Full text of "The Wisconsin archeologist"

reference
collection
book

Kansas city
public library
Kansas city,
missouri

From the collection of the

3 <j

X P V m

o JtTelinger
u ^ JLlibr a ary
t p

San Francisco, California
2007

THE WISCONSIN
ARCHEOLOGHT

NOTES ON AZTALAN SHELL-TEMPERED POTTERY 1

by Peter Bleed

CANADIAN TRADE SILVER FROM INDIAN GRAVES 21

IN NORTHERN ILLINOIS by James W. VanStone

THE THEORY OF THE BIRDSTONE AS AN ATLATL 31

HANDLE GRIP, REVISITED by Robert Ritzenthaler

CLOVIS AND SANDIA-LIKE POINTS FROM DANE 35

COUNTY, WISCONSIN by Robert Ritzenthaler

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee, Wisconsin
Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. ML, Milwaukee Public
Museum, September thru May

OFFICERS
PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Qstberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Greenwald.

TREASURER

Wayne J. Hazlett
3768 S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct., Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr. Robert . Ritzenthaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL

Dr. David Baerreis, Dr. Joan Freeman, Elmer Daalmann,
Robert Hruska, W. O. Noble, R. W. Peterman, Ernest Schug,
Frank Squire, E. K. Petrie, J. K. Whaley, Mrs. Phillip
Wiegand, Paul Koeppler, Tom Jackland, Gale Highsmith,
Robert VanderLeest, Dr. Melvin Fowler, Robert Salzer,
Leo Klecker.

MEMBERSHIP FEES

The Wisconsin Archeologist is distributed
as part of their dues.
Annual Members, $5.50

All communications in regard to the Wisconsin Archeological Society
and contributions to the Wisconsin Archeologist should be addressed to
Mrs. Edward Flaherty, Secretary, N62-W15127 Tepee Ct., Menomonee
Falls, Wis. 53051. Entered as Second Class Matter at the Post Office at
Lake Mills, Wis., under the Act of August 21, 1912. Office of Publica-
tion, 316 N. Main St., Lake Mills Wis. 53551.

THE WISCONSIN ARCHEOLOGIST

New Series
MILWAUKEE, WISCONSIN - MARCH, 1970

Published Quarterly by The Wisconsin Archeological Society

NOTES ON AZTALAN
SHELL-TEMPERED POTTERY

Peter Bleed

University of Wisconsin,

Madison

Because of the long history of archeological research at the site and
the unusual artifacts found there, Aztalan is the most famous prehis-
toric site in Wisconsin. Early excavations which culminated in Bar-
rett's 1933 report indicated clearly that Aztalan was unlike any other
archeological site in Wisconsin in that it showed close architectural and
ceramic parallels to Cahokia and other sites of the Mississippian
tradition. Since Barrett's report was published, several authors have
cited ceramic similarities as evidence that Aztalan was a "northern
colony" or "transplanted village" of Mississippian culture. In making
such interpretations, these authors have glossed over the many
typically "non-Mississippian" ceramic items illustrated by Barrett.
More recently, Barreis and Freeman (1958) have clearly shown that
grit-tempered, cord marked pottery is very common at the site, and
that this pottery is closely related to Late Woodland types from Wis-
consin. Since Barrett's work, however, Aztalan shell-tempered pottery
which looms so large in linking the site to the Mississippian tradition,
has not been described in print. Before the specific relationships of
Aztalan can be known, an exhaustive analysis of all the material re-
covered from the site will have to be made. This 'paper does not
attempt such an enormous task. It rather seeks to give some insights
into the ceramic assemblage from the site by defining a series of at-
tributes which describe the Aztalan shell-tempered rim sherds.

The present study was based on the Aztalan pottery collections
housed at the University of Wisconsin, Department of Anthropology
and the State Historical Society of Wisconsin. The University col-
lections were obtained in excavations conducted by the State Archeolo-

The author would like to thank Professor James B. Stoltman and Dr.
Joan E. Freeman for their assistance with the present study and for
critically reading the manuscript, and Dr. Freeman and Professor David
Baerreis for making available the collections on which the study is
based.

2 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

gical Survey in 1949. Sherds recovered by the Historical Society in
1964 and most of the material collected in 1967 were also included
in the sample. In total there are 11,888 sherds catalogued in these
collections. On the basis of the tentative identifications recorded
in the site catalogues, it was determined that of the total sherd sample,
9,161 or 77.1% are grit-tempered and the remaining 2,727 or 22.9% are
shell-tempered. These figures are only tentative but they show clearly
that shell-tempered pottery is in the minority at Aztalan.

For this study body sherds were not examined and the analysis
was limited to shell-tempered rim sherds. Thus as a first step, the rims
were separated from the mass of pottery. The total sample yielded
318 rim sherds. In an attempt to reconstruct a vessel count, broken
rims were mended and sherds from a single vessel were treated to-
gether. This process indicated that 263 vessels were represented in the
sample. In fact, of course, most vessels were represented by only a
single sherd. The next step entailed separating the rims into groups
which corresponded to pottery types established by Griffin (1949)
on the basis of sherds recovered from the Cahokia site which is located
near East St. Louis, Illinois. This separation seemed highly sub-
jective, but the groupings obtained in this way fit comfortably into
the established Mississippian types. Until these are redefined in a more
specific and detailed manner, the Aztalan pottery discussed here will
fit within their bounds. When they are redefined, the categories dis-
cussed will probably form varieties of Mississippian types.

Attributes, or identifiable elements of the sherds, were isolated and
defined for each of the type categories which had been established. In
order to describe the type categories, the frequency of the various at-
tributes and associations between various combinations of attributes
were recorded. When the samples were large enough, the associations
between various attributes were tested with a simple statistical test for
chi-square which indicates the presences of non-random or apparently
preferred attribute clusters (see Spaulding 1953). In most cases the
rim sherds in the sample were too small to be accurately measured.
When possible, the mouth diameters were determined with circular
templates which were graduated in 1 cm. intervals. Maximum vessel
diameters and shoulder heights were not recorded since very few rim
sherds were large enough to allow for these measurements.

In all, four Mississippian pottery types and one miscellaneous cate-
gory of shell-tempered vessels are represented in the Aztalan collection
studied. The frequency of these categories is summarized in Table 1
and they are described in the following pages.
Powell Plain

Griffin (1949:49-50) originally defined the Powell Rain pottery

Aztalan Shell-Tempered Pottery

type on the basis of sherds collected at Cahokia. Pottery of the type
sample was shell-tempered with fine to medium paste. It varied in color
from smudged black to light buff and was usually formed into globular
jars with distinct shoulders, incurving shoulder areas and rounded out-
curved rolled rims. Supplementary vessel shapes including simple
bowls, water bottles, beakers and large pans were also associated with
the type as originally defined.

In the Aztalan collection studied for this paper, 203 rims repre-
senting 178 vessels (67.9% of total) can be placed into the Powell
Plain type. An additional 21 rolled rim sherds which are too
fragmentary to classify may also be included here. Since only shell
tempered sherds were considered for this study, the present Aztalan
sample matches the type collection in temper. Like the Cahokia
material, Powell Plain sherds from Aztalan show fine to medium paste.
Exterior surfaces are universally smoothed and no evidence of cord
marking was noticed. Rubbing which was used to finish the surfaces
also oriented the small plates of shell temper in the general plane of
the vessel walls. Although there are no reconstructed or whole vessels
in the sample discussed here, most of the sherds appear to have come
from globular jars which had a distinct shoulder at the greatest vessel
diameter. Above this point, the shoulder areas slope inward to the
point of greatest constriction. At this point, the clay was turned out-
ward to form a true rolled rim. A single loop handle was included in
the present sample. It is round in cross section and is added to a round
top, square rimmed black vessel which has an interior angle and a con-
tinuous curved shoulder. Maximum mouth diameters range from
52 cm. to 1 2 cm. with most vessels at the higher end of this range.

To further describe the Aztalan Powell Plain rim sherds, the fol-
lowing six sets of mutually exclusive attributes were defined.

1. Color/Surface finish

Two general color categories, each representing about half of the
total collection were identified. Of the total 178 vessels, 88 ranged
from dark grey to black. The remaining 90 vessels ranged in color from
red-brown to light buff. Two different surface finishes were also noted.
Most of the collection (130 vessels) showed a smoothed dull exterior
surface. This was unlike the polished surfaces which were seen on ap-
proximately 26% (48) of the vessels. In the original type definition
Griffin suggests that a continuum could be established between pol-
ished and dull surfaces. No such continuum was noted in the Aztalan
collection. Plain and polished sherds were easy to distinguish, although
some of the heavily weathered sherds which were included in the dull
category may have had polished exteriors. Color and surface finish

4 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

were combined to form four attributes:

a. Plain Black. The rims of 67 vessels (76 sherds) had smoke-
blackened exteriors which ranged in color from dark grey to
black. Although the exteriors were smoothed, these sherds
showed no sign of polishing.

b. Polished Black. Twenty-six sherds representing 21 vessels had
dark black exterior surfaces which showed fine striation. These
were apparently made by rubbing the surface with a narrow
polishing tool. All but the most heavily weathered sherds in
this category reflected light. No black sherds showed def-
initely slipped surfaces.

c. Plain Buff. Sixty-three vessels (71 sherds) are red to light buff in
color and show smoothed but unpolished surfaces. A few of the
heavily weathered sherds may have been slipped, but it is
impossible to positively identify an added slip on any sherds
in this category.

d. Polished Red. This category consists of 27 vessels represented by
30 sherds which show polishing marks, reflect light and range
in color from brown-red to orange. Nearly all of these sherds
have a thin exterior layer of clay which is either the result
of extensive polishing or an intentional slip. A very few
sherds (3 representing 3 vessels) show smoke smudged areas
on otherwise red surfaces.

The frequency of the four surface finish attributes is reviewed in
Table 2a which shows that plain unpolished pottery, either black or
buff, is most characteristic of Aztalan Powell Plain pottery.

2. Rim Roll

All Powell Plain jars snow a true rolled rim; that is the rim is curved
directly outward at the point of greatest constriction. Except for the
angle at the point of constriction, the vessels have no true neck. An
arbitrary division was made between rims:

a. Those which were rolled 90 or more as measured from the im-
mediately adjacent portion of the shoulder area.

b. Those which were rolled less than 90 as measured in the
same way.

As shown in Table Ib, rims rolled 90or more were found on more
than 75% of the vessels studied (130 vs. 48 vessels, 146 vs. 57 sherds as
shown in Table 2b).

Aztalan Shell- Tempered Pottery

Vessel Count Sherd Count

TYPE number percent number percent

Powell Plain 178 67.7 203 63.8

Ramey Incised 21 8.0 36 11.3
Unassigned Rolled

Rim 21? 8.0 21 6.6

Monk's Mound Red 11 4.2 19 6.0

Cahokia Cord Marked 1 .4 1 .3

Bowls 13 4.9 18 5.7

Everted Lip Jars 12 4.6 14 4.4

Beakers 4 1.5 4 1.3

Plate 1 .4 1 .3

Applique Design 1 .41 .3

263 100.1 318 100.0

Table 1

Frequency of Aztalan Shell-Tempered Ceramic Categories

3. Rim Edge Form

The rim edge is defined here as the top edge (not surface) of the
rim. Two forms were defined:

a. Angle top rims occur when the rim interior (fig. 1) meets the
lip in a distinct angle. Rim interiors, from the top edge to
the interior point of greatest constriction, were essentially
always slightly rounded. Angle top rims seem to have been the
preferred form since they are found on 65% or 117 of the ves-
sels (132 sherds) in the total collection (Table 2c).
. b. Round rim edges show no distinct angle between the interior
rim and the lip. Rather, the two round together with no distinct
boundary. Sixty-one vessels or 71 sherds had round top rims.

4. Form of Interior Constriction

As noted above, the greatest interior constriction was always the
point of juncture between the shoulder area and the outcurved rim
interior. A division was made between interiors:

a. Angle interior, which occurs wiien the rim interiors meets the
interior of the shoulder area in a distinct angle.

b. Round interiors, which have the rim interiors rounding into
the interior of the shoulder area with no distinct single point
of direction change (fig. 1).

Angle interiors were by far the most common of the two types,
being present on over 90% (161 vs. 11 vessels or 190 vs. 13 sherds) of

6 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

the sample. We may thus assume that angle interiors were preferred by
Aztalan potters.

5. Rim Shape

At the risk of over generalization, two basic rim shapes were
defined:

a. Square rims may have rounded edges, but they show three dis-
tinct sides. The rim interior and exterior are approximately
parallel and are joined by a lip which is perpendicular to them
(fig. 1). Square rims are present on 122 of the vessel (140 sherds)
studies.

b. Triangular rims show only two sides with the rim exterior
sloping upward to meet the interior at the rim edge. They

Vessel Count

Sherd Count

Plain Black

Polished Black

Plain Buff

Polished Red

Rin

Rim Rolled less than 90 48 57

Rim Rolled more than 90 130 146

Angle Top Rim 117 132

Round Top Rim 61 71

Interior Angle 167 190

Interior Round 11 13

Square Rim 122 141

Triangular Rim 56 62

Straight Shoulder 32 36

Partial Straight 71 78

Curved Shoulder 75 89

Table 2

Frequency of Aztalan Powell Plain Attributes

Aztalan Shell-Tempered Pottery

Rio Edge

Shoulder

Riffl Interior

Point of Greatest Interior
Constriction

Him Gross Section
Showing!

Rim Rolled less

than 90,

Angle top Square

rim,

Interior Angle,

Straight Shoulder.

Rim Oroas Section
Showing:

Riot Rolled less

than S

Angle top Triangular

rim,

Interior Angle,

Partial Straight

Shoulder.

Oro0 Section
Showings
Rim Rolled more
than 90

Round top Square rim,
Interior Round,
Ourred Shoulder.

Figure 1

Terms used to describe Powell Plain and Ramey Incised Jars

thus lack a lip. Triangular rims are present on 56 of the vessels
(62 sherds) studied (Table 2).

6. Shoulder Area Form

The shoulder area is defined as the inward sloping vessel section
which is located between the shoulder and the point of greatest
constriction. Three shoulder area forms were noted (fig. 1).

a. Straight shoulder areas run from the point of greatest con-
striction to the shoulder in a direct straight line. This type of
shoulder area occurred on 32 vessels or 17.9% of the total sample
(36 sherds). Maximum thickness on straight shoulders occurs
near the rim. It ranges from 4 mm. to 9 mm.

b. Partial straight shoulders are composite forms which have a

8 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

straight or very slightly curved section near the vessel shoulder.
This straight section generally accounts for 2/3 to 3/4 of the
shoulder area. It is separated from the rest of the shoulder area
by a fairly sharp curve. The short section near the rim is either
straight or curved. On larger vessels, the area of the curve and the
upper section are thickened so that the exterior of the shoulder
area is straighter than the interior. Maximum thicknesses vary
from 3mm. to 9 mm. Partial straight shoulders were seen in 71
(39%) of the vessels (78 sherds).

c. Continuous curved shoulder areas form an even curve from the
shoulder to the rim with no significant change in the rate of
curvature. Continuous curved shoulder areas tend to show uni-
form thickness throughout their total length. Maximum thick-
ness, ranges from 4mm. to 9mm. This style of shoulder area was
seen on 75 (40%) of the vessels studied (89 sherds).
The dominance of curved and partial straight shoulder areas as
seen in Table 2 indicates that trjey were the preferred forms.

In an attempt to determine if any specific combinations of the
above attributes were preferred by the Aztalan makers of Powell Plain
pottery, statistical tests for chi-square were run on all possible attri-
bute combinations. The data for these tests are contained in the lower
part of Table 3 which shows the occurrence of each attribute with all
others. The results of the tests are shown in the top portion of the
chart. Combinations which occurred in a highly non-random manner
(greater than .01, or with a statistical probability of less than 100
to 1 of showing such a distribution by chance alone) are indicated with
a plus sign. On the basis of these tests, we may conclude that certain
intra-type tendencies can be identified in the Powell Plain vessels of
Aztalan. First, most of the potters preferred rims rolled more than 90.
When they made such a rim, they also preferred round top square rims
and partial straight shoulders. There was also an identifiable tendency
for vessels with rims rolled more than 90 to be made with plain black
surfaces. Vessels which were made with rims rolled less than 90
are associated with and were apparently preferred with traingular rims
and continuous curved shoulder areas. Similar traits were also pre-
ferred on polished red vessels. Finally, although the statistics are not
positive, it appears that analysis of a larger sample would show a sig-
nificant association between round interiors and round lips.

Ramey Incised

The Ramey Incised type was defined by Griffin (1949:51) as the
decorated companion of Powell Rain. In the Aztalan collection being

Aztalan Shell-Tempered Pottery

Plain Black

Polished Blaok

Plain Buff

Polished Rod

Rim Rolled
> 90

Rim Rolled
<C 90

Angle Top
Rim

Round Top

Interior Aagle

Interior Round

Square Rim

Triangular
Rim

Straight
Shoulder

Partial

Straight

Shoulder

Continuous

Ourred

Shoulder

Table 3

Association of Powell Plain Attributes
lower number = vessel count
upper numer = sherd count
pluses (+) indicate positive non random association ( .01)

considered, only 36 sherds representing 27 vessel (8.0% of the total
sample) fall into this type. These vessels differ from those of the Powell
Plain type in that they have had a design added to the exterior of the
shoulder area. No polished black decorated sherds were noted although
some were so heavily weathered that it was impossible to determine
what their original finish may have been. With this exception, the at-
tributes defined for the Powell Plain type may be used to describe the

1 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 1

Plain
Buff

Polished
Red

Rim Rolled
> 90

Rim Relied
<90

Angle Top
Rim

Round Top
Rim

I nter lor
Angle

I nterior
Round

Square
Rim

Triangle
Rim

Straight
Shoulder

Partial
Straight

Curved
Shoulder

Trailed
Des i gns

I ncised
Des i qns

\IO

XXX

Engraved
Des i gns

\
(X

Chevron
Pattern

Chevron w/
Horizontal

xxvvv

Table 4

Frequency and Association of Aztalan Ramey Incised Attributes

shape and surface finish of all Ramey Incised vessels. The occur-
rence of these attributes and the others which are about to be dis-
cussed is shown in Table 4.

