plateau interior this wall appears
only about 3 ft high. Test excavations conducted at
Windsor Fort by Lee, as part of an archaeological
reconnaissance of Ashtabula County, failed to uncover any
evidence of deep midden deposits or village debris.
Artifacts recovered were limited to small amounts of lithic
debitage and three cordmarked grit-tempered body sherds.
These limited data suggested to Lee that Windsor Fort was
18
Belovich and Brose
No. 47
FIGURE 14. Selected shale knives and tools from the Greenwood Village site (33Su92). a, lanceolate shale knife; b-c, unifacial shale
scrapers; d-e, shale discs; f, shale “hoe. ”
similar to Greenwood Village and more likely to date to the
early Late Woodland period rather than the late prehistoric
Whittlesey period (Lee, 1987).
Sites with artifact assemblages similar to Greenwood
Village have also been noted in other regions of Ohio. The
Water Plant site (33Fr 155) in central Ohio, sits atop a high
bluff overlooking Big Walnut Creek, a tributary of the
Scioto River (Dancey, 1988). Dancey identified eleven
discrete debris clusters which he interpreted as distinct
household units (Dancey, 1988, p. 223) although no
structural evidence for houses was observed. The artifact
assemblage from this site is characterized by Chesser
1992
Late Woodland Fortifications in Northern Ohio
19
Notched projectile points and cordmarked grit-tempered
pottery. As at Greenwood Village, ceramic decoration on
Water Plant site sherds is rare. When present it consists
of poorly incised lines nearly parallel to the rim (Dancey,
1988). Dancey states that the ceramics resemble the type
Newtown Cordmarked (c.f. McMichael, 1984). The
lithics, ceramics, and radiocarbon dates place this small
village site firmly within the early Late Woodland
period.
Another site in northeast Ohio, the Columbia Road
Village site (33Su87), is situated atop a high ridge along
the west side of the Cuyahoga River Valley (Belovich and
Brose, 1982). The plateau is surrounded on three sides by
very steep ravines with small intermittent streams. The
fourth (south) side trails into a narrow neck which
descends to another deep ravine. While no earthworks were
encountered at this site, four features were excavated.
Feature 2 was a random deposit of broken sherds. Feature 1
was rock-lined fire pit with a double post mold (Feature 3)
adjacent to it. Feature 4 was partially excavated and
appeared to be a fire pit. The large number of artifacts
recovered from these excavations included projectile
points, point fragments, tools, chert and shale flakes, and
lithic debitage. The points and point fragments are
morphologically similar to Chesser Notched points
reported from the Ohio Valley (Mayer-Oakes, 1955, p. 83b;
Prufer, 1967, p. 54, fig. 5, n-z). Some of these points also
resemble Lamoka, Jack’s Reef Corner Notched, and Lowe
Flared point types (Ritchie, 1971; Reidhead and Limp,
1974). Grindingstones, hammerstones, a pipe bowl and
shale gorget fragment complete the stone artifact inventory
(Belovich and Brose, 1982).
All of the sherds recovered from the Columbia Road
Village site were grit-tempered and some were
cordmarked. Rims were cordmarked, with the cordmarking
usually oriented vertically. In some instances cordmarking
occurred on the inside of the rim. Most rims were straight
sided or slightly excurvate while lips were usually flat.
Decoration was rare; only one everted rim with finger-nail
punctates was collected (Belovich and Brose, 1982).
Pottery from the Columbia Road Village site resembles
the limestone-tempered Watson Cordmarked ceramics from
the Upper Ohio Valley (Mayer-Oakes, 1955, p. 193-195)
and the limestone-tempered cordmarked sherds from pre-
Fort Ancient contexts at the Haag Site (ca. A.D. 700 - ca.
