3. Paradise
Regained and Lost Again: The Anderson Earthwork,
Ross County, Ohio (33RO551) By William H. Pickard
and Laurie A. Gray Pahdopony
Department of Anthropology, The Ohio State University
Columbus, Ohio
Introduction
Ross County, Ohio was at one time
perhaps the most archaeologically rich locale
in North America. From the time of Squier and
Davis
into the present, the once bountiful resource
found there had been explored, exploited and
abused to
a point where only a small portion remains
and little else could have been overlooked. Yet
as
recent as the 1970's, there were still significant
finds to be made. A case in point is the Anderson
Earthwork, a sizable square enclosure positioned
between Hopewell and Mound City, only discovered
in 1975. Although less spectacular than its
more famous neighbors, Anderson contained a section
of wall that was still mostly intact and original
to the time of construction. In 1993, limited
excavations were conducted at Anderson in the
face of a proposed
development on the site. As a result, valuable
insights were gained into building techniques
used by the builders that could possibly be
carried
over to the interpretation of other sites.
More importantly, however, sufficient material
for
radiocarbon
dating was recovered to provide evidence for
a Hopewellian origin of the Anderson Earthwork. History and Background
The Anderson
Earthwork is a slightly irregular low walled square
enclosure measuring between 235 and 287 meters
on a side located in extreme southern Union Township
in Ross County, about 9 km. west of Chillicothe.
It was named for the nearby unincorporated village
of Anderson. It is situated at about 700 feet
above sea level on a glacial outwash terrace that
is bounded on the west side by Bier's Run and
on the east by an unnamed seasonally intermittent
stream. Both streams flow into the North Fork
of Paint Creek about .7 km to the south. The high,
level nature of the terrace provides excellent
Visibility in all directions, and its placement
in a saddle between four sets of hills allows
easy passage from the Scioto valley to the northeast
and the Paint Creek and North Fork valleys to
the south and southwest (Anderson 1980: 31).
The earthwork was inadvertently discovered
in 1975 by vocational archaeologist, Jerrel
Anderson, while examining a 1938 Department
of Agriculture aerial photograph of Ross
County. He was studying the various local
earthworks, attempting to locate ground features
that had since been destroyed by modern farming
methods.
| |
|
click on image to enlarge |
| About
2 km east on the same terrace as the
Hopewell Earthwork, he noticed a large
square enclosure that to his knowledge had
not yet been described in the literature. Using
a combination of the USDA aerial,
modern infrared aerials and ground surveys,
Anderson was able to construct the plan
map seen in (Figure 1).
Although heavy cultivation had taken
its toll on the site, the map represents
a reasonable plan of the earthwork's
original geometry. This general configuration
was also confirmed by James Marshall's
1975 ground survey using theodolite and
steel tape (pers. comm.). |
|
|
On the ground, much
of the earthwork is traceable by
the bright red color of the soils
used in its construction. This effect
is
most apparent on the enclosure
south
of the B&O Railroad. North of the
railroad, plowing and a drainage
problem created
by the track bed had all but obliterated
the earthwork. The exception was
that part of the enclosure located
in the
three corner field formed by the
railroad, a fence row, and the
farmer's lane
(Figure 1 c). Here the wall is
visible as an elongated rise 10
to 12 m.
wide, .5 m. high with a total length
of about
120 m. This well preserved nature
of the wall was due to the fact
that it
had never been plowed. According
to the land owner, Rufus Riehle,
that
particular parcel had been an orchard
or in pasture for the several decades
he had been on the land (Anderson
1980: 31). This was confirmed in
1993 by
Mr. Riehle's cousin, Phillip Riehle,
who had managed the land since
1988. He stated that a tenant farmer
had
removed the fence row in 1989 to
create one large field but that
he had only
plowed the ground once. That was
the only time he had ever seen
it plowed.
Aside from this wall section no
other prehistoric constructs were
visible
on the terrace.
As is common with many earthworks,
there was a distinct lack of habitation
debris near the earthwork, although
a Middle Woodland site was said to
be located at the junction of Bier's
Run and North Fork, nearly a kilometer
to the south (Anderson 1980:32). A
local collector reported a few Archaic
and Early Woodland lithic scatters
along Bier's Run, but there were few
places in Ross County where such sites
could not be found. Additionally, the
1993 excavations would reveal a paucity
of artifacts in the subsurface.
1993 Excavations
By 1993, the terrace
north of the railroad had been purchased
by developers and slated to become
a housing sub-division known as Golf
View
Estates. The overall plan of the
project was such that what remained
of the
earthwork would be completely destroyed.
In an
ironic twist of fate, the centerline
of the main thoroughfare lay almost
directly on the centerline of the
east-west wall
of the square. Additionally, the
preserved section of the north-south
wall was
located on one of the first housing
lots to be
developed. A sorry state of affairs
indeed. In the summer of 1993, an
agreement was
reached with the developers to conduct
limited test excavations on the site
prior to the start of construction.
On July 27, two .75 x l0 m. trenches,
later
extended to 20 m., were laid out
across the wall in the area that had
once
been the three corner field. The
first unit,
trench A (Figure 1 a), cut across
the wall 20 meters north of the railroad
and the second unit, trench B, was
placed across the east-west wall
a
few meters
short of the old fence row (Figure
1 b). Excavation was by mattock,
square shovel and trowel. Both trenches
produced
straight forward profiles of similar
design, but differing somewhat in detail.
| |
|
click
on image to enlarge |
| It would appear that the construction
sequence began with the stripping
of the original vegetation and
top soils down to the B horizon.