1. Decorative Technique

The decoration on Aztalan, Ramey Incised pottery was executed in
one of three ways:

a. Trailed lines were made by drawing a blunt tool through the sur-
face of soft clay. The grooves formed in this way are wider

Aztalan Shell-Tempered Pottery 1 1

than they are deep and have round bottoms. Sometimes small
ridges of clay were built up along the margins of the trailed lines
and occasionally the trailing formed cameos on the vessel in-
terior. Trailed lines were present on 14 sherds which came from
11 different vessels.

b. Incised lines were drawn with a tool which actually removed a
part of the vessel surface. They generally show a "V" shaped
cross-section and are deeper than they are wide. Incised lines
were apparently made after the vessel was quite dry, and they
never showed cameo impressions. Seven vessels (18 sherds) had
incised decorations.

c. Engraved lines, which are found on three vessels (four sherds)
of the present sample, were scratched into the vessel surface
after firing had been completed. Engraved lines were quite crude,
showing irregular, rough edges and generally penetrating into the
reduced core of the vessel wall.

2. Decorative Motif

The small number of decorated sherds and their small size makes it
impossible to present a detailed inventory of the patterns used. Four
general types of motifs can be identified:

a. Chevron patterns, consisting of concentric "V's" or parallel ob-
lique lines, occur on 8 vessels (8 sherds).

b. Chevron patterns on 5 vessels (8 sherds) were interspersed with
and connected by parallel horizontal lines.

c. Scroll patterns occurred on 5 vessels represented by 10 sherds.
These sherds are so small that it is impossible to determine the
exact patterns involved. Interlocking series of single scrolls and
concentric scrolls were both apparently used.

d. Irregular, crude curvilinear lines were engraved on two vessels
(two sherds).

Because of the small sample size which makes statistical analysis
impractical, it is difficult to point out significant clusters of attri-
butes for the Ramey Incised type. Table 4 shows that the same formal
attributes which were favored for Powell Plain vessels were preferred
on decorated vessels. To determine if attributes which tend together on
Powell Plain jars are commonly associated with one another on deco-
rated vessels, chi-square tests were run on the association of formal at-
tributes within the combined population of plain and decorated vessels.
These tests show that decorated vessels have the same attribute clusters
(e.g. round rim top, square rim, and partial straight shoulder area as-
sociated with rims rolled 90 or more, continuous curved shoulder
areas and triangular nms associated with rims rolled less than 90)

1 2 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

as those characteristic of the Powell Plain type. In other words, in
terms of form, the two types are identical.

Analysis of the present sample has not pointed up any definite as-
sociations between formal and decorative attributes. Engraving was
used to make the irregular designs but except for this, it appears that
there were no preferred techniques for making any of the various
motifs used to decorate the pots.

Monk's Mound Red

The Monk's Mound Red type was defined by Griffin (1949:52-3)
to include red painted spherical jars, bowls and plates from Cahokia.
The temper of the type sample consisted of large quantities of crushed
limestone along with fragments of mussel shell. In the Aztalan collec-
tions, 1 9 sherds, representing 1 1 nearly spherical jars, fit into the origi-
nal type definition at least in terms of vessel shape and surface finish.
Other possible examples of the type will be discussed with the miscel-
laneous vessels. Most of the sherds included in this category have red
slipped exterior surfaces. The few sherds which show no slip are all
heavily weathered so that it does not seem unreasonable to presume
that they were originally slipped. The paste of all these sherds is not
significantly different from that seen in other shell tempered pots in
the collection. Shell fragments seem to have served as the main
aplastic agents in all cases. A few sherds may contain small amounts of
finely crushed limestone (and in one case, sand) in addition to shell
fragments. These are, however, in a distinct minority. Wall thick-
ness ranged from 4 mm. to 7 mm., and the measurable mouth diam-
eters ranged from 14 cm. to 18 cm.

The 11 vessels included in the Monk's Mound Red type all had
continuously rounded sides with the greatest construction at the lip.
All these jars were slipped on the exterior with the added coating of
clay extending just past the lip to the vessel interior. Seven attributes
were defined to further describe the vessels.

1. Surface Finish

a. Seven vessels represented by 13 sherds were red-brown with
polished slipped exterior surfaces. The polishing was finely
done so that it is hard to detect striations caused by the polish-
ing tool.

b. Two sherds from different vessels have chalky deep red un-
polished surfaces.

c. Two vessels (two sherds) have polished surfaces which show
zones of two different colors. One sherd shows polished red
areas interspersed with areas of buff slip. The sherd is too small

Aztalan Shell-Tempered Pottery

to determine the exact pattern of zoning, but the buff zone ap-
parently formed a narrow band around the vessel lip. There were
also triangular buff colored zones on the vessel wall. Another
sherd has dull buff slip added in zones which were outlined with
broad trailed lines.

2. Rim Form

a. Six of the spherical jars, represented by 14 sherds, showed double
thickened rims which apparently had a layer of clay added to the
rim interior, and the change in thickness is only apparent on
the rim interior. These vessels all have rounded lips.

b. The remaining jars (5 vessels 5 sherds) have unthickened rims
with rounded lips.

3. Decoration

a. A single sherd is decorated with an oblique row of punctates.
The punctates are 3 mm. in diameter and were apparently made
with a reed or a hollow tube since each shows a raised center.

Pol ished Red

Chalky Red

Zoned SI ipping

Doubled Rim

Unthickened
Rim

Punctates

Trailed Line
Des ? gn

Rim
Indentation

Table 5

Frequency and Association of Aztalan Monk's Mound Red Attributes

1 4 WISCONSIN ARCHEOLOGIST Vol. 51, No. 1

The punctates are located on the thickened portion of a doubled
rim and perhaps for this reason they show no cameo.

b. Another single sherd shows a small section of a curving broad
trailed line. The line is 4 mm. wide and quite shallow.

c. One vessel represented by two sherds has a deep trailed line or in-
dentation on the exterior surface at a point which corresponds
to the end of the rim thickening on the vessel interior.

Table 5 represents the associations of the attributes of Monk's
Mound Red jars. The sample is far too small to allow any positive state-
ments on the clustering of these attributes. Perhaps analysis of a large
sample may show that punctates or trailed decoration were restricted
to vessels with doubled rims.

Cahokia Cord-Marked

A single large rim sherd from Aztalan fits into the Cahokia Cord-
Marked type. This sherd is from a rolled rim, shouldered, round-bottom
vessel. In shape it is identical with several Powell Plain sherds with a
square angle top rim rolled more than 90, an interior angle, and a par-
tial straight shoulder area. The exterior below the rim exterior is cov-
ered with Z twist cord wrapped stick rolling. The diameter at the inter-
ior point of greatest constriction is 18 cm. The interior surface is
smoothed. The sherd is tempered with very fine shell fragments and
also may contain some fine particles of grog.

Figure 2

Cross Sections of Unrolled Rim Jars

a. Thickened Interior Beveled Rim

b. Unthickened Round Rim

Aztalan Shell-Tempered Pottery

Miscellaneous

Included in the collection studied for this paper were five cate-
gories of vessels which could not be positively linked to any of the
Cahokia types discussed thus far.

A. SHOULDERED JARS WITH UNROLLED RIMS

This group of 14 rim sherds, which represents 12 different vessels,
is probably related to the Powell Plain type. Vessels in this category
are globular jars which have a fairly distinct shoulder at the greatest
diameter. The shoulder area slopes inward past this point and ends in
a slightly everted but unrolled lip (fig. 2). The surface finishes and the
shoulder area forms seen on these jars may be described in terms of
the attributes defined for the Powell Plain type. Two more sets of at-
tributes were defined to describe other features of these vessels. The

\S* I I

& Oi

fl>

Frequency

Pol ished Black

Plain Black

Plain Buff

SI ipped Red

Beveled Rim

Round Rim

Square Rim

Table 6

Frequency and Association of Unrolled Rim Jars Attributes

1 6 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 1

frequency and associations of these attributes are summarized in
Table 6.

1. Rim Form

a. Seven vessels (8 sherds) have interior beveled rims. The interior
and exterior surfaces of the shoulder areas of these vessels paral-
lel one another until they reach a point a few millemeters below
the rim edge. The exterior surface continues past this point with
no change in direction. The rim interior, however, turns slightly
outward at this point and meets the exterior in an angular rim
edge (fig. 2). The turn in the rim interior generally forms a dis-
tinct angle.

b. Round rims are seen on four sherds from four different vessels.
These sherds have both edges of the rim rounded.

c. A square rim is seen on a single vessel represented by two sherds.
The rim of this vessel has a flat lip which is approximately per-
pendicular to the axis of the shoulder area.

Frequency

Interior
Rim Ri doe

Beveled
Rim

Square
Rim

Round
Rim

Tni ckened
Rim

-o

Q.
Q.

O TJ
.- 0)
i- Q.
Q> Q_
4J
C

Table 7

Frequency and Association of Aztalan Bowl Attributes

Aztalan Shell-Tempered Pbttery 1 7

2. Vessel Size

Six of the everted rim jars (7 sherds) may be classed as miniature
vessels. The maximum interior diameter of these jars ranges from 4 to
9 cm., and the maximum thickness of the shoulder area wall varies
from 3mm. to 6mm. Generally these vessels are quite crudely made and
show c high shoulder angle. It was impossible to accurately determine
the mouth diameter of the other vessels in this category, but all seem
to have been larger than 14 cm. Maximum shoulder thickness for the
larger vessels ranged from 5 to 9 mm.

Once again it is impossible to isolate any definite clusters of at-
tributes which are typical of this category.

B. BOWLS

The remains of 13 bowls evidenced by 18 sherds were included in
the analyzed sample of Aztalan pottery. Both the Plain Powell and
Monk's Mound Red types were defined to include bowl-forms. It was
impossible to link the Aztalan bowls to one or another of these types,
so they are described separately here. All these vessels have slightly
convex walls which blend into round bottoms. No rim effigies are defi-
nitely associated with them, and they all have medium to fine paste
which includes shell fragments. One bowl has a noticeable amount of
sand in addition to the shell. Maximum diameter of the bowls ranges
from 1 2 cm. to 20 cm., and maximum wall thickness is between 4 mm.
and 7 mm. Two sets of attributes have been defined to describe the
bowls. The frequency and association of these attributes is summarized
in Table 7.

1. Surface Finish

a. Five bowls (9 sherds) have slipped and polished exteriors. The
color of the slipped surfaces ranges from orange to red-brown.
These finishes are well polished and rarely show burnishing stria-
tions. The unslipped interior surface is smoothed but not pol-
ished and buff or brown in color.

b. Eight bowls (9 sherds) have both the interior and exterior slip-
ped and polished finish. The color of these vessels varies from
dark brown to red.

2. Rim Form

a. One bowl represented by two sherds has a rounded straight rim
which has a small modeled ridge on the rim interior. This ridge
increases the thickness of the vessel wall from 6 mm. to 9 mm.
(fig. 3).

b. Three bowls (3 sherds) have interior beveled rims like those

1 8 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

Figure 3

Cross Section of Aztalan Bowl Rim Forms

a. interior ridge

b. interior bevel d- round

c. square e. thickened

described for the everted rim jars (fig. 3).

c. Four bowls (4 sherds) have square rims with flat lips which meet
both vessel surfaces at distinct angles (fig. 3).

d. Round rims are seen on three bowls represented by three
sherds (fig. 3).

e. Two bowls represented by six sherds had their rims thickened by
having a part of the edge rolled onto the exterior surface,
(fig. 3).

A larger collection of bowls is needed before a definite statement
can be made about the typical characteristics of this class. The bowls
studied for this paper are more elaborately finished than most of the
Powell Plain vessels discussed earlier. They seem more similar to the
Monk's Mound Red jars, but the variation in the surface color seen of
the bowls, and the fact that they are shell tempered makes it diffi-
cult to place them neatly in that type.

C. BEAKERS

Four straight vertical-sided beakers, each represented by a single
sherd, were included in the analyzed sample. One sherd has a polished
back finish on both the interior and exterior surfaces. It has a round
lip. The remaining sherds all have red slipped surfaces which were
apparently polished, although they are heavily weathered now. One of
them has an unthickened rim with a round lip. The remaining two have
double thickened rims with exterior indentations at the end of the
thickened section. All are, of course, shell tempered and have smooth
exteriors. Maximum diameters range from 14 cm. to 36 cm., and maxi-
mum wall thickness varies from 4 mm. to 6 mm. These beakers may be-
long to either the Powell Rain or Monk's Mound Red type.

D. PLATE

The only plate in the sample was represented by a single sherd.

Aztalan Shell-Tempered Pbttery 19

The plate is 18 cm ; in diameter and is very slightly concave. It has a
round lip and a smoothed surface which is so heavily weathered that it
is impossible to determine the nature of its original surface. Both the
Powell Plain and Monk's Mound Red types include plates.

E. Only a single shell-tempered sherd from Aztalan has no parallels at
Cahokia. This sherd is from a highly polished red slipped shouldered
jar which has a slightly curved row of applique nodes running obliquely
across the shoulder area. The sherd has an angle top square rim which
is rolled less than 90 . The shoulder area is partially straight. The red
slip, which covers the exterior of the sherd, is thicker than that seen
on either Powell Plain or Monk's Mound Red jars. It shows extensive
burnishing. The paste* of the sherd is quite coarse and includes fairly
large fragments of mussel shell. The mouth diameter of the vessel is
18 cm., and the maximum thickness of the shoulder area wall is 5 mm.

CONCLUSION

This paper has answered a few of the many questions which exist
about the Aztalan ceramic assemblage. It has shown that a number of
pottery types which are found at Aztalan can be favorably compared
to Mississippiari types from Cahokia. The Mississippian types found at
Aztalan share a few attributes of temper, decorative technique, and
surface finish which isolate them from the dominant grit-tempered
cord-marked pottery types present at the site. Few if any attributes
are found in both the grit-tempered and in the shell-tempered types. If
we assume that the collections analyzed for this paper represent a
random cross section of the Aztalan ceramic inventory, the study has
shown that Mississippian pottery accounts for less than a quarter of the
total ceramic assemblage from the site. With the exception of Powell
Plain and Ramey Incised the Mississippian types present at Aztalan are
distinguishable from one another on the basis of vessel form. At Azta-
lan, the Powell Plain and Ramey Incised types share the same formal
attributes and in both, these attributes are associated with one another
in approximately the same manner. Aside from the decoration on the
latter, these two types appear to be identical. This study has not dealt
with the exact similarities between Aztalan shell-tempered pottery and
that recovered from Cahokia or other sites. Such a statement will have
to wait until the Cahokia types have been defined in greater detail.
Finally, the pottery studied for this paper is but a small portion of the
material which has been recovered from Aztalan. The attributes and
attribute clusters which have been suggested as characteristic of Azta-
lan shell- tempered pottery may be evaluated and amplified by an
analysis of the total sample of pottery which is available from the site.

20 WISCONSIN ARCHEOLOGIST Vol. 51 , No. 1

BIBLIOGRAPHY

Baerreis, David A. and I.E. Freeman

1958 "Late Woodland Pottery in Wisconsin as Seen from Azta-

lan," Wisconsin Archeologist, Vol. 39, No. 4, pp. 35-61.
Barrett, S.A.

1933 "Ancient Aztalan," Bulletin of the Museum of the City of

Milwaukee, Vol. XIII, pp. 1-602.
Griffin, J.B.

1949 "The Cahokia Ceramic Complexes," Proceedings of the

Fifth Plains Conference for Archeology, pp. 44-58.
Spaulding, Albert C.

1953 "Statistical Techniques for the Discovery of Artifact
Types," American Antiquity, Vol. XVIII, No. 4, pp.
305-313.

Canadian Trade Silver 21

CANADIAN TRADE SILVER FROM INDIAN GRAVES
IN NORTHERN ILLINOIS

James W. VanStone
Field Museum of Natural History

The archaeological specimens which form the basis of this paper
were excavated in the early years of the present century from Indian
burial mounds in Forest Home Cemetery on the banks of the Desplaines
River. Today the cemetery is in Forest Park, a western suburb of Chi-
cago. Although details are lacking, the burial mounds were apparently
leveled to fill in a swampy area of the cemetery and at least thirteen
skeletons were uncovered along with a quantity of artifacts. These ma-
terials, stored for many years in the cemetery office, were eventually
turned over to the Forest Park Library where I examined them in Oct-
ober, 1968. l

The collection includes a variety of laminated sheet silver trade orna-
ments, most of which are reasonably well preserved. Some have identi-
fying marks that make dating within fairly narrow limits possible. Such
ornaments were made for use in the fur trade by silversmiths in Canada,
England, and the United States and are considered to be the best cri-
terion for dating archaeological sites in the western Great Lakes area be-
tween 1760 and 1820 (Quimby 1966:91). Although there is a con-
siderable literature on various aspects of Indian trade silver, relatively
few collections of known provenience have been described and illus-
trated in detail; none at all from the northern Illinois area.

The collection to be described here was not excavated under con-
trolled conditions and it has been impossible to obtain more data con-
cerning provenience than the scanty information noted above. Never-
theless, these ornaments reveal interesting details about the spread of

I would like to express appreciation to Mrs. Isabel B. Wasson of
River Forest, Illinois for calling my attention to the silver ornaments
described in this paper. Thanks are also due to Miss Josephine Austin,
Librarian of the Forest Park Library, for permission to remove the
artifacts to Field Museum of Natural History for study. Mr. John E.
Langdon of Toronto read an early draft of this paper and offered
several valuable suggestions.

22 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

trade silver into Illinois and its use by the Indians.

Because of the considerable movements of people into and out of
northern Illinois during the eighteenth and early nineteenth centuries,
it is difficult to determine with certainty the specific Indian tribe rep-
resented by these graves on the banks of the Desplaines River. It is
known, however, that the Potawatomi began moving info northeastern
Illinois from the Green Bay region early in the eighteenth century, and
by 1 800 there were sizeable settlements of these Indians in the Chicago
area. .Many qf the Potawatomi villages contained numbers of Ottawa
and Chippewa Indians who were closely related to the former both
linguistically and culturally (Bauxar 1959:54; Blasingham 1961:166-
169). It is very likely, therefore, that the occupants of the graves, those
who wore the ornaments about to be described, were members of one
of these tribes. The major Potawatomi withdrawal from Illinois began
in 1832 and had been virtually, completed four years later (Bauxar
1959:54).

The silver ornaments from graves in the Forest Home Cemet *y can
most conveniently be considered under five headings which correspond,
for the most part, to the areas of the body where they were worn;
wristbands, armbands, brooches, gorgets and earrings. All of these ob-
jects, when first examined by the author, were covered with a heavy
green patination. This was caused by the copper alloy in the silver, a
substance that resulted in the preservation of a number of textile frag-
ments with which the silver had been in close contact. Prior to study, as
many of the ornaments as possible were cleaned by means of a glass
bead blasting technique.