A.D. 950) in southeastern Indiana (Reidhead and Limp,
1974, p. 9). Similar grit-tempered ceramics were found at
the Gillie Rock Shelter near Twinsburg, Ohio, dated
between A.D. 200 and A.D. 700 (Bernhardt, 1973). While
Columbia Road Village site ceramics appear similar to
Peters Cordmarked from Chesser Cave (Prufer and
McKenzie, 1966, p. 60), Prufer stated that the ceramics
from 33Su87 are somewhat thicker, are better fired, and
had finer fabrics used in their manufacture than those he
had found at Chesser (personal communication, 1985). By
far the strongest resemblances are with the grit-tempered
cordmarked and plain ceramics from nearby Hale Farm
(Brose, 1985), and the Boston Ledges Rock Shelters A and
B (Brose and Scarry, 1976). Though absolute dates are not
available for either of those sites, their assemblages suggest
a chronological placement between A.D. 600 and A.D.
1000. Charcoal samples from Feature 1 at the Columbia
Road Village site (Figure 12) date between A.D. 780 and
A.D. 1005 (A.D. 930, 1090 + 85 B.P., DICARB - 2605 and
A.D. 850, 1160 + 60/-50 B.P., DICARB - 2606) (c.f. for a
discussion of pertinent issues surrounding DICARB dates
Belovich, 1985b; Belovich and Brose, 1982). These dates
plus the regional comparison of the site assemblage clearly
indicate an early Late Woodland chronological placement
(Figure 12).
The multicomponent Stanford Knoll site (33Su99),
located east of the Cuyahoga River had Early Woodland
(950 B.C.), Middle Woodland (A.D. 170-300) and early
Late Woodland (A.D. 600) manifestations (Lee, 1986).
While no structures were encountered, cordmarked, grit-
tempered ceramics dating to the early Late Woodland
period were recovered from Feature 6, a 53 cm diameter,
26 cm-deep pit. These well-fired ceramics were from 8.4-
14.0 mm thick. Shallow cord impressions were “fairly
widely spaced (average 1.6 mm)” (Lee, 1986, p. 40). They
are virtually indistinguishable from ceramics recovered
from the Greenwood Village site and could be lost within
its ceramic assemblage. A thermoluminescence (Figure 12)
date obtained for one of these sherds was A.D. 620 ( 1330 +
150, B.P. ALPHA-2621) (Lee, 1986). A carbon sample
(Figure 12) from Feature 11 placed a rim sherd, virtually
indistinguishable from ceramics from the Greenwood
Village site, as early as A.D. 235 (1780 + 60 B.P, BETA-
15012) (Lee, 1986). Two Racoon Side Notched projectile
points recovered from Feature 6 suggest a very early Late
Woodland date (Ritchie, 1965, p. 228, p. 258-260).
Belovich (1985b) clearly noted that early Late
Woodland phases, complexes, and cultures were poorly
defined both in terms of their characteristics and
chronology. In an attempt to bring order to these data for
southern Ohio, Seeman (1980), noting the similarity
between sites originally assigned to the Cole Complex,
Newtown Focus, and Peters Phase, suggested they be
grouped together. He then gave them the descriptive, yet
cumbersome name of Central Ohio Valley early Late
Woodland and suggested a time span of A.D. 500 to A.D.
800 (Seeman, 1980, p. 17). Seeman also suggested that
such Central Ohio Valley early Late Woodland cultures
would extend from southwestern Ohio (Peters
Cordmarked), northeastward to Pittsburgh, Pennsylvania
(Watson Cordmarked) (Seeman, 1980, p. 16-17). In
20
Belovich and Brose
No. 47
Seeman’s scheme, the distinct Intrusive Mound Culture
spans the period A.D. 800 to A.D. 1000 (Seeman, 1980, p.
16-18). As Belovich noted, under this scheme the Cole
Complex also remains,
“. . . but it is now represented only by those sites
not already subsumed (with Newtown and Peters)
under the nomen Central Ohio Valley early late (sic)
Woodland. Due to their late temporal placement (post
A.D. 900) and certain shared characters (guilloche
designs) these Cole Complex sites are considered “to
represent a late Late Woodland complex at least
partially contemporaneous with Fort Ancient
components to the south” (Seeman 1980:18). Finally,
Seeman considers the Chesser Phase to be so poorly
defined that not much more than its latest temporal
placement (A.D. 1200) is suggested.”