At this stage a layer of fine angular
gravel in a silty clay matrix was
laid down (Figure 2e).
Generally, this layer was between
5 and 8 cm. thick and up to 7 m.
wide. with these gravels forming
a more or less continuous strata
across all profiles. Onto this
layer the embankment fill itself
was placed (Figure 2b). |
|
 Figure
2. Soil profile of trenches
placed along the walls of the
structure.
|
|
This component was
a bright red (Munsell 2.5YR5/8)
sandy clay composite up to 30 cm.
thick at the apex of the profiles
and tapering to 15-20 cm. at the
ends where it became indistinctly
mixed with the surrounding natural
soils. In both test units this
strata extended approximately 5
m. in both directions from the
apex. Above the embankment fill
was a 10-12 cm. thick sod zone
immediately below the surface (Figure
2a). Between the embankment fill
and sod zone were compacted turnings
of sod (Figure 2c), evidence of
the 1989 plowing described by Phillip
Riehle. It might be suggested that
the very low profile of the walls,
even without a long history of
plowing, was due to the moderately
incompetent nature of these embankment
fill soils. Further, it should
be noted that they do not occur
naturally on the terrace and would
have had to been carried in from
a remote location.
In total, seven possible features
were identified, a surprising number
considering the limited scope of
the excavation. Three were ruled
out as either tap roots or rodent
incursions. Of the four remaining,
three were post molds and the fourth
a basin-like gravel loading. Feature
#1 was the charred remains of a
large wooden post, located in trench
B directly beneath the apex of
the wall and originating below
the embankment fill. The post was
30 cm in diameter and cut to an
acute point that extended 50 cm
into the subsoils. Dee Anne Wymer
of Bloomsburg University identified
the wood as hickory (Carya), noting
that it had been subjected to a
rather intense burning. Charcoal
from this feature returned an AMS
date of 2010 +/- 60 B.P. (Beta
68758/CAMS 10484), slightly early,
but well within the accepted chronology
for Ohio Hopewell. Also the use
of hickory by the builders raises
an interesting point. N'omi Greber
(1983:54) reports that of the posts
sampled at the Edwin Harness Mound,
76% were identified as hickory.
Perhaps this may indicate a continued
utilization of a preferred material
by Hopewell peoples. Then again,
perhaps it doesn't.
Feature #2 and #4 were located
in the western end of the gravel
layer in trench A. As with feature
#1, both originate at a level below
the embankment fill. Unlike feature
#1 however, features #2 and #4
seem to represent sockets that
had silted in after the posts had
been removed. Feature #2 was 20
cm. in diameter and came to a dull
point 21 cm. below the gravel base.
Feature #4 was 15 cm, in diameter
and extended 15 cm. below the gravel
base where it also formed a dull
point. Both features contained
the same dark sandy silt with the
same minor charcoal flecking.
Feature #3 (Figure 2d) was located
in the south profile of trench
A and appeared to be a basin shaped
construct 10-12 cm. thick and about
1.2 m. across. The fill consisted
of about 90% fine rounded gravels
in a sandy matrix that also contained
a few small (<1 cm.) pieces
of charcoal. In plan this feature
continued into the wall 40-45 cm.
and was rather irregular in shape.
Feature #3 probably more accurately
represents a construction loading
rather than a basin per se.
What purpose the posts the aforementioned
post molds represent is unclear.
It is doubtful they formed part
of a stockade and the lack of habitation
debris or major features seems
to point away from any type of
structure. Perhaps considering
their location within the embankment,
they were measuring or surveying
points used by the ancient builders
as they perfected a craft so manifest
in their later works. Not enough
data was recovered however, to
make a concrete determination or
to rule out any of the above.
Conclusion
Squares are not
at all uncommon items in Hopewell earthwork
design. Well known examples include
those at Liberty, Baum, Hopewell,
Seip and at Newark. However,
these examples are exact segmented units
and not the axially aligned square
and circle combinations seen
at Anderson.
| |
|
click
on image to enlarge |
| While the Anderson square
is indeed a rare form, it is
not exactly unique. A very
similar form of comparable
dimensions is seen on James
Salisbury's highly detailed
1863 map of the Newark Earthworks
(Figure 3). Such adaptations
on the square may be an element
of Hopewell earthwork iconography
that was simply heretofore
overlooked, but not as rare
as one might think. |
|
 Figure
3. Newark Earthworks.
This map was created by James
Salisbury in 1863.
|
Finally, why
the Anderson Earthwork waited
until 1975 to be discovered
is somewhat of a mystery.
But suffice it to say it
was like finding a jewel
in the rough, one that is
now forever lost. Although
everything cannot be saved
from development, Anderson's
destruction was a sad commentary
on development at any cost.
Whether those parts of the
Anderson Earthwork that could
have been saved were destroyed
merely in the course of construction
or to keep "nosy" people
away is uncertain, the effect
was still the same.
Acknowledgments Special
thanks go to Jerrel Anderson,
Barbara
Donahue, N'omi Greber,
Brad Lepper, Robert Peterson
and
Dee Anne Wymer. Also to
Jeff Weinberger, John Pack,
Cathy
Wolf Crowser, Larry Wickliff,
Dave Towell to Jan Long,
and especially to Mr. Alva
McGraw,
the best friend the Hopewell
culture has.
Cited References
Anderson,
Jerrel
1980 A Recent Discovery-The
Anderson Earthwork.
Ohio Archaeologist,
30(l), Winter
1980.
Greber, N'omi
1983 Recent Excavations at the Edwin Harness Mound. MCJA Special Paper
No. 5. Kent State University Press.
|
|
|
|