WRISTBANDS are, for the most part, remarkably uniform in ap-
pearance. Out of a total of ten specimens, seven have straight edges and
parallel ridges on the outer surface (PI. 1, 4). All of these are ap-
proximately 16 cm long and 2 cm wide. They are drilled at either end
for the insertion of thongs by which the bands were held and adjusted
to the size of the wearer's wrist. On four, specimens the thongs are still
in place. A single wristband is approximately the same size as the
others, but has scallop edges as well as parallel surface ridges (PI. 1,5).
Two specimens are quite distinctive and appear to have been cut from
armbands. Both exhibit, along one side, the longitudinal border ridges
characteristic of armbands; they are roughly cut along the other side
(PI. 1, 3). Fragments of thongs are attached to the perforations on
both these specimens.

In addition to the typical wristbands with parallel ridges just des-
cribed, there are two virtually identical bands that are much shorter and
narrower (PI . 1 , 8). These do not seem to have been cut down from full
sized wristbands. They were probably worn by small children.

Canadian Trade Silver

Wristbands from sites in Michigan similar to those in the Forest
Home Cemetery collection are described and illustrated by Quimby
(1958:317, fig. 1, 1) and Herrick (1958:11, fig. 29). Such bands were
apparently worn more frequently by women than by men (Quimby
1966:92).

There are five ARMBANDS in the collection, all of which were ap-
parently modified to some extent by their Indian owners. Four are vir-
tually identical, being approximately 7 cm wide, bordered by longitu-
dinal ridges, and from 12 to 14 cm in length. They are decorated with
designs, engraved with a small burin or engraving tool, which reveal an
interesting characteristic of the armbands. In each case the designs are
at one end of the specimen and cut off, suggesting that the Indians cut
the original bands in order to make a larger number of small ones.
These bands, which presumably were worn just above the elbow by men
(see Quimby 1966: fig. 32), originally circled the arm. But cutting the

24 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 1

original band in half, only the outer side of the arm would be covered
(PI. 1-2). Paired perforations at either end served for the insertion of
thongs which stretched around the unexposed side of the arm. Two of
these smaller armbands appear to have been made from the same full
sized one since the engraved design, a long necked bird, is continuous
on the two specimens. Another band also has part of a long necked bird
while a fourth shows the rear half of some four footed animal.

The fifth armband in the collection is fragmentary but appears not
to have been reworked in the manner of those just described. The speci-
men, which is approximately 25 cm long and was at least 10 cm wide,
closely resembles one from Michigan illustrated by Quimby (1958:
fig. 1 , 2). It has longitudinal ridges along the edges and an engraved cen-
tral design showing a seated human figure facing a seated animal, prob-
ably a dog. Only the lower halves of these figures are visible.

The most abundant ornaments in the Forest Home Cemetery col-
lection are BROOCHES of which there are four types. These belong-

Canadian Trade Silver 25

ing to type 1 , five in number, are concavo-convex disks with central cir-
cular openings. On two specimens a hinged pin for fastening the brooch
to the garment lies across the opening. These brooches, four of which
are illustrated (PI. 2, 7-8, 1 1-12), show a variety of sizes and simple en-
graved decorations. Type 1 brooches are illustrated in nearly every pub-
lished collect on of Indian trade silver (see Quimby 1958:fig. 1, 7; fig.
2, row 2; Alberts 1953: PI. 3, j).

Type I brooches are similar in shape to those just described but tend
to be la: D er and are much more ornate. All are ornamented with a sym-
etrical arrangement of oval, triangular, diamond shaped and semi-lunar
perforations in combination with engraved designs. To show the range
in size and decoration, four specimens are illustrated (Pi. 2, 1-3, 5).
Like those of type 1, the type 2 brooches have also been widely des-
cribed and illustrated (see Quimby 1937: fig. 2, first row; 1958: fig. 1,
5-6; 1966: fig. 20, lower left; Alberts 1953: PI. 3, e, PI. r, e-f:'
Herrick 1958: PI. 3, figs. 30-31).

There are two specimens belonging to type 3, uic > died "St^r
brooch" (Quimby 1966:93). Like the others that have been descriDcu
these brooches are concavo-convex in cross section with round open-
ings in the center. One has a nine-pointed star design enclosed by a nar-
row rim and a symmetrical arrangement of triangular and semi-lunar
perforations (PI. 2, 4). The other has an eleven-pointed star enclosed by
a narrow scalloped border (PI. 2, 6). There is some engraving on both
specimens. Star brooches would appear to be relatively common in col-
lections of Indian trade silver as they are present in most of the sources
mentioned above.

By far the most common brooches in the collection are those belong-
ing to type 4. These are plain, circular rings of silver, empty in the mid-
dle except for the cross pins which fasten them to garments. Although
they vary somewhat in diameter, there are two general sizes; a large var-
iety that is approximately 2 cm in diameter (PI. 2, 10), and a smaller
one with a diameter of 1.2 cm (PI. 2, 9). The large form occurs in
greater abundance with 64 loose specimens and more than 225 attached
in clusters to garment fragments. In addition there are eleven attached
to braids of human hair (PI. 3, 3). A similar use for such brooches has
been reported from a burial at the Ada site, Kent County, Michigan
(Herrick 1958:6). Eighty-three of the small brooches are attached to
textile fragments and there are six loose specimens.

Of considerable interest are these garment fragments with clusters of
type 4 brooches worn as decoration (PI. 3, 1, 4). 2

For assistance in identifying the garment fragments, I wish to thank
Mr. Harold B. Burnham, Textile Department, Royal Ontario Museum.

26 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 1

The greatest number of brooches, both of the large and small var-
iety, are attached to seven fragments of wood block printed cotton.
Materials of this type were produced in many parts of France and, to a
lesser extent, in England and Scotland. At least three of these frag-
ments appear to have been part of a one-piece, straight cut garment
with separate attached collar, probably a woman's dress. An interesting
feature of this dress, as shown on one large fragment, is a false shoulder
seam sewn on with two-ply linen thread. This seam may have been of a
different color than the rest of the garment.

A smaller number of large type 4 brooches were fastened to a frag-
ment of napped woolen flannel, slightly fulled, of average quality. The
best materials of this type were always of English manufacture with the

Canadian Trade Silver

average and poorer qualities being either English or French. This frag-
ment is probably part of a blanket, perhaps worn thrown over the
shoulder. Brooches attached in clusters to garments frequently are
shown in contemporary portraits of Indians, but they are not often
found in an archaeological context.

Thirty-one small brooches are fastened to a narrow strip of leather
with a tassel of worsted ribbon at one end (PI. 3, 2). A few brooches
were apparently attached to this ribbon as well. The object may have
been a dangle intended for wear as decoration on some garment.

Two types of GORGETS are recognized by students of Indian trade
silver and there is one specimen of each type in the Forest Home Ceme-
tery collection. A crescent-shaped gorget is concavo-convex in cross sec-
tion with suspension holes at the points of the crescent. The convex
surface of this specimen is engraved with small dots in two parallel,
wavy lines (PI. 1, 9). According to Quimby (1966:92), such ornaments
were an adaptation of gorgets worn by British army officers. This speci-
men is very simple and plain when compared to many illustrated speci-
mens in other collections (See Woodward 1945: Fig 3; Quimby 1958:
fig. 1,3; Barbeau 1942:12, upper left hand corner, bottom).

A fragmentary round gorget is also concavo-convex in cross section
(PI. 1, 10). It has an engraving of a fox or similar animal in the center

Figure 1

(fig. 1) and an engraved border on the concave side. There are two sus-
pension holes. Quimby (1966:93) believes that the round gorget had its
origin in the aboriginal shell gorget of the Indians.

There are two types of earrings in the collection. Three specimens
have a double-barred cross in combination with a small cone-shaped
pendent suspended on a thin piece of silver wire (PI. 1, 7). The sec-
ond type, of which there are ten specimens, is simply a cone-shaped
pendant suspended from a circular piece of wire (PI. 6). Similar ear-
rings are illustrated by Quimby (1958: fig. 1, 10), Herrick (1958:P 1.
3, 21) and Alberts (1953: PI. 9, e).

28 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

Silversmiths who produced ornaments for the Indian trade fre-
quently stamped their wares with their individual mark. Thirteen
specimens from Forest Home Cemetery are punch marked in this
manner and all marks are those of Canadian craftsmen. Three brooches,
two bracelets and one armband are stamped with the roman letters
"PH" separated by a pellet and in an oblong cartouche (see Quimby
1966: fig. 21, 3). This is the punch mark of Pierre Huguet dit Latour
(1749-1817), a Montreal silversmith and merchant, and his son (1771-
1 829). The elder Huguet sold large quantities of trade silver to North
West Company traders in 1797 and 1798 (Langdon 1966:94). Two bra-
celets with the Huguet mark also have the word "Montreal" in a long,
rectangular cartouche, while a single bracelet without maker's initials
is similarly marked.

There are three armbands which have punch marks consisting of
the roman letters "IS" in an oval cartouche (see Quimby 1966: fig. 21,
m). This is the mark of the Quebec silversmith Joseph Schindler and,
after his death, his wife. Schindler was living in Quebec City as a silver-
smith at least as early as 1767 and worked there until his death in 1786.
The Widow Schindler, as she was called, moved to Montreal and pro-
vided Indian trade silver from 1 797 until she died in 1 802. According
to Langdon (1966:125-126), it is not clear from the records whether
she actually made silver or simply acted as an agent. Traquair (1938:5)
suggests that she may have carried out her work as a kind of home in-
dustry (Alberts 1953:28; Barbeau 1940:33; Quimby 1966:199).

An important figure in the silver trade in Montreal was Narcisse
Roy (1765-1814). His punch mark, which occurs on one bracelet, con-
sists of the script letters "NR" in a trapezoidal cartouche (see Quimby
1966: fig. 21, h). Roy, who was apprenticed to the famous craftsman
Robert Cruickshank, made and sold silver to the Northwest Company,
Outfit 1799-1803 (Langdon 1966:121).

The Montreal silversmith Charles Arnoldi (1779^1817) is represented
by a single brooch (PI . 2, 8). His punch mark consists of the letters
"CA" in script capitals separated by a pellet (see Quimby 1966: fig.
21, g), and documentary evidence dates his work from at least 1784
to 1810 (Alberts 1953:26-27; Langdon 1966:42; Quimby 1966:198)
Barbeau (1942:13) illustrates a fine double-barred cross by Arnoldi
which is now in the American Museum of Natural History.

The final marked specimen to be discussed here is a bracelet with
the roman letters "PM" in a rectangular cartouche (see Langdon 1966:
103). This punch mark is probably that of Paul Morand (1775-1886).
He was apprenticed to Pierre Huguet dit Latour in 1 802.

On the basis of the information presented above, it can be said with
reasonable certainty that the graves in Forest Home Cemetery date be-

Canadian Trade Silver 29

tween approximately 1780 and 1820. During that period, the fur trade
was becoming increasingly important to the economy of Canada and a
large number of traders, including such well known names as McTavish,
Fraser and McGillivray, were closely associated with the North West
Company and its struggle with the Hudson's Bay Company for control
of the trade. These Montreal-based traders, many of whom traded in
the western Great Lakes region, tended to patronize local silversmiths
to a greater extent than did those of the Hudson's Bay Company who
generally ordered their supplies from England (Langdon 1966: 1 7-20).

In spite of considerable efforts made by the Americans to secure the
support of Indians in the Great Lakes area during the Revolutionary
War, most of the native peoples of this region favored the British. The
intensive activities of Canadian traders were at least partly responsible
for this situation and English influence continued to be felt until after
the War of 1812. In northern Illinois and contiguous areas the Chip-
pewa, Ottawa and Potwatpmi were all British allies. Quimby (1937:18,
20) believes that this explains the large proportion of Canadian silver
ornaments found in the area. It probably also explains why the trading
contacts of Indians buried in Forest Home Cemetery were primarily, if
not exclusively, with Canadian traders.

BIBLIOGRAPHY

Alberts, R.C.

1953 "Trade silver and Indian silver work in the Great Lakes
region." The Wisconsin Archaeologist, vol. 34, no. 1, pp.2-
121.
Barbeau, M.

1940 "Indian trade silver." Transactions of the Royal Society of

Canada, Section II, pp. 27-41.
1942 "Indian trade silver." The Beaver, Outfit 273 (December),

pp. 10-14.
Bauxar, J.J.

1959 "The historic period," Illinois Archaeological Survey, bul-
letin no. 1, pp. 40-58.
Blasingham, E.J.

1961 "Indians of the Chicago area ca. 1650-1816." Illinois Arch-
aeological Survey, bulletin no. 3, pp. 163-173.
Herrick, R.

1958 "A report on the Ada site, Kent County, Michigan." Michi-
gan Archaeologist, vol. 4, no. 1, pp. 1-27.

Langdon, I.E.

1 966 Canadian Silversmiths 1 700- 1 900. Toronto.
Quimby, G.I.

1937 "Notes on Indian trade silver ornaments in Michigan."
Papers of the Michigan Academy of Science, Arts, and Let-

30 WISCONSIN ARCHEOLOGIST Vol. 51, No. 1

ters, vol. 22, pp. 1 5-24.

1958 "Silver ornaments and the Indians." Miscellanea Paul Rivet,
Octogenario Dicata (XXXI Congreso Internacional de Am-
ericanistas), pp. 318-337.

1966 Indian culture and European trade goods. University of Wis-
consin Press.
Traquair, R.

1938 "Montreal and the Indian trade silver." Canadian Historical

Review, March, pp. 1-8.
Woodward, A.

1945 "Highlights on Indian trade silver." Antiques Magazine,
June, pp. 328-331.

Atlatl Handle Grip

THE THEORY OF THE BIRDSTONE AS AN ATLATL
HANDLE GRIP, REVISITED
By Robert Ritzenthaler

On a visit to the Mujica Museum of Gold near Lima, Peru, I noted
with some excitement a collection of Nazca spear-throwers. Due to the
desert conditions of coastal Peru, they had survived for nearly 2QOO
years completely intact as to wooden shaft, wooden hook lashed to the
rear, and bone or wooden handle-grip lashed to the front (Fig. 1). The
striking aspect was that the shape of some of the handle-grips were re-
markably similar to the stone birdstones found in late Archaic times
in the Great Lakes Area of the United States, and to which no agreed
upon function has been forthcoming. Usually catalogued by museums
or described in the literature as "ceremonial objects" or "problemati-
cals," their real function has hung in limbo. The purpose of this arti-
cle is to present my conviction that the birdstone was used by our pre-
historic Indians as an atlatl hand grip or brace.

I photographed the Mujica specimens and upon my return to the
Milwaukee Public Museum made a reconstruction. This consisted of a
one inch dowel 30 inches in length with one flat, planed surface. To
this surface, starting one inch from the front, a series of four vertical
holes were drilled positioned so that they could accommodate cords
passing through the front and rear holes found at the base of a bird-
stone specimen selected from the museum collections. The birdstone
was bound on with the head facing the distal end of the shaft (Fig. 2).
Additional lashings were passed around the shaft and over the back of
the birdstone, giving further firmness to the attachment and follow-

Fig. 1. Nazca Atlatls, Mujica Museum, Lima, Peru.

3 2 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 1

Fig. 2. Atlatl Reconstruction with Wisconsin Birdstone.

ing the Peruvian model.

As the reconstruction was to be used for testing purposes, a hook
was required to stay the end of the spear. This posed somewhat of a
problem as atlatl hooks have not been reported for the Midwest, al-
though the Kentucky type could have served as a model. About this
time, however, a curiously modified bear canine was discovered in a
cave near Platteville, Wisconsin, containing Archaic (and other) mater-
ial (Fig. 3). No atlatl hooks of this shape or material have been re-
corded, but a wooden replica was made and set into the distal end of
the shaft to test its practicality as a hook, as an additional phase of the
experiment. While it worked well in the test throwing of a spear, its
original function as an atlatl hook is, at this stage of our knowledge,
only hypothetical. The functioning of the birdstone as a grip, however,
was completely convincing in the tests. Besides being a convenient and
practical grip, it serves to steady the spear shaft before and during the
release. It can be gripped with the last three fingers over the back of
the birdstone and index finger and thumb holding spear. It also works
well with the last three fingers braced against the base of the birdstone.
(Fig. 4).

In checking the literature, I had first sought out Kellar's monograph,
"The Atlatl in North America," the only comprehensive study thus far

Fig. 4. Test throwing was done either with three fingers
over back of birdstone or braced against birdstone as shown
here.

Atlatl Handle Grip 33

done. While Kellar (p. 3 13 and 317) cites references to Kentucky and
Indiana atlatl hooks found with drilled antler sections indicative of
altatl handles, he does not deal with the problem of birdstones as han-
dles. Turning to Townsend's "Birdstones of the North American In-
dian," I was in for a surprise: a chapter titled "The Atlatl Handle Grip
Theory" picturing the Peruvian atlatls, describing his experiments with
birdstones as handle grips, with the cautious conclusion that it may be
the solution as to the use of at least some of the birdstone types. His
experiments in throwing spears with reconstructed atlatls using bird-
stones as grips satisfied him as to their usability including some of the
more unique types as the popeyed and bust. He noted that the wear in
the birdstone perforations was always at the center, never at the sides,
which one would expect from bindings subjected to a forward-back-
ward stress which would occur in atlatl use. Townsend (p. 1 69) also pic-
tures two cut bear teeth found with a birdstone in a red ocher burial in
Emmet County, Michigan, with the speculation that they may have
served as atlatl hooks. Each is different in shape and neither is the shape
of the Platteville specimen, although all three are of such form to be
functional as hooks.

There are other factors favoring the theory of the birdstone as atlatl
grip. The occurrence of the birdstone in Early Woodland, and es-
'pecially, Archaic horizons, are horizons that have provided the bulk of
evidence for the presence of the atlatl, particularly in the form of at-
latl hooks such as were found at Indian Knoll and other Kentucky Ar-
chaic sites. Then too, the relatively large percentage of birdstones bro-
ken at the neck or with broken perforations suggest that they were sub-
jected to some form of hard usage such as would occur in atlatl use. The
breakage factor certainly argues against the old theory of ceremonial
use. The argument that broken specimens may have been "killed" is
ruled out by the presence of repaired specimens and the unlikelihood of
killing have taken place on the perforations. The Peruvian analogy: the
fact that similar objects definitely served as handle grips, albeit in a dis-
tant part of the New World, would appear to offer the strongest sup-
port for a similar use of the birdstone.