(Belovich, 1985b, p. 166)
It bears repeating, that the Greenwood Village ceramics
share traits with Mixter Cordmarked, Cuyahoga
Cordmarked, and Fairport Flarbor Cordmarked (var.
Willoughby) ceramics which may date from as early as
A.D. 600. Few Middle Woodland sites from which a local
antecedent for the Greenwood Village ceramics can be
identified are known for northeast Ohio. Belovich (1985b)
suggested that the early Late Woodland Greenwood Village
ceramics will in fact be found to develop out of indigenous
Middle Woodland styles such as those types seen at
Stanford Knoll (Lee, 1986) and at other late Hopewellian
sites such as Everett Knoll (Brose, 1974), the North Benton
Mound (McGrath, 1945) and the Huntington Road Village
site (Evangilista and Dodd, personal communication).
Conclusions
It now seems clear that the late prehistoric Whittlesey
materials Brose reported in 1973 (Belovich and Brose,
1983) cannot be associated with the construction of any
earthwork reported by Whittlesey at Greenwood Village.
Nor do they represent any significant occupation within the
enclosure. Similar late Whittlesey materials were recovered
in 1985 from contexts associated with ephemeral
occupations outside the enclosure and approximately
107 m further east along the narrow hogback entrance to
the site, where there is some geophysical evidence for an
embankment not recorded by Whittlesey (c.f. Belovich,
1985b).
These investigations place the occupations responsible
for the earthwork constructions at a pre-Whittlesey date
between A.D. 600 and A.D. 800 (Belovich, 1985a, 1985b;
Belovich and Brose, 1983). The absence of structures, deep
midden deposits, or storage pits with significant faunal or
floral remains, and the numerous single-use fire pits,
suggest limited domestic activities which seem
functionally, structurally, and even chronologically more
closely related to preceding Middle Woodland phenomena
than to the subsequent late prehistoric Whittlesey
Tradition. Indeed, there is unambiguous evidence that
burial mound construction continued into the 14th century
A.D. (Belovich, 1986). The Greenwood Village earthworks
and mounds further suggest that much of the cultural
resources and energy expended at the site was focused on
the construction activities themselves. Greenwood Village
can therefore be seen as a continuation of traditional
mound and earthwork construction first seen in the Early
Woodland period. Construction techniques were little
changed, despite the apparent social and organizational
changes presumed to have occurred by the Late Woodland
period. The apparent absence of exotic artifacts, mound
burials, or cremations or other internments at Greenwood
Village may very well be attributed to the erosion of the
mounds once located along the northwestern edge of the
site, and the poor preservation of bone (Belovich, 1985b).
Their absence at other early Late Woodland sites may also
be an artifact of preservation or it may be due to the
collapse of the elaborate social ceremonialism of the
Middle Woodland period.
Additional research in northern Ohio may clarify the
nature of this transition from horticultural Middle
Woodland, to more agricultural early Late Woodland
villages, to the economic dependence on cultigens
characteristic of the late prehistoric Whittlesey Tradition.
While further excavations may be needed before any
definitive statements can be made regarding social function
at such sites as Greenwood Village or Fort Ancient, these
investigations have clearly supported Belovich’s (1985, p.
191) hypothesis that not all hilltop enclosures are defensive
works, nor can it be assumed that all of Whittlesey’s
“forts” date to the late prehistoric Whittlesey period .
Acknowledgements
We wish to acknowledge the help and contributions of
Hiram College, Cuyahoga Community College, The
Cleveland Museum of Natural History and especially the
William Bingham Foundation. Without their financial
support this research would not have been possible.
Thanks must also go to F. Calabrese and Mark Lynott,
of the Midwest Archaeological Center and to Lewis Albert,
Superintendent, the Cuyahoga Valley National Recreation
Area for granting permission to test the site. We also wish
to mention the immeasurable contributions of Catherine
Hovey, Sean Coughlin, Alfred M. Lee, David McNickle,
Robert R Mensforth, Laura Pompignano, Judy Placko, and
Mark Ohlberg in the areas of organization, fieldwork,
photography, drawing and data analysis. Wendy Wasman
and Dan Flocke lent us assistance, as did a skilled crew of
1992
Late Woodland Fortifications in Northern Ohio
21
field and laboratory volunteers who shared in this project.
The hard work, enthusiasm, and good humor from all of
these folk were important contributions to the success of
this project. The authors would also like to thank the
reviewers for their helpful suggestions.
Finally, the senior author extends a very special thanks
to the late Patricia S. Essenpreis for providing the impetus
to write this paper by requesting the inclusion of my
Masters research in her and Robert Riordan's Woodland
Enclosures Symposium at the 1986 meetings of the Society
for American Archaeology. Thanks Pat: Miss you.
References
Allman, John C. 1957. A new Late Woodland culture for Ohio.
The Lichliter Village site near Dayton. Ohio Archaeologist
7(2):59-68.
Allman, John C. 1961. Late Woodland stone discs from Preble
County, Ohio. Ohio Archaeologist 11(4): 136.
Baby, Raymond S., and Martha A. Potter. 1965. The Cole
Complex: A Preliminary Analysis of the Late Woodland
Ceramics in Ohio and Their Relationship to the Ohio
Hopewell Phase. Ohio Historical Society, Papers in
Archaeology 2. Columbus.
Belovich, Stephanie J. 1985a. Evaluative testing at the
Greenwood Village Site (33Su92): 1984 season.
Archaeological Research Report 56. The Cleveland Museum
of Natural History, Cleveland.
Belovich, Stephanie J. 1985b. The Greenwood Village site
(33Su92): an early Late Woodland Site along the Cuyahoga
River. Unpublished M.A. thesis, Kent State University, Kent.
Belovich, Stephanie J. 1986. The Staas site (33Cu224): 1985
excavations. Archaeological Research Report 66. The
Cleveland Museum of Natural History, Cleveland.
Belovich, Stephanie J., and David S. Brose. 1982. Survey and
evaluative testing within the Cuyahoga Valley National
Recreation Area: Sites 33Su87 and 33Sul02.
Archaeological Research Report 40. The Cleveland Museum
of Natural History, Cleveland.
Belovich. Stephanie J., and David S. Brose. 1983. Evaluative
testing at the Greenwood Village site (33Su92).
Archaeological Research Report 47. The Cleveland Museum
of Natural History, Cleveland.
Bernhardt, Jack E. 1973. Gillie Rockshelter: A Late Woodland
phase in Summit County, Ohio. Unpublished M.A. thesis,
Kent State University, Kent.
Braun, E. Lucy. 1950. Deciduous Forests of Eastern North
America. Blakiston Co., Philadelphia.
Brose, David S. 1973. A preliminary analysis of recent
excavations at the South Park site, Cuyahoga County, Ohio.
Pennsylvania Archaeologist 43( 1 ):25-42.
Brose, David S. 1974. The Everett Knoll: a late Hopewellian
site in northeastern Ohio. Ohio Journal of Science 74(1):
36-46.
Brose, David S. 1976a. An initial summary of the Late Prehistoric
period in northeastern Ohio. In The Late Prehistory of the
Lake Erie Drainage Basin: A 1972 Symposium Revised,
edited by D. Brose, The Cleveland Museum of Natural
History, Cleveland.
Brose, David S. 1976b. Locational analysis in the prehistory of
northeast Ohio. In Cultural Change and Continuity: Essays
in Honor of James Bennett Griffin, edited by C. Cleland, p.