In six months of confronting local and visiting colleagues with this
theory, I found that it was as new to them as it had been to me. I view
my role here not as a discoverer, but as a reviver and perhaps as a rein-
forcer of what seems to me to be the most logical theory accounting for
the function of the birdstone.

The archaeological confirmation of this theory borders on the impos-
sible. It is unlikely that an intact Great Lakes atlatl (preserved by
water?) will ever be found. About the best that can be hoped for is
that careful excavation of a Glacial Kame or Red Ocher burial site will

34 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 1

reveal a birdstone lined up with a hook in original atlatl spacing and
position.

BIBLIOGRAPHY

Kellar, James H. The Atlatl in North America. Indiana Historical Soc-
iety, Prehistory Research Series, Vol. Ill, No. 3, June, 1955.

Moore, C.B. Some Aboriginal Sites on Green River, Kentucky. In Jr. of
the Academy of Natural Sciences of Philadelphia, 2nd Series, XVI
(1916), 440-87.

Townsend, Earl C. Birdstones of the North American Indian. Privately
printed, Indianapolis, Indiana, 1959.

Uhle, Max. Peruvian Thro wing-Sticks. American Anthropologist N.S.
Vol. II, No. 4, p. 624-627.

Clovis and Sandia Points

CLOVIS AND SANDIA-LIKE POINTS

FROM DANE COUNTY, WISCONSIN

By Robert Ritzenthaler

Mr. Theodore Kouba reports the finding of four single-shouldered
points occuring as surface finds on the Kouba Site (Fig. 1). The basal
edges are ground. One has a concave base and is fluted reminiscent of
the Sandia type found at the Lucy Site, New Mexico. The other three
are somewhat similar to the Sandia II type. While these are out of the
range of the true Sandia points of the Southwest, the similarity is worth
noting. It is of some interest also that they were found on a site which
has produced a number of other surface finds of Paleo-Indian artifacts
such as Clovis points and nearly 400 gravers.

The site has also recently produced two unusually long Clovis points
(Fig. 2). They measure 5 5/8ths and 6 inches in length and are charac-
terized by flaring basal edges and extremely long fluting on both sur-
faces. They were found within three feet of one another.

Fig. 1. Sandia-Like Points.

36 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 1

Fig. 2. Clovis Points.

COMMITTEE ASSIGNMENTS
WISCONSIN ARCHEOLOGICAL SOCIETY

MEMBERSHIP COMMITTEE: Phillip H. Wiegand, Chairman.
Herman Zander, J. K. Whaley.

PUBLICITY: Tom Jackland.

FRAUDULENT ARTIFACTS: J. K. Whalcy, Chairman. Martin
Greenwald, Robert Hruska, Neil Ostberg, Dr. Robert Ritz-
enthaler. Paul Scholz, Phillip H. Wiegand.

SURVEYING AND CODIFICATION: Paul Koeppler, Chairman.
Darryl Be land, Elmer Daahnann, Wayne Hazlett, Ernest
Schug.

PRESERVATION OF SITES: Robert Hruska, Chairman. Dr.
Richard Peske, Dr. Robert Ritzenthaler. (Chairman has op-
tion to make further appointments.)

EDITORIAL: Dr. Robert Ritzenthaler, Chairman. Dr. David A.
Baerreis, Dr. Joan Freeman, Dr. Lee Parsons, Dr. Melvin
Fowler.

PROGRAM: Dr. Diehard Peske, Chairman. Paul Turney, Mrs.
P. H. Wiegand, Dr. Robert Ritzenthaler.

LAPHAM RESEARCH MEDAL. Dr. Robert Ritzenthaler,
Chairman, Dr. David A. Baerreis, Phillip H. Wiegand.

THE CHARLES E. BROWN CHAPTER

(Meets second Tuesday, 7:45 P. M., State Historical Society,
October thru May)

President: James Stoltman
Secretary: Peter Storck
Treasurer: Lathel Duffield

THE FOX VALLEY CHAPTER

Oshkosh

(Meets second Monday, 7:30 P. M., Oshkosh Public Museum,
September thru May)

President: Elmer Daalmann

Vice President: Richard Mason

Secretary.' Robert R. Jones

Treasurer: Nancy L. Hoist

THE WISCONSIN
RCHEOL06UT

THE BELOIT MOUND GROUPS: FIFTY YEARS LATER 37

by Frederick W. Lange and Mark W. Kristensen

PIONEER ARCHEOLOGY IN AN ILLINOIS MIDDLE 49

WOODLAND MOUND by Stuart Stuever

EVIDENCE SUPPORTING THE THEORY OF CLIMATIC 57

CHANGE AND THE DECLINE OF HOPEWELL by

D. Vickery

ANOTHER RADIOCARBON DATE FOR THE OCONTO 77

SITE by Robert Ritzenthaler

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee, Wisconsin
Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. M., Milwaukee Public
Museum, September thru May

OFFICERS

PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Ostberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Green vvald .

TREASURER

Wayne J. Hazlett

3768 S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct., Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr. Robert . Ritzenthaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL

MEMBERSHIP
The Wisconsin Archeologist is distributed to members
as part of their dues.
Annual Members, $5.59

All communications in regard to the Wisconsin Archeologica! Society
and contributions to the Wisconsin Archeologist should be addressed to
Mrs. Edward Flaherty, Secretary, N62-W15127 Tepee Ct., Menomonee
Falls, Wis. 53051. Entered as Second Class Matter at the Post Office at
Lake Mills, Wis., under the Act of August 21, 1912. Office of Publica-
tion, 316 N. Main St., Lake Mills Wis. 53551.

THE WISCONSIN ARCHEOLOGIST

New Series

MILWAUKEE, WISCONSIN - JUNE, 1970
Published Quarterly by The \Visconsin Archeological Society

THE BELOIT MOUND GROUPS: FIFTY YEARS LATER

Frederick W. Lange Mark W. Kristensen

University of Wisconsin-Madison Beloit College

A resurvey of the Beloit-Rock River series of mound groups described by
Ira M. Buell (1919) was made to select possible sites for excavation and to de-
termine the loss of mounds in the Beloit area in the past fifty years. Buell 's work
built on previous surveys by the Rev. Stephen Peet (1 898) and by Robert Becker
(1913) and ended with the following comments:

The permanent preservation of some of the distinctive mound groups in this
region is much to be desired and should be undertaken by the cities of Be-
loit or Rockford or by the counties or townships in which they are located.
The rate at which these ancient monuments have been disappearing makes
this now imperative (p. 149).

The recent survey allows description of the present condition of the Beloit-
Rock River and Turtle Creek mound series, with a corresponding inference to
the impact of Buell's admonition. The previously unreported excavation of one
mound, the Yost mound, is also presented.
THE BELOIT-ROCK RIVER SERIES: Township 1 N Range 12 East Beloit

Township, Rock County, Wisconsin.

The Beloit College Group (Ro-15) NWfc of SWtt of Section 36
The original map of this group, published in Lapham's "Antiquities of Wiscon-
sin" in 1855, shows 14 conical mounds and 5 linears. Buell notes the destruc-
tion of one linear and the addition of a previously unmapped turtle and linear
mound. The most recent report on the Beloit College group was by Bastian
(1958). His map (see Figure 2) shows a total of 27 mounds: 6 linear, 13 coni-
cal, 6 oval, and 2 turtle effigies. Of these, 1 turtle, 1 linear, and 2 conical
mounds have been destroyed. Two conicals, 2 ovals, and 1 linear have been ex-
cavated and are discussed in Bastian 's report. Mounds in the immediate vicinity
of the campus have been added to the group through acquisition of properties
by the College. Recent building by the College has not damaged remaining
mounds, nor have commercial excavations revealed additional indications of
prehistoric activities, such as habitation areas.

38 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 2

Figure 1. Map of Beloit area, showing mound groups surveyed: 1) Roth Group;
2) Yost Mound; 3) Weirick Group; 4) Henderson Mound; 5) Poe Mound; 6) Ad-
ams Group; 7) Water Tower Group; 8) Beloit College Group; 9) Baldwin Group;
10) Joint Switch Group; 11) Stone Mill Group; 12) State Line Group; 13) Hill-
crest Group; 14) Strong Group; 15) McLenegan Group; 16) Murphy Group; 17)
Lathers Group; 18) Pierce Group; 19) Baarz Group.

Water Tower Group (Ro-19) NEtt of NEtt of Section 35
The presence of this group was reported by Rev. Peet. By Buell's time the
mounds were only vague outlines and could no longer be associated with defi-
nite shapes. Even these vague outlines are no longer visible; the only suggestion
of a mound is a small rise between two oak trees on the crest of the hill.
Adams Group (Ro-33) SE& of NE 1 A of Section 26

The mound group is located at an area known as Pageant Park, on the north
end of a tract belonging to Fairbanks, Morse Co. The original group consisted of
13 members: 3 turtle mounds surrounded by 4 conicals and 6 oval and linear

The Beloit Mound Groups

mounds. Bull-dozing and reworking of the terrace have eliminated traces of most
of these. One turtle and one oval mound are still recognizable (Figure 3.).

Poe Mound (Ro-34) SW& of NE& of Section 26

Directly across the Rock River from the Adams group was a small conical mound

less than twenty feet in diameter and about one and one-half feet high, from

.27

BELOIT COLLEGE
MOUA/D CROUP ft IS)

PUNE T/IBU SUAVErZD, /7Sf
Cawrovm BASCD o/vsgftRr sawf-nMAf

US.GS

SCALE -rctr

COLLEGE

Figure 2. Map of Beloit College Group (from Bastian, 1958). Dotted lines ind
cate mounds which have been destroyed. Mounds 2, 3, 4, 5 and 6 have been e?
cavated.

40 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 2

ADAMS GROU P

Figure 3. Adams Mound Group. Blackened mounds survive.

The Beloit Mound Groups

which seven burials were removed during the 19th century. This mound had
already been obliterated by the time of Buell's survey.
Henderson Effigy (Ro-27) SE 1 ^ of SE& of Section 23

Half a mile north of the Adams group, on the edge of the terrace, is a solitary
effigy mound. The mound appears to be of the turtle variety and Buell des-
cribed it as being the central figure in the Rock River series. This mound is ex-
tant, being located in the front yard of a private residence.
Weirick Group (Ro-30) Online between NE& and SE& of Section 23
The original group consisted of three clusters totaling 1 5 mounds. The north-
ernmost cluster consisted of 1 dumbell-shaped mound, 2 bearlike mounds, 2
linear mounds, 1 conical mound and 1 incomplete effigy figure, possibly a tur-
tle. South of this cluster were 3 linear mounds, 2 being oriented in parallel east-
west positions and the third lying slightly to the south and approaching the other
two at a north-eastward angle. East of this cluster, on the second terrace of the
nver, were 2 turtle mounds, 1 incomplete turtle figure (dumbell), 1 oval mound,
and 1 conical mound. Two linears from the second cluster and the conical and
the dumbell mounds from the third cluster remain (Figure 4).
Baldwin Group (Ro-28) NE# of SE& of Section 14

While Buell reported that one of the three conical mounds mapped by Wilson in
"Prehistoric America" (p. 146) was still visible, examination of this area in

H \

ROCK RIVER
Figure 4. Weirick Group. Blackened mounds survive.

42 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 2

1968 yielded negative results.

Roth Group (Ro-21) SW& of SW& of Section 1

This group was located by Charles Hyde of Beloit College Class of 1914. Two

distinct oval mounds and faint traces of others were located along the terrace of

the Rock River where it changes its course from north-south to east-west and

widens into a broader vallet just north of the township line. No mounds were

visible when this area was re-examined.

Yost Mound (Ro-23) NW& of NE& of Section 14

This mound was reported by Buell as being a solitary linear mound 80 feet long
and 16 feet wide (1919:126). It lies atop and near the western edge of a rela-
tively low bluff, approximately one mile north of the Baldwin group and one-
eighth of a mile east of the Rock River. Allen Wells, a student at Beloit College,
excavated this mound in the fall of 1963. The following summary is presented
from his unpublished notes on the excavation:

Figure 5. The Yost Mound.

The Beloit Mound Groups

Wells described the mound as being 22 feet wide and 69 feet long. The dis-
crepancy in distances between this and BuelTs observation is apparently due to
gravel borrowing from the edge of the bluff with subsequent erosion on the
south and west sides of the mound. No other disturbance was noted in the
mound. The orientation of the mound is along a north-south line. A 27-foot by
3-foot trench was excavated in 3-foot square units across the north end of the
mound. Excavation revealed that the original humus line has been removed prior

STATI LINE MOUND GROUP (Ro-39)
(AFTER B U E I l , I 9 1 9)
DESTROYED

A-C-D-E EXCAVATED, 196>
B UNEXCAVATED

Figure 6. The State Line Mound Group.

44 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 2

to construction of the mound and that the mound was constructed directly
on the outwash gravels. The mound fill consisted of a sandy soil inter-
spersed with rock fragments and pebbles.

Excavation produced three artifacts: two bone fragments, both broken and
poorly preserved which Wells tentatively identified as human; and one chert pro-
jectile point, the base of which was missing.

The Yost Mound is illustrated in Figure 5.

TURTLE CREEK MOUND GROUPS: Township 1 N Range 12 East, Beloit
Township, Rock County, Wisconsin.

State Line Mound Group (Ro-39) SE& of the SE 1 ^ of Section 36
This group at one time traversed the Illinois- Wisconsin state line. The original
group contained 8 mounds. The two northern-most members of the group, a
linear mound and a turtle mound, on the Wisconsin side of the line, were des-
troyed by undermining of the second outwash terrace by Turtle Creek. The third
northern-most mound, an oval on the Illinois side, was destroyed by road con-
struction. The 5 remaining Illinois-side mounds, 1 turtle, 2 conicals, and 2 ovals
were found in good condition. These mounds, except for the effigy, were ex-
cavated by the authors and other students from Beloit College during the sum-
mer of 1967 (Lange 1968). (See Figure 6).
Stone Mill Group (Ro-38) NEtt of SW& of Section 36

This group originally consisted of 2 turtle mounds and 4 conicals, of which none
remain.

Hillcrest Group (Ro-147) NWfc of SW& of Section 31

This group lies across the bottoms of Turtle Creek from the State Line Group.
At the time of Buell's survey, it consisted of six mounds: one tapering linear,

o O

Figure 7. The Hillcrest Mound Group. Blackened mounds survive.

The Beloit Mound Groups

two turtle effigies, and three conicals. The linear and one turtle were located on

the east side of the railroad tracks; the former and the eastern half of the linear

remain. The three conicals and one turtle to the west of the tracks have been

destroyed. (See Figure 7).

Joint Switch Group (Ro-142) NE& of SEtt of Section 25

This was the largest known group in the Beloit area. As originally surveyed by

the Rev. Peet, there were 39 mounds none remain. Five of these were panther

effigies, two were effigies of an undetermined nature, four were linear mounds,

with the rest being conical and oval tumuli. Resurvey of the area indicated that

the mounds in the southern portion of the group (those lying close to the first

JOINT SWITCH GROUP
Figure 8. The Joint Switch Group. Completely destroyed.

46 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

terrace of Turtle Creek and south of the present gravel-mining operations) were
destroyed by the installation of a large subterranean concrete structure. The
northern portion of the group, farther back on the terrace, was in the area which
has been excavated for commercial purposes by the gravel company. An arc of 9
conicals to the west was destroyed by residential development (See Figure 8).
TURTLE TOWNSHIP (T 1 N R13E Rock County, Wisconsin)
Strong Mound Group (Ro-143) NE& of SWfc of Section 30
A group of four mounds consisting of 2 ovals, 1 conical, and 1 turtle effigy was
reported by Buell. The first three were partially destroyed by the construction
of Arrowhead Drive; the eastern edges are vaguely visible. The effigy mound has
been completely destroyed by suburban development.

McLenegan Group (Ro-141) NWH of SE!4 of Section 19

One large mound 45' x 60' is described by Buell; a mound known to be to the

east of this oval was reported to already have been destroyed. The large mound

survives in the frontyard of a farm house.

Murphy Group (Ro-140) SE& of SE& of Section 18

A group of seven conical mounds in this location were described by Buell. None

remain.

Lathers Group (Ro-139) NEfc of SW% of Section 1 8

This group originally consisted of 2 linear mounds, 1 conical mound, and 1 oval

mound. Examination of the area indicated that all had been destroyed.

Baarz Mounds (Ro-136) NE& of SE& of Section 7

Buell described this group as containing 3 linear mounds, forty-five feet, sixty

feet, and ninety feet in length. None survive.

Pierce Group (Ro-138) SEfc of NWH of Section 17

This group consisted of 3 oval mounds, 1 linear mound, and 1 panther mound. A

housing development along the creek terrace has destroyed any traces.

SUMMARY:

This survey covered the location of 1 27 mounds reported in Buell's survey of
1919 and represents the main groups in the immediate Beloit area. Of these
127 mounds, 41 remain at the present time, a loss of 67.7% from these major
groups. Except for the Beloit College group, no effort has been made either to
document or to preserve mounds within the city or area. Buell's admonition for
the preservation of prehistoric remains, evidently unheard, should be re-empha-
sized. In addition to the remaining mounds, this is especially true of possible
village/habitation areas which are equally liable to destruction and even less vis-
ible on the surface. It is these sites that will provide significant non-ceremonial
information.

What conditions permit these losses? Are academic or civic leaders responsi-
ble for the conservation of these prehistoric resources as a part of the public

The Beloit Mound Groups 47

trust they have assumed? As they are often involved with the forces of expan-
sion and development that so often bring about the unstudied destruction oi
prehistoric remains, academic and civic agencies should become more prac-
tised in their latent role as custodians of this irreplacable scientific wealth that is
rapidly diminishing. As Fred Wen dor f wrote:

Conservative forecasts of future population and land use indicate that we
may expect the population to increase at the rate of 1 8 to 20% per DECADE.
But, even more significantly for archaeology, that portion of the population
living in the urban areas is expected to grow from the 70% of today to 85%
by 1980. The result will be an expansion of the urban areas laterally, so that
within the next 20 years the cities will sprawl over the remaining nonurban
sectors of the valleys and the foothills, the only areas suitable for either mod-
ern or aboriginal habitation and the major areas where our remaining arch-
aeological resources occur (1963:288).

While it is true that Wendorf is referring specifically to the American South-
west, the same basic problems and needs are applicable to other areas.

There is a need for academic institutions to show an increased awareness of
the vestiges of prehistory within their individual regions of influence, and to
combine with local, state and federal resources to preserve the material evidences
of prehistoric America.

Aside from the salvaging of data from threatened locations, there is an impor-
tant need for the maintenance and conservation of archaeological resources. This
pressing current problem is perhaps even more critical for coming generations, so
that there are sites for them to explore with as yet undeveloped techniques of
excavation and analysis which this scientific age can provide for archaeologists.

ACKNOWLEDGEMENTS:

The authors would like to recognize the assistance of Dr. Joan Freeman of the
State Historical Society of Wisconsin; of Mr. David A. Trapp, who assisted in
portions of the survey; and Mr. Richard Rhey, who prepared Figure 7.

BIBLIOGRAPHY

Bastian, Tyler

1958 The Beloit Mound Group (Ro-15), A Preliminary Report. WISCON-
SIN ARCHEOLOGIST, Vol. 39, No. 3, pp. 1 55-1 71 . Milwaukee

Becker, Robert H.

1913 Turtle Creek Mounds and Village Sites. WISCONSIN ARCHEOLO-
GY. Old Series, Vol. 1 2, pp. 7-20, Milwaukee.

48 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

Buell, Ira M.

1919 . Beloit Mound Groups. WISCONSIN ARCHEOLOGIST. Old Series,
Vol. 18, No. 4, pp. 119-151. Milwaukee.

Lange, Frederick W.

1968 The Excavation of the State Line-Mound Group (Ro-39), Beloit,
Wisconsin. WISCONSIN ARCHEOLOGIST. Vol. 49, No. 3, pp. 109-
125. Milwaukee.

Peet, Stephen B., Rev.

1898 PREHISTORIC AMERICA, VOLUME II. Chicago: American Anti-
quarian.

Wells, Allen

1 963 Unpublished notes on Yost Mound excavation. In files of Logan Mu-
seum of Anthropology, Beloit College.

Wendorf, Fred

1963 Archaeology and Private Enterprise: A Need for Action? AMER-
ICAN ANTIQUITY. Vol. 28, No. 3, pp. 286-288. Salt Lake City.

Wilson, James Jr.

1898 Map published in Stephen Peet's PREHISTORIC AMERICA, VOL-
UME II.

Pioneer Archeology In Illinois 49

PIONEER ARCHEOLOGY IN AN ILLINOIS
MIDDLE WOODLAND MOUND

By Stuart Struever
Dept. of Anthropology
Northwestern University

William McAdams was a pioneer archeologist who lived and excavated most
of his life in the area near the confluence of the Illinois and Mississippi rivers, a
few miles north of St. Louis. McAdams was an intrepid archeologist who, like
others of his time, was fascinated with the problem of understanding the ori-
gins of the people and culture responsible for the vast number of earth mounds
strung along the major river valleys south of the Great Lakes. His book, RE-
CORDS OF ANCIENT RACES IN THE MISSISSIPPI VALLEY (1887), along
with others of his publications, including "The Ancient Mounds of Illinois" and
"Antiquities of Cahokia," were explorations of this problem. For almost 40
years, beginning prior to the Civil War, McAdams excavated innumerable mounds
and habitation sites perhaps more than we will ever know in the lower Illinois
and adjacent Mississippi Valleys near St. Louis in an attempt to throw some
light on ". . .The Prehistoric Races of America, With Some Suggestions As To
Their Origin" (McAdams. 1887:i).

Between 1881 and 1890, the Bureau of Ethnology of the Smithsonian In-
stitution conducted a "Mound Survey" throughout the eastern woodlands of
the U.S. This Survey was directed by Cyrus Thomas who later published a sum-
mary of this work (Thomas, 1894). William McAdams was employed in the fall
of 1882 as one of Thomas' archeological assistants and for a time carried on
mound explorations in the St. Louis area under Bureau of Ethnology sponsor-
ship. However, both prior to and after his affiliation with the Bureau's "Mound
Survey." McAdams carried out independent excavations, largely in mounds lo-
cated in the Illinois aad Mississippi Valleys in the area near the confluence of the
two rivers.

This paper deals with one of these investigations.

McAdams established close ties with Major. J.W. Powell, Director of the Bu-
reau of Ethnology, to whom he wrote regularly about his discoveries. Some of
the McAdams-Powell correspondence is currently housed in the U.S. National
Museum. *In one of the unpublished letters to Major Powell, dated January
14, 1881, McAdams describes the results of a specific mound excavation:

'The writer is indebted to Mrs. Margaret Blaker, Archivist of the Department of
Anthropology of the U.S. National Museum, for "excavating" this and other
McAdams letters from the Museum Archives.

50 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

Otterville, Ills.
Jan. 14th, 1881

My Dear Major

I have just come in from a digging expedition, had to heave the dirt up a little
lively with the thermometer 10 and 1 5 below zero.

Explored two good mounds the contents of which may interest you.

The first is situated in the bottom, on the bank of the Illinois river, in Cal-
houn Co. Ills. It is oval in shape, truncated, and 1 75 feet in length on top, and
50 feet wide on top, sloping [plan view sketch of mound showing mound dim-
ensions and location of the excavation] gradually to the level below.

Commencing in the centre we carried an excavation down and at the depth
of 16 feet came to a shallow basin of clay, made hard, perhaps by fire. This basin
was filled with white sand, covered up in the sand was, perhaps a waggon load of
raw material for implements, flint, quartz, jasper, chalcedony, and all choice
pieces, some of which were blocked out in the shape of the implement or orna-
ment intended. The flint is very peculiar, being a dark brown and all broken
from large nodules, three of which, weighing from 20 to 40 Ibs, we found entire.
These nodules are white outside and dark brown within, resembling the chalk
nodules.

Laying on the sand were three large sea shell vessels, of the capacity to hold
one to two quarts.

[Sketches of two sea shells and a burial pit containing 3 skeletons]

In a sitting posture on the sand, at the north end of the basin were two skele-
tons, and old man, from the teeth and the obliterated sutures of the [?] and a
young woman with beautiful even teeth and delicate bones. About the neck of
the old man was a string of peculiar beads, made of shell. Some of these were
round [sketch of round bead] , others cylindrical, [sketch of three cylindrical
beads] , but a majority of them were cylindroids with the perforations in the
ends coming out laterally, [sketch of 3 cylindrical beads showing perforation
holes] On the breast of the old man was a plate of copper, circular in shape and
some 4 inches across [sketch of circular object with hole drilled in center] and
with it another curious copper implement or ornament like a short spool, being
two disks two inches across connected by a short tack in the centre, [sketch of
bicymbal ear spool] These were much corroded. At the left-hand of the man
was a very beautiful pipe of mottled catlinite, polished and smooth as glass, one
end of the flat, curved base being perforated to form a stem. It is a splendid
specimen of aboriginal workmanship without a scratch or mar upon it. [side
view sketch of Hopewell plain platform pipe].

The woman had no ornaments about her, nor was anything of this nature
found with a skeleton at the south end of the basin. Ashes were found about the
edges of the basin, showing that fire had been used in the burial ceremony.
Further exploration in the mound may reveal something more, but I am in-

Pioneer Archeology In Illinois

clined to believe that at the burial of this old mound-builder at least two of his
subjects were buried with him. . .

Wm. McAdams Jr [signed]
Otterville, Ills.

The above portion of this remarkable letter from McAdams to Major Powell
is actually a valuable descriptive report of a mound excavation; a somewhat re-
vised and shortened account of this same excavation was published in the Smith-
sonian Institution ANNUAL REPORT OF 1882. (McAdams, 1884:684). As
noted by the bracketed phrases above, McAdams' letter is liberally sprinkled
with skillfully rendered ink sketches of artifacts. Figures 1 and 2 present a few
of these sketches. They leave no doubt that McAdams' 1881 excavation reached
a central burial feature in a Hopewell Mound. Depicted in the letter are two of
at least three marine shell (BUSYCON sp?) containers (Fig. 1), one circular
copper breast plate, one copper bicymbal ear spool and one plain platform pipe
(Fig. 2), all from the central burial feature. In addition, McAdams describes
"a waggon load" of artifact preforms and perhaps cores of "flint, quartz, jasper,
chalcedony." His description of some of the flint nodules in this cache strongly
suggests they are examples of the blue-grey "hornstone" which weathers out of
Menard limestones in Union County in southwestern Illinois and elsewhere
(Fowke, 1928:530-532). This is the same flint referred to as "Dongola" by

Figure 1. Marine shell containers from Merrigan Mound as sketched by William
McAdams in unpublished letter to Major J.W. Powell, January 14, 1881.

52 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

Figure 2. Plain platform pipe from Merrigan Mound as sketched by William
Me A dams in unpublished letter to Major J.W. Powell, January 14, 1881.

Winters (1967:54).

The "catlinite," or more likely Ohio pipestone, pipe depicted in McAdams'
letter and shown in Figure 2 is the characteristic "smokestack variety" plain
platform pipe that is diagnostically Hopewell and found in large numbers at the
Tremper (Mills, 1916) and Mound City (Mills, 1922) sites in south-central Ohio.
Struever and Houart (n.d.) briefly define this Hopewell pipe style and discuss
its distribution.

Both the copper bicymbal ear spool and plain platform pipe are Middle Wood-
land horizon markers in the eastern U.S., establishing this Illinois Valley mound
as Hopewell. This interpretation is supported by the occurrence of marine shell
containers in association with the pipe and ear spool. The discovery of a large
cache of preforms and cores of quartz, "hornstone" (?), and other exotic stones
in a prepared pit with a white sand matrix is also suggestive of Hopewell.

In 1950 almost 70 years after the McAdams excavations- two amateur arch-
eologists, Walter Wadlow and Gray LaDassor, excavated the Merrigan Mound in
the lower Illinois Valley floodplain. Merrigan is located 300 yards west of the
Illinois River and four miles south of the town of Hardin (Wadlow and LaDas-
sor, 1951). This mound, and perhaps others now plowed down, along with an
adjacent habitation site are located 35 miles north of St. Louis in Calhoun
County, Hardin Township, Section 14.

Wadlow and LaDassor recovered 630 "roughly chipped blanks," over 1 ,000
"chunks" of hornstone and novaculite, one broken ornament "made from the
jawbone of some animal belonging to the cat family," as well as four pins or
skewers made of the long bones of what is probably the white-tailed deer, all
from a central feature at the base of the mound. In addition to these mate
possible Hopewell diagnostics included:

a) a prepared sand floor; b) a centrally located sub-floor burial pit; c) three

Pioneer Archeology In Illinois 53

superimposed mounds over tliis burial pit; and d) a layer of limestone slabs
covering the primary mound.

Wadlow and LaDassor (1951) encountered quantities of habitation debris in
the fill of the Merrigan Mound. They point out that the decoration and temper-
ing of the Merrigan sherds are closely similar to sherds from Snyders (Griffin
n.d.) and Sconce-Schudel, two well-known Havana-Hopewell habitation sites in
the area.

In July, 1960, the writer collected 66 Havana and two Hopewell series sherds
from the surface of the Merrigan Mound and from areas immediately to the east
and north of the mound. Nine of these sherds were classifiable as either Naples
Straight Dentate Stamped or Havana Zoned Straight Dentate Stamped, while one
was Naples Ovoid Stamped (cf. Griffin, 1952). This and other surveys indicate a
definite Havana-Hopewell habitation site bordering the east and north sides of
the Merrigan Mound, while similar debris littering the mound surface suggests
that the moundfill itself contains Havana-Hopewell occupational debris. Careful
study of the descriptions of both the McAdams and Wadlow-LaDassor excava-
tions strongly suggests that both pertain to the Merrigan Mound.

The following evidence supports this interpretation:

1) William McAdams' 1881 letter to Major J.W. Powell indicates the mound
was oval in plan view and flat-topped ( a plan view sketch of the mound is in-
cluded in the letter). This is essentially the shape of the Merrigan Mound as il-
lustrated by Wadlow and LaDassor (1951).

2) McAdams describes the mound as 1 75 feet long and 50 feet wide; Wadlow
and LaDassor indicate that Merrigan is 1 75 feet long by 75 feet wide. The ad-
ditional 25 foot width of the Merrigan Mound may be accounted for by spread-
ing of the mound soils through repeated plowing across the lateral axis of the
earthwork during the past half century.

3) McAdams describes tunnelling the mound, beginning in the approximate
center of its top. Wadlow and LaDassor discovered that the Merrigan Mound had
been previously tunnelled, and thus tunnel began in the top of the mound near
the center. They also point out that shoring timbers used by the earlier ex-
plorers had almost completely decayed, indicating that considerable time had
elapsed since the original tunnelling. In these shoring timbers they found "old
wrought iron nails," again suggesting that this tunnel dates to at least the 19th
or very early 20th century.

4) McAdams indicates that the central burial feature was located at a depth of
16 feet below the mound surface. Wadlow and LaDassor give this distance as
15 feet.

5) McAdams recovered a copper breast plate and a copper earspool in burial
context. Wadlow and LaDassor recovered copper stained bones, but no copper
artifacts.

6) McAdams discusses finding a "waggon load" of chert and chalcedony nod
ules, cores and partially worked artifacts or "blanks." Wadlow and LaDassc

54 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

found the same range of raw material, cores and "blanks" made from a variety
of exotic raw materials. They indicate removing about 1650 pieces of stone,
again reminiscent of the massive quantity described in Me Adams' 1881 letter,
Both McAdams and Wadlow-LaDassor found the mass of raw material and arti-
facts in the central burial feature. Apparently, McAdams found this material but
did not remove it; it was later rediscovered and removed by Wadlow and La-
Dassor.

7) McAdams (1884) in his later published report on this excavation, indicates
that the mound was located in the "Illinois River bottom, 15 miles from the
mouth." He also notes that it was in Calhoun County. The Merrigan Mound
is located in the bottomlands of the Illinois Valley, in Calhoun County, and al-
most exactly 1 5 miles upstream from the river mouth.

In sum, it appears that both McAdams and Wadlow-LaDassor excavated the
Merrigan Mound. Using both McAdams' unpublished letter to Major Powell and
the brief report by the latter excavators, it is possible to reconstruct a picture of
the mound structure and contents.

The prehistoric builders stripped away the top soil from a small area and here
prepared a "clay-lined" pit, 25 to 30 feet long, and 20 feet wide. The pit was
filled with "white" sand in which at least 1,000 nodules, cores, preforms, etc. of
flint, quartz, chalcedony, jasper and perhaps novaculite were buried. At least
three articulated human skeletons were laid on top of this sand-filled pit. The
aforementioned marine shell, copper and other artifacts were associated with
these burials. Both McAdams and Wadlow-LaDassor indicate that much of the
flint recovered at Merrigan was the well-known southern Illinois-Indiana horn-
stone. McAdams' 1881 letter describes the distinctive hornstone nodules very
clearly. It is quite likely that many, if not all, of the 630 bifacially chipped
"blanks" referred to by McAdams and removed by Wadlow and LaDassor were
the hornstone disks that are a familiar mortuary association in Hopewell sites in
Ohio, Illinois and elsewhere. These "blanks" were either in the sand pit fill or im-
mediately on top of it with the skeletons.

Both McAdams and the later excavators describe evidence of fire around the
edges of this central burial feature.

Primary and secondary conical mounds were then built over the central fea-
ture. No burials were found in the fill of either mound. A layer of limestone
slabs was placed over the primary mound. Finally, two tertiary mounds were
built, one slightly to the south, the other to the north of the central feature.
These two mounds overlap at the feature, giving this tertiary earthwork an oval
or elliptical plan view. Wadlow and LaDassor (1951) report discovering three in-
trusive burials in the two tertiary mounds.

Except for the three intrusive burials, the only skeletons recovered from the
entire 1 75 by 50-foot Merrigan Mound came from the central feature which also
yielded ALL of the aforementioned artifacts.

The concentrating of skeletons in a central feature in association with cere-

Pioneer Archeology In Illinois 55

monial, or at least non-subsistence technology artifacts is another diagnostic
characteristic of SOME Hope well burial mounds in the Havana and Crab Or-
chard traditions in Illinois (Struever, 1965).

Evidence from the Wadlow-LaDassor excavations, along with surface surveys
by the University o Michigan and the writer, indicate Havana-Hopewell and
Rke-Hopewell habitation sites located immediately adjacent to the Merrigan
Mound, and the mound itself appears to have been built of soil frpm these sites.

Finally, this brief report calls to our attention the likelihood that the U.S.
National Museum in Washington represents a potential reservoir of unpublished
information on the excavations of the Bureau of Ethnology "Mound Survey"
and other 19th century archeological projects in the eastern United States. The
Smithsonian, as parent institution for the Bureau and the focal point of much
late 19th century science in this country, was the organizer of and advisor to in-
numerable pioneer archeological undertakings. It also became the recipient of
the collections and records of these projects, many of which-as in the case of
McAdams' excavation of the Merrigan Mound remain to be analyzed and des-
cribed in the literature.

"In April, 1948, an expedition of the University of Michigan, Museum of An-
thropology, under the direction of Dr. James B. Griffin (personal com-
munication) recovered 1 38 Havana and 23 Hopewell series sherds from the sur-
face at Merrigan.

REFERENCES CITED

Fowke, Gerard

1928 "Archaeological Investigations II," Bureau of American Ethnology,
44th ANNUAL REPORT; pp. 530-532. Washington.

Griffin, James B.

1952 "Some Early and Middle Woodland Pottery Types in Illinois," in

HOPEWELLIAN COMMUNITIES IN ILLINOIS (Thorne Deuel,

ed.), Illinois State Museum, SCIENTIFIC PAPERS, Vol. 5, No. 3;

pp. 93-129. Springfield,
n.d. "The Snyders Site, Calhoun County, Illinois," Greater St. Louis

Archaeological Society; St. Louis.

56 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

McAdams, William Jr.

1884 "Mounds of the Mississippi Bottom, Illinois," Smithsonian Institu-
tion, ANNUAL REPORT FOR 1 882; pp. 684-686. Washington.

1887 RECORDS OF ANCIENT RACES IN THE MISSISSIPPI VAL-
LEY. C.R. Barns Publishing Co., St. Louis. 120 pp.

Mills, William C.

1916 "Exploration of the Tremper Mound," in CERTAIN MOUNDS AND

VILLAGE SITES IN OHIO, Vol. 2, No. 3; pp. 105-240. Columbus.
1922 "Exploration of the Mound City Group," OHIO ARCHAEOLOGI-
CAL AND HISTORICAL QUARTERLY, Vol. 31, No. 4; pp. 422-
584. Columbus.

Struever, Stuart

1965 "Middle Woodland Culture History in the Great Lakes-Riverine
Area," AMERICAN ANTIQUITY, Vol. 31, No. 2; pp. 211-223.
Salt Lake City.