3-18. Academic Press, New York.
Brose, David S. 1983. A statistically dendritic analysis of
Whittlesey ceramics with an illustrated type-variety
classification. Manuscript on file in The Cleveland Museum
of Natural History Archaeological Archives.
Brose, David S. 1985. The prehistoric occupation of the Hale
Farm, Bath Township, Summit County, Ohio. Kirtlandia
41:35-61.
Brose, David S. 1992. The South Park Site and the Whittlesey
Tradition of Northeast Ohio. The Cleveland Museum of
Natural History, Cleveland.
Brose, David S., S. Belovich, M. Brooslin, R. Burns, Jr., J.
Hall. H. Haller, C. Pierce, and C. Ubbelohde. 1981.
Archaeological investigations in the Cuyahoga Valley
National Recreation Area. Archaeological Research
Report 30. The Cleveland Museum of Natural History,
Cleveland.
Brose, David S., and John F. Scarry. 1976. The Boston Ledges
Shelter: comparative spatial analyses of early Late Woodland
occupations in Summit County, Ohio. Midcontinental
Journal of Archaeology 1 (2): 1 79-228.
Converse, Robert N. 1963. Ohio flint types. Ohio Archaeologist
13(4):77- 1 2 1 .
Converse, Robert N. 1973. Ohio Flint Types. (Revised)
Archaeological Society of Ohio.
Converse, Robert N. 1984. Fishspear points. Ohio Archaeologist
34(4):5.
Dancy, William S. 1988. The community plan of an early Late
Woodland village in the middle Scioto River valley.
Midcontinental Journal of Archaeology 1 3(2):223-258.
Essenpreis, Patricia S., and Michael E. Moseley. 1984. Fort
Ancient: citadel or coliseum? Field Museum of Natural
History Bulletin 55(6):5-10, 20-26.
Fischer, Fred William. 1974. Early and Middle Woodland
settlement, subsistence, and population in the central Ohio
valley. Unpublished Ph.D. dissertation, Washington
University, St Louis.
Fitting, James E. 1964. Ceramic relationships of four Late
Woodland sites in northern Ohio. Wisconsin Archaeologist
45(4): 160-175.
Gordon, Robert B. 1969. The natural vegetation of Ohio in
pioneer days. Bulletin of the Ohio Biological Survey 3(2): 1 -
1 13. Columbus.
Greenman, Emerson F. 1935a. Excavation of the Reeve Village
site. Lake County, Ohio. Ohio State Archaeological and
Historical Quarterly 44( 1 ):2-64.
Greenman, Emerson F. 1935b. Seven prehistoric sites in northern
Ohio. The Ohio State Archaeological and Historical
Quarterly 44(2):220-237.
22
Belovich and Brose
No. 47
Greenman, Emerson F. 1937. Two prehistoric villages near
Cleveland, Ohio The Ohio State Archaeological and
Historical Quarterly 46(4):305-366.
Halsey, John Robert. 1976. The Bussinger site: A
multicomponent site in the Saginaw Valley of Michigan with
a review of early Late Woodland mortuary complexes in the
northeastern woodlands. Unpublished Ph.D. dissertation.
University of North Carolina, Chapel Hill.
Hemmings, E. Thomas. 1984. The Fairchance Site: Middle
Woodland settlement and subsistence in the upper Ohio
Valley. West Virginia Archaeologist 36( 1 ):3-5 1 .
Johnson, William C. 1976. The Late Woodland period in
northwestern Pennsylvania: a preliminary survey and
analysis for the symposium on the Late Woodland period in
the Lake Erie drainage basin. In The Late Prehistory of the
Lake Erie Drainage Basin: A 1972 Symposium Revised ,
edited by D. Brose, p. 48-75. The Cleveland Museum of
Natural History, Cleveland.
Lee, Alfred M. 1982. Archaeological Survey of Trumbull County.