Struever, Stuart and Gail L. Houart

n.d. "An Economic Analysis of Hopewell," in PREHISTORIC ECO-
NOMIC AND SUBSISTENCE PRACTICES (Edwin M. Wilmsen,
ed.) IN PRESS.

Thomas, Cyrus

1 894 "Report of the Mound Explorations of the Bureau of Ethnology,"
Bureau of Ethnology, 12th ANNUAL REPORT. Washington.

Wadlow, Walter and Gary LaDassor

1951 "Excavation Merigan Mound, Calhoun County, Illinois," Greater
St. Louis Archaeological Society, BULLETIN No. 6; pp. 19-23. St.
Louis.

Winters, Howard D.

1967 "An Archaeological Survey of the Wabash Valley in Illinois," Il-
linois State Museum, REPORT OF INVESTIGATIONS No. 10.
Springfield.

Theory of Climatic Change 57

EVIDENCE SUPPORTING THE THEORY OF
CLIMATIC CHANGE AND THE DECLINE OF HOPEWELL

By Kent D. Vickery
Indiana University

The decadent character of early Late Woodland cultures as compared with the
cultural florescense of Hopewell has long been recognized in the prehistory of
the eastern United States. Among the theories which have been proposed con-
cerning the decline of Hopewell, Griffin (1952:361) mentions the following:
"cultural fatigue" (a concept borrowed from Kroeber involving a kind of cul-
tural "stagnation"^; the continuing destruction and burial of fine art products;
and the disintegrating effect of encroaching Mississippian peoples.

Other attempts have since been made to account for the decline of Hopewell.
Prufer (1964:66) mentions the possibility of a plague. He also believes that hill-
top enclosures in Ohio were "terminal Hopewell" fortifications built for protec-
tion against invading Fort Ancient peoples. Elsewhere, Prufer (1965) has sug-
gested that "certain disturbances" disrupted the elaborate Hopewell exchange
network, and as a consequence, "the system itself, like a house of cards, began
to collapse" (Prufer, 1965:136). Wray and MacNeish (1961:67) see the decline
of Hopewell as the result of "internal political differences, or, possibly, a reas-
sertion of local tribal independence after the use of agriculture had become fam-
iliar and trustworthy."

Most of the theories mentioned above are speculative, and remain untested by
either contradictory or supporting evidence. Inferences such as cultural fatigue
and internal political differences cannot easily be made from archaeological evi-
dence alone, and explanations involving plagues or defensive measures against
hostile cultures are simply unsubstantiated. Prufer's suggestion that the Hope-
well exchange network was disrupted is probably true, but no explanation is
offered as to what might have caused this.

In a later article, Griffin (1960a) recognized the unsatisfactory nature of
earlier theories attempting to explain the decline of Hopewell, and proposed in-
stead the ecological explanation which is summarized below. *

The present paper suggests that a minor phase of cooler climate in the nor-
thern Mississippi Valley and Great Lakes areas was a significant contributing
factor to a decline in the reliability of agricultural products, which in turn re-
sulted in the decline from Hopewell, or Hope well-Woodland, to early Late
Woodland cultural forms (Griffin, 1960a:21-2).

Specifically, Griffin (1960a:28) noted that "The gradual decline and demise
also Griffin, 1960b, 1961a, 1961b, 1965 and 1967.

58 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

of the Hopewell culture in the Ohio and northern Mississippi Valley appears to
correlate well with the cold period from around A.D. 200 to A.D. 700."

Griffin proposed the theory of climatic change as a major factor contributing
to the decline of northern Hopewell only. A comparable climatic change cannot
be documented in the southeastern United States, and it is significant to note
that the southern expressions retained Hopewellian traits as they evolved into
the later Weeden Island and related' culture complexes.

Most of Griffin's evidence for climatic deterioration is derived from data in
northern Europe, where paleoclimatologists have been able to reconstruct cli-
matic sequences with a great deal of accuracy by the use of historic records.
Such precise climatic reconstruction is not possible in the eastern United States,
where pollen diagrams, some of which are radiocarbon-dated, must be relied on
in the absence of written records. Griffin realized, however, that valid interpre-
tations of climatic events in the northern Mississippi Valley must rely on these
pollen studies. In attempting to use palynological data from this area, he reports
the following major difficulty:

Most of the American palynologists have been so intrigued with demonstra-
ting the major climatic shifts of the post glacial period that they have not
spent much time on any less apparent changes of the last 3000 to 4000
years (Griffin, 1960a:29).

Since this article was written, however, a number of relevant pollen studies
have been made. It is to this more recent information that I have turned for
data with which to re-examine Griffin's hypothesis.

EVIDENCE FOR MAIZE AGRICULTURE IN HOPEWELL

Griffin assumed that climatic deterioration toward the end of Hopewell had
its greatest effect on the cultivation of corn. This discussion will therefore be
limited to corn agriculture, although there is good evidence that other cultigens
contributed to the Hopewell subsistence economy. In view of the fact that corn
has recently been found in several Hopewell sites, it will be worthwhile to review
the evidence for corn agriculture in the Middle Woodland period:

Wedel (1938:101, 1943:26), reports the discovery of corn at the Renner site
in Platte County, Missouri . "From pit 9, in a layer of charred grass 27 inches
beneath the surface, were taken six short wide kernels of maize." From a more
recent excavation at the Renner site, Roedl and Howard (1957:57) report "a
kernel of what appears to be maize." Four radiocarbon dates were obtained from
the Renner site: A.D. 1 + 250 (M-572), A.D. 100 + 200 (M-571), and A.D. 430
J-200 (M-573), Crane and Griffin, 1959). A date published earlier, A.D. 680 +
250 (M-454) Crane and Griffin, 1958), was rejected as being too late (Griffin,

See Fig. 1 for location of archaeological sites.

Theory of Climatic Change

LEGEND

A EARLY WOODLAND SITE
4 ECOLOGICAL STUDY

HOPEWELL SITE

POLLEN STUDY

LAKE CARLSON
KIRCHNER MARSH

DlSTERHAFT'S FARM BOG

McGRAW DAINES
TWIN MOUNDS % A f~

PEISKER MACOUPIN

Figure 1. Location of archaeological sites and pollen studies mentioned in text.

1958:20).

The discovery of corn in Mound 4 of the Marksville group in Louisiana is
mentioned by Fowke (1928:420-21) and Setzler (1933:14): "Among the re-
mains was one decorated pot 2 inches high containing minute desiccated frag-
ments of corn, squash, and perhaps other forms of food. A leaf, apparently a
corn blade, had been placed over the top" (Fowke, 1928:420-21).

Corn from the Harness site is recorded by Putnam in a list of additions to the
Peabody Museum: ". . .seeds, nuts, corn and grass, all charred. . .from mounds
belonging to the Liberty group on the land of Edwin Harness, esq., in Liberty
township, Ross Co., Ohio" (Putnam, 1886:426). In 1882, Putnam and Metz dis-
covered corn in a fireplace near the center of Mound 1 of the Turner group in
Hamilton County, Ohio: "Mixed with the ashes were charcoal, flint flakes,
charred corn and corn-cobs" (Willoughby, 1922:29). The disposal of the corn
from Harness and Turner is unknown, and consequently has not been analyzed.

Fowler (1952:169) recovered "a few fragments of charred corn" from the
Gear Lake site in Tazwell County, Illinois. A Hopewell association for this

60 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

corn is assumed, but the context is not given, and Mississippian material is also
known from the site. The possible occurrence of corn is also mentioned in a
trait list for the Mitchell Mound of the Ogden-Fettie group in Fulton County,
Illinois: "Corn stalk (?) below find of projectile points. . .1" (Cole and Deuel,
1937:181). No further information regarding the corn stalk is given, and it has
not been analyzed.

Corn was discovered at the Knight (Ansell) site in Calhpun County, Illinois:
"Not only was there more than a cupful of charred, well-preserved corn kernels,
but mixed with them were a few fragments of charred cornstalks, and what
might have been very small segments of husk" (McGregor, 1958:169). Hugh Cut-
ler, who examined the corn, reports that "Most of the grains are from 1 2-rowed
ears with relatively straight rows. . .The ears probably resembled the Guatemalan
Tropical Flints more than they resembled the historic ears of the region" (in Me
Gregor, 1958:170). The Knight site is represented by a single radiocarbon date
of A.D. 250 1 300 (M-164, Crane, 1956). Griffin (1960a:24) has questioned the
Hope well association of this corn. He points out that the pit from which the
corn was recovered contained Late Woodland and Mississippian sherds as well as
Hopewell material.

The Peisker site in Calhoun County, Illinois, has yielded eight fragments of
corn cob unquestionably associated with the Hopewell occupation of the site
(Stuart Struever, personal communication). Hugh Cutler examined the Peisker
corn, and reports an average row number of 12.5 for all cob fragments analyzed
(Cutler and Blake, n.d.). The corn was associated with Pike ceramics, thus indi-
cating an age of A.D. 100 A.D. 400. The Hopewell component of the Peisker
site is represented by radiocarbon dates of A.D. 180 +_ 130 (M-1405, Crane and
Griffin, 1966); A.D. 70+120 (M-1570), and A.D. 250 1 120 (M-1569, Crane
and Griffin, 1968).

William Gardner recovered corn from the Jasper Newman site in Moultrie
County, Illinois: "Almost 2 ft. down in a pit containing nothing but Havana ma-
terial some dozen fragments of maize were recovered" (in Hall, 1966:905 ).
The corn was analyzed by Cutler, who reports an average of 1 rows per cob
(Cutler and Blake, n.d.). Two radiocarbon dates were obtained from the Hope-
well occupation of this site: 50B.C. 140 (M-1789) and 80 B.C. + 140 (M-
1790, Crane and Griffin, 1968). The latter date was obtained from charcoal
taken from the pit which yielded the corn.

A partially carbonized ear of corn was recovered by Frank Rackerby in a pure
Havana excavation unit at the Macoupin site in Jersey County, Illinois (Racker-
by, 1 969 and personal communication), and a single grain of maize pollen also
associated with the Havana-Hope well occupation of this site was identified by
James Schoenwetter (personal communication). The cob fragment was analyzed
by Cutler, who reports that it was 1 2-rowed (Cutler and Blake, n.d.). Samples
for radiocarbon analysis have been submitted to the University of Michigan.

Prufer found "A medium-sized, 1 2-rowed ear. . .and a distorted medium-

Theory of Climatic Change 61

sized kernel" (Cutler, 1965:107) at the late Hopewell McGraw site in Ross
County, Ohio. Yarnell, who was present when the corn was discovered, re-
ports that "No indication was found that could lead to a conclusion that the
corn was not Hopewell" (Yarnell, 1964:105). Cutler examined the cob, and
found it to be "larger than cobs of small 1 2-rowed flint and pop corns of the
central and midwestern states; small when compared to cobs of the 8-rowed Nor-
thern Flints" (Cutler, "1965:108). Cutler also noted that the only preserved
grain was slightly crescent-shaped, and came from an 8- or 10-rowed ear. Seven
radiocarbon dates with a range of 230 B.C. to A.D. 481 were obtained from the
McGraw site. Prufer (1965:104-6) accepts three of these as being most repre-
sentative of this late Hopewell village: A.D. 440 + 80 (UCLA-679 C. Berger,
Fergusson, and Libby, 1965); A.D. 435 +_ 166 (OWU-62) and A.D. 481 + 65
(OWU-61, Ogden and Hay, 1964).

Several maize kernels and cob fragments were recently found at the Twin
Mounds village site in Hamilton County, Ohio (Fred W. Fischer, personal com-
munication). The corn was found scattered throughout the undisturbed portion
of a basin-shaped depression. Analysis of the corn is being undertaken by
Richard I. Ford. Radiocarbon dates are not available, but Fischer believes that
the Hopewell component of this village site should date around A.D. 100 to A.
D. 300, judging from the presence of McGraw Plain and McGraw Cordmarked
ceramics in nearly equal amounts.

At the present time, the earliest reported occurrence of corn in the northern
Mississippi Valley is from the Daines II mound in Athens County, Ohio, where
James L. Murphy (personal communication) found a single carbonized ear of 10-
rowed corn still enclosed in the husk (Hugh Cutler, personal communication;
Cutler and Blake, n.d.). The corn was recovered from the undisturbed base of the
mound, five feet from the outer edge. Murphy states that the mound is unques-
tionably Adena, as evidenced by the presence of a bark-covered burial, prepared
clay floor, and Adena Stemmed points. The Daines II mound is represented by a
radiocarbon date of 280 B.C. 1 140 3 .

The discovery of several carbonized corn kernels at the Leimbach site in Lor-
ain County, Ohio, is another possible instance of maize in an Early Woodland
context. The possibility that the corn might have been instrusive from the Late
Woodland component of the site has not been dismissed, but it was recovered
from an Early Woodland LEIMBACH PHASE feature near the top of a midden
zone, and no internal disturbance by plowing or rodent activity was noted (Or-
rin C. Shane, III, personal communication). The corn will be submitted to Hugh
Cutler for analysis. Radiocarbon dates of 520 B.C. + 310 (OWU-185, Ogden and
Hay, 1967); 510 B.C. + 260 (OWU-250), and A.D. 15+240 (OWU- 251, Ogden
and Hay, 1969) have been obtained from the site, but Shane rejects the latter
two dates on the basis of associated artifacts and possible sample contamina-

This sample was run by the University of Michigan.

62 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

tion. Since the only accepted date of 520 B.C. came from the bottom of the
midden zone, Shane believes that if the corn is genuinely early, it probably
dates to the first century B.C.

With the exception of the Leimbach site and the Daines II mound, all of the
sites in which corn has been found are apparently Hope well rather than more
generalized Middle Woodland expressions in which typical Hopewell artifact
classes are absent. 4 This does not imply that regional variability is lacking, but
there may be a difference between the subsistence economies of Hopewell and
other Middle Woodland cultures. Specifically, the cultivation of corn appears to
be associated with the presence of Hopewell status items indicating participation
in the Hopewell Interaction Sphere.

EVIDENCE FOR CLIMATIC CHANGE

Environmental reconstruction in palynology relies on the premise that the
changes in vegetation observable in pollen diagrams reflect corresponding
changes in climatic conditions. Occasionally they do not, since vegetation res-
ponds to a variety of factors in its environment, not all of which are climatic.
The fact that vegetation is generally more sensitive to changes in climate than to
anything else, however, justifies the use of pollen analysis as the most reliable
method of climatic reconstruction in prehistory.

The pollen zones first defined by Deevey (1939) in the New England area
have provided a model for the interpretation of pollen diagrams in the Midwest.
Of particular interest is the beginning of pollen zone C 3, which is roughly
contemporaneous with the end of the Hopewell period, and which provides
evidence for a clinjatic shift to cooler and moister conditions.

Wright, Winter, and Patten (1963), who analyzed pollen from Lake Carlson
and Kirchner Marsh, faj, Dakota County; Minnesota, use the same zonal desig-
nations as those in New England. They observe that:

The general dominant in the two areas are not necessarily the same but the
climatic changes implied by Jhe vegetational changes correlate in a general
way the C Zones, with oak and various other tree genera dominant in Min -
nesota, imply in both areas a drykjg or warming followed by return to a
moister or cooler climate (Wright, Winter, and Patten, 1963:1378).

The pollen spectrum from Kirchner Marsh shows that oak begins to attain
maximum frequency slightly before sediments in the C 3 zone radiocarbon-
dated at A.D. 290. The decline of oak correlates with an increase of pine at a
slightly later date. In the Lake Carlson diagram, the oak maximum occurs at a
comparable level. The authors reach the following conclusions from these two

4 See Struever (1965) and Prufer (1965) for discussions of the distinction be-
tween Hopewell and generalized Middle Woodland.

Theory of Climatic Change 63

pollen studies:

In Zone C-c the recovery and dominance of the QUERCUS curve implies a re-
turn to moister conditions than those that prevailed during the Zone C-b
'interval The regional rather than local character of the recorded vegetation
is indicated by the same features, even in the minor curves, in both diagrams
(Wright, Winter and Fatten, 1 963 : 1 3 87).

The increase in pine is generally believed to signal the onset of a cooler cli-
mate in this area, if not a delayed immigration phenomenon (Wright, 1 968).

Me Andrews (1966, 1967) analyzed several pollen diagrams from an east-west
transect across four counties in the Itasca region of northwestern Minnesota.
He observed that "About 2,000 years ago PINUS pollen increased to over 35%
in locations on the Itasca moraine; the presence of pine trees and the arrival of
the modern pine-hardwood forest is thereby indicated" (Me Andrews, 1967:234).
He concludes that "the invasion and subsequent dominance of pine on Itasca
moraine is attributed to decreased summer temperatures, a shortening of the
growing season, and an increase in winter snowfall" (McAndrews, 1966:66).

The pine maximum has also been encountered in pollen diagrams in the
Michigan area. Zumberge and Potzger (1956:283) observe that "Since the end of
the Xerothermic, an increase in pine at the Hartford site (O- to 2-foot levels)
suggests climatic deterioration." From pollen studies at Sodon Lake in Oakland
County, Michigan, Cain and Slater (1948) define Period V as a zone which rep-
resents "the recent period which seems to be cooler and. more moist than IV"
(Cain and Slater, 1948:499).

West (1961) has noted the dominance of oak and pine and an increase in
birch in Zone 8 of the Disterhaft's Farm Bog pollen diagram in Green Lake
County, Wisconsin, possibly indicating the presence of an oak savanna type of
vegetation in this part of the Great Lakes following Xerothermic. The presence
of oak, pine, elm, and beech, as well as an increase in birch and the appearance
of hemlock characterize Zone 8 in the nearby Seidel Lake diagram in Kewaunee
County, thus indicating the establishment of the modern Lake Forest.

At Gray's Lake in Lake County, Illinois, Sears (1963) noted the presence of
an oak maximum between two hickory maxima. He concludes that the oak
maximum probably represents a dry period, but that "the two CARYA maxima
suggest somewhat more humid intervals" (Sears, 1963:426). The upper hickory
maximum correlates with the period of Hopewell decline.

The existence of beech-maple maxima along the C-l/C-2 and C-2/C-3 boun-
daries is generally recognized in many of the diagrams from Indiana and Ohio.
Frey (1959) recognized the presence of a beech maximum at the C-2/C-3 boun-
dary in his pollen diagram from Myers Lake in Marshall County, Indiana. He ob-
served that "This is a fairly typical sequence for the central Midwest, of gradual
warming to a maximum in C2 time, with a dry period corresponding to the
CARYA peak, followed by cooling and a moister climate" (Frey, 1 959: 1 34).