Prepared by the Trumbull County Planning Commission and
The Cleveland Museum of Natural History, Cleveland, and
Warren, Ohio.
Lee, Alfred M. 1986. Excavations at the Stanford Knoll site,
Cuyahoga Valley National Recreation Area. Archaeological
Research Report 65. The Cleveland Museum of Natural
History, Cleveland.
Lee, Alfred M. 1986. Unpublished type descriptions for Anthony
Side Notched projectile points in author’s possession.
Lee, Alfred M. 1987. Archaeological reconnaissance in Ashtabula
County, Ohio. Archaeological Research Report 78. The
Cleveland Museum of Natural History, Cleveland.
Lee, Alfred M., and Stephanie J. Belovich. 1985. Field notes
and forms from test excavations in 1985 at Fort Hill
( 3 3Cu 1 ) on file at The Cleveland Museum of Natural
History, Cleveland.
MacGrath. Willis H. 1945. The North Benton Mound: a Hopewell
site in Ohio. American Antiquity 11(1 1:40-46.
Maslowski, R. F. 1973. An analysis of cordmarked Watson Ware.
Pennsylvania Archaeologist 43(2): 1-12.
Mayer-Oakes, William J. 1955. Prehistory of the upper Ohio
valley. Annals of Carnegie Museum 34. Anthropological
Series No. 2.
McMichael, Edward V. 1984. Appendix A: Type descriptions for
Newtown series ceramics. Posthumously published In The
Pyles Site (15MS28). a Newtown Village in Mason County
Kentucky, edited by J. Railey, p. 132-135. William S. Webb
Archaeological Society, Occasional Paper 1. Lexington.
(Publication of ms circulated in the 1960s).
Morgan, Richard G. 1946. Fort Ancient. Ohio State
Archaeological and Historical Society, Columbus.
Morgan, Richard G. 1970. Fort Ancient. The Ohio Historical
Society, Columbus.
Morgan, Richard G., and H. Holmes Ellis. 1943. The Fairport
Harbor site. The Ohio State Archaeological and Historical
Quarterly. 52(l):3-64.
Morgan, Richard G., and Edward S. Thomas. 1950. Fort Hill. The
Ohio State Archaeological and Historical Society, Columbus.
Murphy, James L. 1971a. The Fairport Harbor site (33La5) Lake
County, Ohio. Pennsylvania Archaeologist 41 (3):26-43.
Murphy, James L. 1971b. Whittlesey ceramic types. Ohio
Archaeologist 2 1 ( 1 ):298-303.
Murphy, James L. 1971c. The Lyman site (33La2), Lake County,
Ohio. Pennsylvania Archaeologist 41(3): 12-25.
Oehler, Charles. 1950. Turpin Indians. Cincinnati Museum of
Natural History Popular Publication Series 1. (Revised and
reprinted for The Journal of the Cincinnati Museum of
Natural History 23[2] 1973).
Oplinger, Jon. 1981. Wise Rockshelter: A multicomponent site in
Jackson County, Ohio. Kent State Research Papers in
Archaeology 2. Kent State University Press, Kent.
Ormerod, Dana E. 1983. White Rocks: A Woodland rockshelter
in Monroe County, Ohio. Kent State Research Papers in
Archaeology 4. Kent State University Press, Kent.
Potter, Martha A. 1968. Ohio’s Prehistoric Peoples. The Ohio
Historical Society, Columbus.
Potter, Martha A., and Edward S. Thomas. 1970. Fort Hill. The
Ohio Historical Society, Columbus.
Prufer, Olaf H. 1965. The McGraw Site: A Study in Hopewellian
Dynamics. Scientific Publications of The Cleveland
Museum of Natural History 4(1).
Prufer, Olaf H. 1967. Chesser Cave: a Late Woodland phase in
southeastern Ohio. In Studies in Ohio Archaeology, edited by
O. Prufer and D. McKenzie, p. 1-62. The Press of Western
Reserve University, Cleveland.
Prufer, Olaf H. 1981. Raven Rocks: A specialized Late Woodland
rockshelter occupation in Belmont County, Ohio. Kent State
Research Papers in Archaeology 1. Kent State University
Press, Kent.
Prufer, Olaf H., and Douglas H. McKenzie. 1966. Peters Cave:
two Woodland occupations in Ross County, Ohio. Ohio
Journal of Science 66(3):233-253.
Prufer, Olaf H., and Orrin C. Shane, III. 1976. The Portage-
Sandusky-Vermillion River region in Ohio. In The Late
Prehistory of the Lake Erie Drainage Basin: A 1972
Symposium Revised, edited by D. Brose, p. 283-304. The
Cleveland Museum of Natural History, Cleveland.
Riordan, Robert V. 1982. The Pollock Works: report on the 1981
field season. Wright State University Laboratory of
Anthropology, Reports in Anthropology 6.
Reidhead, Van A., and William F. Limp. 1974. The Haag site
( 1 2D 19): A preliminary report. Indiana Archaeological
Bulletin 1(1 ):4- 19.
Ritchie, William A. 1965. The Archaeology of New York State.
New York: Natural History Press.
Ritchie, William A. 1971. A Typology and Nomenclature for New
York Projectile Points. New York State Museum and Science
Service, Bulletin No. 384. Second edition, revised. Albany.
Schatz, John W. 1957. Late Woodland projectile types from
central Ohio. Ohio Archaeologist 7(4): 1 34- 1 35.
1992
Late Woodland Fortifications in Northern Ohio
23
Seeman, Mark F. 1980. A taxonomic review of southern Ohio
Late Woodland. Paper presented at the 25th Annual
Midwestern Archaeological Conference, Chicago, Illinois.
Seeman, Mark F. (editor). 1992. Cultural Variability in Context:
Woodland Settlements of the Mid-Ohio Valley. MCJA Special
Paper No. 7. The Kent State University Press, Kent.
Shane, Orrin C., III. 1967. The Mixter Site: a multicomponent
hunting station in Erie County, Ohio. In Studies in Ohio
Archaeology , edited by O. Prufer and D. McKenzie, p. 121-
186. The Press of Western Reserve University, Cleveland.
U. S. Department of Agriculture. 1971. Soils, the Seventh
Approximation (with supplements). Washington, D.C.
Department of Agriculture, Soil Survey Service.
White, George W„ with Stanley M. Totten. 1982. Glacial geology
of northeastern Ohio, with a chapter on Pleistocene beaches
and strandlines bordering Lake Erie. State of Ohio,
Department of Natural Resources, Division of Geological
Survey, Bulletin 68. Columbus.
Whittlesey, Col. Charles. 1850. Descriptions of ancient works in
Ohio. Smithsonian Contributions to Knowledge, Vol. II,
Article VII, pp. 17-18, Plate VII, No. A.
Whittlesey, Col. Charles. 1867. Early History of Cleveland,
Ohio. Fairbanks, Benedict and Company.
Whittlesey, Col. Charles. 1871. Ancient earth forts of the
Cuyahoga Valley, Ohio. Western Reserve and Northern Ohio
Historical Society Tracts: Number 5, Fairbanks, Benedict
and Company.
Whittlesey, Col. Charles. 1888. Ancient earthworks - northern
Ohio. Western Reserve and Northern Ohio Historical
Society, Tract Number 41, Vol. 11:34-35.
Williams, Arthur B. 1949. The native forests of Cuyahoga
County, Ohio. Scientific Publications of The Cleveland
Museum of Natural History 9( 1 ): 1 -90.
•NATURAL HISTORY*
Published by
The Cleveland Museum of Natural History
1 Wade Oval Drive
Cleveland, Ohio 44106-1767
Production by Melody Oakes
HECKMAN |X|
BINDERY INC. p|
OCT 98
Bound 'To -Pleas^1 fN*gR