64 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

From a radiocarbon-dated pollen diagram from Pretty Lake in LaGrange
County, Indiana, Alice S. Jones (personal communication) found that beech
and maple began to increase in sedim'ents dated at A.D. 120. Both attained
maximum frequency at a later date. The boundary between zones 3b and 3c
in Ogden 's (1966) diagram from Silver Lake in Logan County, Ohio, is brac-
keted by radiocarbon dates of A.D. 640 and A.D. 790. The rise in the percen-
tage of beech and maple pollen, however, begins in the earlier 3b zone, and
peaks in sediments above the 3b/3c boundary..

Ogden offers the following statement on the correlation of these two pol-
len spectra:

. . .the extraordinary correspondence between Pretty Lake and Silver Lake,
Ohio, pollen records indicates that present vegetational differences between
the two areas have existed for at least 10,000 yr and that vegetational changes
have been in phase and apparently synchronous (Ogden, 1969:144).

The studies of Potzger (1946, 1948), Engelhardt (1960), and others sub-
stantiate the widespread nature of this climatic deterioration toward the end of
Hopewell. To demonstrate the effect that this shift could have had on the cul-
tivation of corn, however, it is necessary to determine the effect it could have
had on the growing season.

CORRELATION OF CLIMATE AND GROWING SEASON

Lindsey and Schmelz (1965) undertook a study of Donaldson's Woods in
Lawrence County, Indiana. They found that during the decade of 1954 to 1963,
the forest had changed from a mixed deciduous stand to a forest dominated pri-
marily by beech and maple. Specifically, the authors observe that

. . .during the decade, the combined density of beech and sugar maple in-
creased in nearly all size-classes, while oaks and hickories lost numbers in all
classes below 22 in d.b.h. This clearly suggests a future shift from mixed
mesophytic composition to beech-maple (Lindsey and Schmelz, 1965:176).

The Paoli weather station, located 13 miles from the forest, has maintained
climatic records since 1899. When the authors consulted these records, they
found that the average precipitation for the period April 1 to August 31 was 19.
4" for the years 1899 to 1953, as compared with an average of 21.2" for the
years 1954 to 1963. The average temperature for the same period was 68.3
for the years 1899 to 1953, as compared with an average of 67.1for the years
1954 to 1963. In other words, the decade of 1954 to 1963 saw a 1.8" ."
erage increase in precipitation, and 1.2*"decrease in temperature.

Lindsey and Schmezl summarize their findings as follow: "Thus, the decade
which saw beech-maple increase and oak-hickory decline had climate consis-

Theory of Climatic Change 65

tent with this change, i.e., moister and cooler than normal" (Lindsey and Sch-
melz, 1965:176).

A review of the climatic records from Paoli revealed that the average length of
the growing season for the years 1899 to 1953 was 172 days, as compared with
an average of only 158 days for the years 1954 to 1963. This represents a re-
duction of 14 days in the length of the growing season for the decade in which
beech and maple achieved dominance in the forest.

Yarnell (1964) has recently undertaken studies of present-day climate and
growing season data as related to prehistoric site distribution in the Great Lakes
area. He determined that the average length of the frost-free period for the entire
area is 158 days. Hopewell sites that have produced the remains of cultivated
plants, however, are not found in areas with less than an average of 1 70 frost-
free days. Both Yarnell (1964:132) and Struever (1964:99) agree that the
length of the frost-free period may have been a factor in the selection of site lo-
cations by prehistoric peoples in the northern Mississippi Valley, and Yarnell
believes that this site selection probably resulted from "agricultural experience."

Yarnell was unable to find any early historic reports ot :oi gi ^wing in an
area with an average frostless season of less than 120 days, nor was he able iv
find any prehistoric sites with corn outside of areas that today mark the 1 20-
day average frost-free period. He concludes that:

. . .the Indians who lived in areas where the frost-free period was less than
1 20 days could not depend upon maize agriculture for subsistence, nor could
they do so in areas with longer frost-free periods where other climatic fac-
tors and soil conditions or physiography were not suitable (Yarnell, 1964:
129).

DISCUSSION

Even though the ecological study of Donaldson's Woods and the growing
season data provided by Yarnell are based on present-day climate, both are
valuable because of their ability to reveal trends which may have been opera-
tive in the past. The study of Donaldson's Woods, for example, provides a model
for the interpretation of pollen diagrams featuring a beech-maple maximum.
The dominance of these two species was achieved in only 10 years, reflecting
the vegetational responses to a slight climatic shift toward cooler and moister
conditions. It was shown that a !.*' increase in precipitation and a 1.2 c decrcase
in temperature also shortened the growing season by 14 days in the same 10-
year period.

If we accept Yarnell's conclusion that a 120-day growing s^a:on was neces-
sary for the reliable cultivation of corn, we may assume that a climatic fluctua-
tion resulting in a 14-day reduction in the length of rhe growing season every

66 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

10 years could reduce the frost-free period in any particular area to less than
120 days in a matter of a few years. For a culture located in an area of 190
frost-free days, for example, it is possible' that the cultivation of corn would be-
come unreliable in a period of only 50 years.

I am not suggesting that corn agriculture became undependable in a matter of
50 or even 100 years following the onset of the climatic deterioration. The
changing climate was most likely a gradual trend with many minor fluctuations
which made the cultivation of corn possible for certain years when a late killing
frost lengthened the growing season. It is also possible that certain cultures were
situated in areas with unusually long growing seasons.

It is as difficult to determine exactly when the Hope well to Late Woodland
transition occurred as it is to delimit the period of climatic deterioration. Both-
the pollen and archaeological evidence, however, suggest that by about A.D.
400-450, the decline of Hope well and the climatic shift were nearly complete.

From the occurrence of corn in Hopewell and later sites, it appears that the
climatic shift caused the cultivation of corn to be abandoned.

Corn grown during the Hopewell period is of the small-eared, 10- to 14-rowed
variety sometimes referred to as Tropical Flint (Cutler and Blake, n.d.). Ano-
ther type of corn, called Basketmaker, is usually 1 2-rowed, but is a later variety.
It is found in the Cahokia region, where it was able to survive well into Missis-
sippian times because of an unusually long frost-free growing season in the area
(Yarnell, 1964:114, 135). Some post-Hopewell corn shows characteristics of
both Basketmaker and another variety of corn known as Eastern Complex or
Northern Flint. Eastern Complex corn is large-eared and 8-rowed, and probably
diffused from the Southwest (Galinat and Gunnerson, 1963). It is of common
occurrence in Fort Ancient and Mississippian sites throughout the northern
Mississippi Valley, and is occasionally found in Late Woodland contexts.

There is no clear evidence that either Basketmaker or Eastern Complex corn
evolved directly from the 10- to 14-rowed corn grown in the Hopewell period.
At the present time, the latest occurrence of Hopewell corn is from the Me
Graw site, which dates around A.D. 450. There seems to be a gap in the culti-
vation of maize from this time until about A.D. 700-800 (Cutler, personal
communication), a date which corresponds well with the introduction of
8-rowed maize into the eastern United States. Any occurrence of corn between
these dates will most likely be restricted to areas with unusually long growing
seasons.

It appears that the Tropical Hint corn could not survive the shift to a cooler
and moister climate, and the cultivation of this type of corn had to be aban-
doned toward the end of the Hopewell period. About A.D. 700-800, a better-
adapted and higher-yielding corn diffused from the Southwest, and continued to
be grown well into Mississippian times. This theory is supported by Yarnell's
study, where he found that the average length of the frost-free period for all
sites which have yielded Eastern Complex corn is 159 consecutive days, while

Theory of Climatic Change 67

those with corn of "other types" average 190 days. This suggests that the Eas-
tern Complex corn was better adapted to the environment, for apparently it
could mature in a shorter growing season than that required for the earlier
Hopewell corn.

The theory of a climatic change affecting corn agriculture toward the end of
Hopewell is supported by the data presented above. Several archaeologists, how-
ever, have taken issue with Griffin concerning this theory.

Prufer, for example, believes that

. . .it is debatable whether the climatic deterioration postulated by Griffin
for the northern Mississippi Valley was of sufficient intensity to cause a sub-
stantial change in agricultural productivity (Pruefer, 1964:66).

Yarnell, however, implies that even a mild climatic fluctuation can have a
disastrous effect on the cultivation of corn:

In addition to minimum growing season, crops require adequate night tem-
peratures, daytime temperatures, and average temperatures for certain
lengths of time before they will mature. They also need enough, but not too
much, moisture and sunshine. Thus, all minimum conditions must be met
before a crop can mature (Yarnell, 1964: 149).

I would suggest that the minimum conditions for the growth of corn as men-
tioned by Yarnell were not met after the onset of climatic deterioration.

Baerreis and Bryson (1 965) have proposed a sequence of "climatic episodes"
in which the European-derived Sub-Atlantic period is correlated with Hope-
well. The Sub-Atlantic, which lasted from about 500-600 B.C. to A.D. 300-400,
is characterized by a "more severe" climate. Following the Sub-Atlantic is a
warmer period which lasted until about A.D. 800-900. This period is labeled
the "Scandic episode."

The authors explain why the Sub- Atlantic is a period of more severe climate:
"It is the glacial advances and associated eustatic Roman Emergence that justify
in any way the term "more severe" (Barreis and Bryson, 1965:215). The fol-
lowing "Scandic episode" is so named because of a climatic amelioration in
Greenland, Iceland and Scandinavia.

Apparently, Baerreis and Bryson are forced to rely on paleoclimatological
evidence from other parts of the world because "Little direct evidence is avail-
able from the Wisconsin-Great Lakes region to support the interpretation of the
character of the late climatic episodes" (Baerreis and Bryson, 1965:218). In the
preceeding pages, I have shown that there is ample evidence in the form of pol-
len studies in the northern Mississippi Valley. Furthermore, the concept of the
"Sub- Atlantic" is not universally applicable, a fact which the authors them-
selves acknowledge. In the Southwest, for example, this period is characterized
by a warmer climate.

68 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

In Baerreis and Bry son's scheme, the transition from the Sub- Atlantic to the
Scandic would involve a climatic shift from cooler to warmer conditions. I sug-
gest that this interpretation be abandoned in favor of the hypothesis of cli-
matic deterioration for the following reasons:

1) Baerreis and Bryson's conclusions are based on data derived from other
parts of the world, while the pollen studies cited above constitute direct evi-
dence from the northern Mississippi Valley;

2) Vegetational changes in pollen diagrams reflect climatic changes more
rapidly and more accurately than the movement of glaciers or the rise and fall
of sea levels.

deland (1966) adopts a critical attitude toward Griffin's hypothesis, but
fails to offer climatological evidence supporting his contradictory interpreta-
tions. Apparently, he feels that the acceptance of Baerreis and Bryson's cli -
matic episodes is sufficient justification for ignoring the pollen evidence avail-
able from the Great Lakes region. It is significant to note that Celland fails to
cite even a single pollen study to support his claim that the "Hopewell Epi-
sode," which lasted from 300 B.C. to A.D. 300, was a period of cooler climate,
although pollen evidence figures prominently in his discussions of earlier cli-
matic periods. Following Barreis and Bryson, deland believes that the "Scandic
Episode" was a period of warmer climate.

deland cites his reasons for rejecting Griffin's hypothesis in the following
statement:

Griffin anticipated one of the potential weaknesses in his theory by pre-
dicting that it would be criticized on the grounds that it has not been dem-
onstrated that Hopewellian culture was reliant upon maize agriculture. Not
only is this still true, but most of the corn which has been reported from
Middle Woodland sites is from sites which date during the latter part of this
period, the period of supposed climatic-deterioration (Cleland, 1966:28).

There is little doubt that sites such as Renner, McGraw, and Turner are late
Hopewellian manifestations, but this is not true of other Hopewell sites where
corn has been found. The 280 B.C. and 80 B.C. radiocarbon dates for the Daines
II mound and the Jasper Newman site, as well as the possible occurrence of corn
in the Early Woodland period at the Leimbach site, attest to the early cultivation
of maize in the northern Mississippi Valley.

Clelarfa"s other criticism concerns one of the basic assumptions of Griffin's
hypothesis, and deserves further consideration. The problem is obviously one of
trying to assess the role of corn in the Hopewell subsistence economy, for it is
doubtful if any Hopewellian cultures were reliant exclusively upon maize agricul-
ture.

Struever (1968) believes that agriculture was practiced by some Hopewell
cultures, but suggests that a pattern of "Intensive Harvest Collecting" may have

Theory of Climatic Change 69

obtained in certain locales. Struever's concept of Intensive Harvest Collecting,
which is closely related to Caldwell's (1958) idea of "primary forest efficiency,"
involves:

- an adaptation centering on exploitation of SELECTED, HIGH- YIELD-
ING natural food resources characteristic of certain biomes that have a sharp-
ly restricted geographic distribution within the woodlands of northeastern
United States (Struever, 1968:305).

I do not regard Struever's concept of Intensive Harvest Collecting as incon-
sistent with the evidence for corn agriculture among some Hopewell peoples.
Perhaps local environmental adaptations favored one over the other, but I hardly
think that agriculture and intensive gathering need be mutually exclusive. The
fact that more and more Hopewell sites with corn continue to be found, how-
ever, in addition to the recent evidence for corn in the Adena period; leads me to
the conclusion that corn may have played a more important role in the Hopewell
subsistence economy than has formerly been suspected.

The number, and distribution of Hopewell sites which have yielded corn pro-
vides evidence that the knowledge of maize cultivation was widespread, and im-
plies that it was of some importance to all Hopewell cultures in the northern
Mississippi Valley. Furthermore, corn does not reproduce year after year-it
must be cultivated and tended by man. Whenever corn is found in an archaeo-
logical site, deliberate planting, purposeful storage, and conscious selection from
a previous harvest is indicated (Griffin, personal communication). For these
reasons, it seems to me that the contribution of corn to the Hopewell diet was
considerably greater than is indicated by the few kernels or cob fragments that
have been accidentally preserved by charring at any particular site.

If a climatic shift to cooler and moister conditions forced those cultures for-
merly . dependent to a large degree upon maize agriculture to revert to a- subsis-
tence economy based primarily on hunting and gathering, it is unlikely that the
Interaction Sphere could have been sustained. The withdrawal .of certain stra-
tegically-located cultures having access to the sources of important raw mat-
erials may have caused the breakdown of the Interaction Sphere and paved the
way for the transition to early Lake Woodland.

CONCLUSION

Griffin's theory of climatic change will be supported or contradicted only by
continued excavations of Hopewell sites which yield evidence ^subsistence ac-
tivities, and by additional pollen studies in which greater emphasis is placed on
the delimitation of climatic fluctuations during the past 2000 years. If Struever
and Yarnell are correct in their belief that prehistoric peoples in the northern
Mississippi Valley selected sites on the basis of the growing season, it would be

70 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

valuable to investigate Hopewell sites which are located in areas with long frost-
free periods and possessing other environmental conditions favorable to the
growth of corn and other cultigens.

Considering the present state of our knowledge about the Hopewell to Late
Woodland transition, Griffin's theory of climatic change seems to be a valid
explanation for the decline of Hopewell.

ACKNOWLEDGMENTS-

I wish to thank James B. Griffin, James H. Kellar, and Donald R. White-
head for reading the manuscript and for helpful comments and suggestions.

REFERENCES

Barreis, David A., and Reid A. Bryson

1965 "Climatic Episodes and the Dating of the Mississippian Cultures,"
THE WISCONSIN ARCHEOLOGIST, 46:203-20.

Berger, Rainer, G.J. Fergusson, and W.F. Libby

1 965 "UCLA Radiocarbon Dates IV," RADIOCARBON, 7:336-71 .

Cain, Stanley A., and J.V. Slater

1948 "Palynological Studies at Sodon Lake, Michigan, Part III. The Se-
quence of Pollen Spectra, Profile I, " ECOLOGY, 29: 492-500.

Caldwell, Joseph R.

1958 TREND AND TRADITION IN THE PREHISTORY OF THE
EASTERN UNITED STATES. American Anthropological Associ-
ation, Memoir No. 88.

Cleland, Charles Edward

1 966 THE PREHISTORIC ANIMAL ECOLOGY AND ETHNOZOOLOGY
OF THE UPPER GREAT LAKES REGION. University of Michi-
gan, Museum of Anthropology, Anthropological Papers No. 29. Ann
Arbor.

Cole, Fay-Cooper, and Thome Deuel

1937 REDISCOVERING ILLINOIS. Chicago: University of Chicago Press

Crane, H.R.

1956 "University of Michigan Radiocarbon Dates I," SCIENCE, 124:
664-72.

Theory of Climatic Change 71

Crane, H.R. and James B. Griffin

1958 "University of Michigan Radiocarbon Dates III, "SCIENCE, 128:
1117-23.

1959 "University of Michigan Radiocarbon Dates IV," RADIOCARBON
SUPPLEMENT, 1:173-98.

1966 "University of Michigan Radiocarbon Dates XI," RADIOCARBON

8:256-85
1968 "University of Michigan Radiocarbon Dates XII," RADIOCARBON

10:61-114.

Cutler, Hugh C.

1965 "Cultivated Plants" in THE MCGRAW SITE: A STUDY IN HOPE-
WELLIAN DYNAMICS, Olaf H. Prufer, pp. 107-12. Cleveland Mu-
seum of Natural History, Scientific Publications, New Series, Vol. 4,
Nc I.Cleveland.
Personal Communication. Letter dated May 28, 1969,

Cutler, Hugh C., and Leonard W. Blake

n.d. "An Adena Corn Ear from the Daines II Mound, Athens, Ohio."
(In Preparation).

Deevey, Edward S., Jr.

1939 "Studies on Connecticut Lake Sediments. LA. Postglacial Climatic
Chronology for Southern New England," AMERICAN JOURNAL
OF SCIENCE, 237:691-724.

Engelhardt, Donald W.

1960 "A Comparative Pollen Study of Two Early Wisconsin Bogs in Ind-
iana," PROCEEDINGS, INDIANA ACADEMY OF SCIENCE, 69:
110-8.

Fischer, Fred W.

Personal Communication. Letter dated May 31, 1969.

Fowke, Gerard

1928 "Archeological Investigations-II," BUREAU OF AMERICAN ETH-
NOLOGY, 44th ANNUAL REPORT, pp. 399-540. Washington, D.C.

Fowler, Melvin L.

1952 "The Clear Lake Site: Hopewellian Occupation" in HOPEWELL1AN
COMMUNITIES IN ILLINOIS, ed. Thome Deuel, pp. 133-74. III.
nois State Museum, Scientific Papers, Vol. 5. Springfield.

72 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 2

Frey, David G.

1959 "The Two Creeks Interval in Indiana Pollen Diagrams, "INVESTI-
GATIONS OF INDIANA LAKES AND STREAMS, 5:131-9.

Galinat, Walton C, and James H. Gunnerson

1963 "Spread of Eight-Rowed Maize from the Prehistoric Southwest,"
HARVARD UNIVERSITY, BOTANICAL MUSEUM LEAFLETS,
20:117-60.

Griffin, James B.

1952 "Culture Periods in Eastern United States Archeology" in ARCH-
EOLOGY OF EASTERN UNITED STATES, ed. James B. Griffin,
pp. 352-64. Chicago: University of Chicago Press.

1958 THE CHRONOLOGICAL POSITION OF THE HOPEWELLIAN
CULTURE IN THE EASTERN UNITED STATES. University of
Michigan, Museum of Anthropology, Anthropological Papers No.
12. Ann Arbor.

1960a "Climatic Change: A Contributory Cause of the Growth and Decline
of Northern Hopewellian Culture," THE WISCONSIN ARCHEO-
LOGIST, 41:21-33.

1960b "A Hypothesis for the Prehistory of the Winnebago" in CULTURE
IN HISTORY: ESSAYS IN HONOR OF PAUL RADIN, ed. Stanley
Diamond, pp. 809-65. New York: Columbia University Press.

196 la "Post-Glacial Ecology and Culture Changes in the Great Lakes Area
of North America," University of Michigan, GREAT LAKES RE-
SEARCH DIVISION, INSTITUTE OF SCIENCE AND TECHNOLO-
GY, Publication 7, pp. 147-55. Ann Arbor.

1961b "Some Correlations of Climatic and Cultural Change in Eastern
North American Prehistory," ANNALS, NEW YORK ACADEMY
OF SCIENCES, 95:710-17.

1965 "Lake Quaternary Prehistory in the Northeastern Woodlands" in
THE QUATERNARY OF THE UNITED STATES, ed. H.E. Wright,
Jr., and David G. Frey, pp. 655-67. Princeton: Princeton University
Press.

1967 "Climatic Change in American Prehistory" in THE ENCYCLOPE-
DIA OF ATMOSPHERIC SCIENCES AND ASTROGEOLOGY, ed.

Rhodes W. Fairbridge, pp. 169-71. New York: Reinhold Publishing
Corporation.

Theory of Climatic Change 73

Hall, Robert L. (Ed.)

1966 "Current Research-Northern Mississippi Valley," AMERICAN AN-
TIQUITY, 3 1 : 904-8.

Lindsey, Alton A., and Damian V. Schmelz

1965 "Comparison of Donaldson's Woods in 1964 with its 1954 Forest
Map of 20 Acres," PROCEEDINGS, INDIANA ACADEMY OF
SCIENCE, 74:169-77.

McAndrews, John H.

1966 "Postglacial History of Prairie, Savanna, and Forest in Northwestern
Minnesota," MEMOIRS OF THE TORREY BOTANICAL CLUB,

22:1-72.

1967 "Pollen Analysis and Vegetational History of the Itasca Region, Min-
nesota," in QUATERNARY PALEOECOLOGY, ed. E.J. dishing and
H.E. Wright, Jr., pp. 219-36. New Haven: Yale University Press.

McGregor, John C.

1958 THE POOL AND IRVING VILLAGES. Urbana: University of Il-
linois Press.

Murphy, James L.

Personal Communication. Letter dated July 24, 1969.

Ogden, J. Gordon, III

1966 "Forest History of Ohio. I. Radiocarbon Dates and Pollen Strati-
graphy of Silver Lake, Logan County, Ohio," THE OHIO JOUR-
NAL OF SCIENCE, 66:387-400.

1969 COMMENT in "Ohio Wesleyan University Natural Radiocarbon
Measurements IV," J. Gordon Ogden, III, and Ruth J. Hay, RADIO-
CARBON, 11:137-49.

Ogden, J. Gordon, III, and Ruth J. Hay

1964 "Ohio Wesleyan University Natural Radiocarbon Measurements I,"
RADIOCARBON, 6:340-8.

1967 "Ohio Wesleyan University Natural Radiocarbon Measurements III,"
RADIOCARBON, 9:316-32.

Potzger, J.E.

1946 "Phytosociology of the Primeval Forest in Central-Northern Wis-
consin and Upper Michigan, and a Brief Post-Glacial History of the

74 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

Lake Forest Formation," ECOLOGICAL MONOGRAPHS, 16:211-
50.

1948 "A Pollen Study in the Tension Zone of Lower Michigan, " BUT-
LER UNIVERSITY BOTANICAL STUDIES, 8:161-77.

Prufer,OlafH.

1964 "The Hopewell Complex of Ohio," in HOPEWELLIAN STUDIES,
ed. Joseph R. Caldwell and Robert L. Hall, pp. 35-83. Illinois State
Museum, Scientific Papers, Vol. 12, No. 3. Springfield.

1965 THE MCGRAW SITE: A STUDY IN HOPEWELLIAN DYNAMICS.

Cleveland Museum of Natural History^ Scientific Publications, New
Series, Vol. 4, No. 1. Cleveland.

Putnam, F.W.

1886 "Report of the Curator" in EIGHTEENTH AND NINETEENTH
ANNUAL REPORTS OF THE TRUSTEES pp. 401-32. Peabody
Museum of American Archaeology and Ethnology, Vol. 3, Nos. 5
and 6. Cambridge.

Rackerby, Frank

1969 "The Macoupin Site: A Lower Illinois Valley Middle Woodland Set-
tlement." Paper read at the 34th Annual Meeting of the Society for
American Archaeology. Milwaukee, May 1-3, 1969.

Personal Communication. Letter dated May 16, 1969.

Roedl, Leo J., and James H. Howard

1957 "Archaeological Investigations at the Renner Site," THE MISSOURI
ARCHAEOLOGIST, 19:52-96.

Sears, Paul B.

1963 "A Latent Record in the Prairie Woodland Ecotone," GRANAPALY-
NOLOGICA, 4:424-7.

Setzler, Frank M.

1933 "Pottery of the Hopewell Type from Louisiana," PROCEEDINGS,
US NATIONAL MUSEUM, Vol. 82, Art. 22, pp. 1-21.

Shane, Orrin C., HI

Personal Communication. Letter dated May 14, 1969.

Theory of Climatic Change 75

Struever, Stuart

1964 "The Hopewell Interaction Sphere in Riverine-Western Great Lakes
Culture History" in HOPEWELLIAN STUDIES, ed. Joseph R. Cald-
well and Robert L. Hall, pp. 85-106. Illinois State Museum, Scien-
tific Papers. Vol. 12, No. 3. Springfield.

1965 "Middle Woodland Culture History in the Great Lakes Riverine
Area," AMERICAN ANTIQUITY, 31:211-23.

1968 "Woodland Subsistence-Settlement Systems in the Lower Illinois
Valley" in NEW PERSPECTIVES IN ARCHEOLOGY, ed. Sally R.

Binford and Lewis R. Binford, pp. 285-312. Chicago: Aldine Pub-
lishing Co.

United States Weather Bureau
1939-1963

CLIMATOLOGICAL DATA. Superintendent of Documents. Washing-
ton, D.C.

United States Department of Agriculture

1941 "Climates of the States-Indiana" in CLIMATE AND MAN, pp.
852-61. Yearbook of Agriculture. Washington, D.C.

Wedei Waldo R.

1938 "Hopewellian Remains near Kansas City, Missouri," PROCEED-
INGS, US NATIONAL MUSEUM,Vol. 86, No. 3045, pp. 99-106.

1943 ARCHEOLOGICAL INVESTIGATIONS IN PLATTE AND CLAY
COUNTIES, MISSOURI. US National Museum, Bulletin 183. Wash-
ington, D.C.

West, R.G.

1961 "Late-and Postglacial Vegetational History in Wisconsin, Partic-
ularly Changes Associated with the Valders Readvance," AMER-
ICAN JOURNAL OF SCIENCE, 259:766-83.

Willoughby, Charles C.

1922 THE TURNER GROUP OF EARTHWORKS, HAMILTON

COUNTY, OHIO. Papers of the Peabody Museum of American
Archaeology and Ethnology, Vol. 8, No. 3. Cambridge.

Wray, Donald E., and Richard S. MacNeish

1961 THE HOPEWELLIAN AND WEAVER OCCUPATIONS OF THE
WEAVER SITE-, FULTON COUNTY, ILLINOIS. Illinois State Mu-

76 WISCONSIN ARCHEOLOGIST Vol. 51, No. 2

seum, Scientific Papers, Vol. 7, No. 2. Springfield.

Wright, H.E., Jr.

1968 "The Roles of Pine and Spruce in the Forest History of Minaesota
and Adjacent Areas," ECOLOGY, 49:937-55.

Wright H.E., Jr., Thomas C. Winter, and Harvey L. Patten

1963 "Two Pollen Diagrams from Southeastern Minnesota: Problems in
the Regional Late Glacial and Postglacial Vegetational History,"
BULLETIN OF THE GEOLOGICAL SOCIETY OF AMERICA,
74:1371-96.

Yarnell, Richard Asa

1964 ABORIGINAL RELATIONSHIPS BETWEEN CULTURES AND
PLANT LIFE IN THE UPPER GREAT LAKES REGION. Univer-
sity of Michigan, Museum of Anthropology, Anthropological Papers
No. 23. Ann Arbor.

Zumberge, James H., and John E. Potzger

1 956 "Late Wisconsin Chronology of the Lake Michigan Basin Correlated
with Pollen Studies, " BULLETIN OF THE GEOLOGICAL SO-
CIETY OF AMERICA, 67:271-88.

Radiocarbon Date for the Oconto Site 77

ANOTHER RADIOCARBON DATE FOR THE OCONTO SITE
By Robert Ritzenthaler

Through the efforts of Mr. Jack Steinbring and the sponsorship of the Uni-
versity of Winnipeg a bone sample from Feature 15, Oconto Site was analyzed
by the Gakushvin University, Tokyo, radiocarbon laboratory. It was dated at
4,540 +-400 (B.P.-1950) or 2,590 B.C.

This fits in very reasonably wich the dates for other Old Copper sites and is
more believable than the dates of 7510 and 5600 previously obtained for Oc-
onto. The latter two were done in the early days of radiocarbon analysis when
the runs were made on solid carbon rather than conversion to gas and less apt to
be reliable.

On the basis of the Tokyo date Oconto would appear to remain the oldest
of the Old Copper sites with B.P. dates for the other sites running as follows:
Osceola-3450. Reigh-3660, and Riverside -six dates ranging from 1950 to 3040.

BOOKS RECEIVED

TESTAMENTS OF TIME-

By Leo Deuel. Penguin Books Inc. Paperback: $3.45.

NOTES

COMMITTEE ASSIGNMENTS
WISCONSIN ARCHEOLOGICAL SOCIETY

MEMBERSHIP COMMITTEE: Phillip H. Wiegand, Chairman.
Herman Zander, J. K, Whaley.

PUBLICITY: Tom Jackland.

FRAUDULENT ARTIFACTS: J. K. Whaley, Chairman. Martin
Greenwald, Robert Hruska, Neil Ostberg, Dr. Robert Ritz-
enthaler, Paul Scholz, Phillip H. Wiegand.

SURVEYING AND CODIFICATION: Paul Koeppler, Chairman.
Darryl Beland, Elmer Daahnann, Wayne Haziett, Ernest
Schug.

PRESERVATION OF SITES: Robert Hruska, Chairman. Dr.
Richard Peske, Dr. Robert Ritzenthaler. (Chairman has op-
tion to make further appointments.)

EDITORIAL: Dr. Robert Ritzenthaler, Chairman. Dr. David A.
Baerreis, Dr. Joan Freeman, Dr. Lee Parsons, Dr. Melvin

PROGRAM: Dr. Diehard Peske, Chairman. Paul Turney, Mrs.
P. H. Wiegand, Dr. Robert Ritzenthaler.

LAPHAM RESEARCH MEDAL. Dr. Robert Ritzenthaler,
Chairman, Dr. David A. Baerreis, Phillip H. Wiegand.

CHARLES E. BROWN CHAPTER

(Meets second Tuesday, 7:45 P. M., State Historical Society,
October lluru May)

Secretary: Peter Storck

FOX VALLEY

(Meets second Monday, 7:30 P. M., Oshkosh Public Museum,
September thru May)

vice President: Richard Mason
Secietaty: Robert R. Jones
L. Hoist

THE WISCONSIN
ARCHEOLO6IST

THE MIDWAY VILLAGE SITE: An Orr Phase Oneota
Site in the Upper Mississippi River Valley
by Guy E. Gibbon

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee, Wisconsin
Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. M., Milwaukee Public
Museum, September thru May

OFFICERS

PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Ootberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Greenwald.

TREASURER

Wayne J. Hazlett
3768 S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct, Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr. Robert E. Ritzen thaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL

MEMBERSHIP FEES

The Wisconsin Archeologist is distributed to members

as part of their dues.

Annual Members, $5.50

All communications in regard to the Wisconsin Archeological Society
and contributions to the Wiscorsin Archeologist should be addressed to
Mrs. Edward Flaherty, Secretary, N62-W15127 Tepee Ct., Menomonee
Falls, Wis. 53051. Entered as Second Class Matter at the Post Office at
Lake Mills, Wis., under the Act of August 21, 1912. Office of Publica-
tion, 316 N. Main St., Lake Mills Wis. 53551.

THE WISCONSIN ARCHEOLOGIST

New Series

MILWAUKEE, WISCONSIN r- SEPTEMBER, 1970
Published Quarterly by The Wisconsin Archeological Society

THE MIDWAY VILLAGE SITE: An Orr Phase Oneota Site
in the Upper Mississippi River Valley

Guy E. Gibbon
University of Illinois

INTRODUCTION

Middle Mississippi cultures are regional representatives of an agricultural,
sedentary way of life generally known as the Mississippian tradition in the eas-
tern United States. The roots of this lifeway begin to appear in the Middle Mis-
sissippi Valley between A.D. 700-800. Oneota designates a number of distinc-
tive prehistoric and early historic complexes in the Upper Mississippi drainage
system exhibiting some "Mississippian" traits. The Oneota settlement-subsis-
tence patterns reflect a less complex level of sociocultural integration and a
greater reliance upon hunting-gathering activities, than is typical of the Middle
Mississippi pattern. Hypotheses as j the relationship of Oneota and Middle
Mississippi cultures have proffered explanations attempting to account for the
less complex nature of Oneota. This research project was designed to test one
facet of one of these hypotheses. 1

James B. Griffin has suggested that the Oneota cultural complex developed
from a northern push of Middle Mississippi peoples during the Neo-Atlantic
warm period (1960, 1961). A marked climatic change which began in approxi-
mately A.D. 1300 produced a relatively cool period that was extreme enough,
Griffin argued, to affect the horticultural practices on which the northernmost
Middle Mississippi cultures were partially based (1961:710-712). The Middle
Mississippi cultures:

"shifted away from a marked dependence on agriculture to a heavier em-
phasis on hunting, and. . .there was a marked drop in cultural level. This
period of gradual decline can now be clearly seen as occurring in the A.D.
1300 to 1650 time period" (ibid.).

As the cultural patterns of these northern complexes adjusted to a new sub-
sistence base, the characteristics, which are now recognized as typifying Oneota,
emerged.

Some Oneota sites in Wisconsin are very large. The Shrake-Gillies site in
Trempealeau County is reported to cover an area of about forty acres, and the
Midway site in LaCrosse County to cover an area of approximately thirty-five

80 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

acres (McKern, 1945:123-124). Aztalan, the largest Middle Mississippi site in
Wisconsin, occupies an area enclosed within a stockade of only twenty-one
acres. The relative sizes of these sites leads to the formulation of an interesting
problem. If the agricultural potential of the Middle Mississippi culture was cur-
tailed to such an extent that a deterioration in culture resulted, a decrease in
the size of the village community would be expected and not an increase of the
proportion found between Aztalan and the larger Oneota sites. The apparent
large size of these sites could be the result of course of other factors such as in-
termittant occupation by small groups. This project was initiated to test the
degree of intrasite cultural homogeneity at the Midway site of the Orr Phase.
The description of the recovered material and the relationship of the site to
other Orr Phase components were also considered integral objectives of the pro-
ject.

An attempt has been made in the organization of the report to separate the
factual data from the interpretations presented by the writer. The primary rea-
son for this approach is to facilitate comparative research by not encumbering
the data with interpretations that are subject to revision. Interpretations de-
rived from the data are presented in a separate section. This latter section is fol-
lowed by a general conclusion in which the problems pertinent to this report are
reviewed.

SITE HISTORY

The Midway village site, 47LC1 9, is situated in the S.W. & of the S.W. &, Sec-
18, T. Onalaska, R. 7W., La Crosse County, Wisconsin, where Halfway Creek
and County Trunk XX intersect. The site was dug to some extent by Profes-
sor A.H. Sanford of La Crosse State College (now university) in the early 1920's
and was subjected to a more intensive investigation in 1929 by a party under
the direction of W. C. McKern of the Milwaukee Public Museum. The results of
the latter excavation were subsequently mentioned in two reports written by
McKern (1931, 1945). Except for this partially described excavation, no other
recorded investigations have been conducted at the site prior to this report. The
site has been extensively cultivated at least since the turn of the century and
has long been a favorite of the private collectors in the area.

SITE DESCRIPTION

The topographic setting of the site is a fairly level area along the outer edge
of the Midway-New Amsterdam terrace system in the Gorge of the Upper Mis-
sissippi River Valley. The site is sixty feet above the adjoining flood plains of the

This report is a condensed version of a Master's thesis prepared under the
guidance and at the suggestion of Dr. David A. Baerreis (Gibbon, 1966).

The Midway Village Site 81

Black and Mississippi Rivers and one mile from the main river channel. The
gorge is part of the geographical area described as the Western Uplands. This up-
land region is within the unglaciated or "Driftless" area of southwestern Wis-
consin and adjacent parts of Dlinois, Iowa and Minnesota. The region is char-
acterized by a broad humid-continental type of climate having four seasons.
The natural vegetation of the gorge is part of the Oak sub-climax of the North
Central United States (Curtis, 1959).

The site surface is treeless and has been under cultivation for at least half a
century (Pis. 1-2). A land survey in 1836 described the area as 'sandy, second
rate, containing Bur, Black Oak and a few Hickory Trees.' The soils covering the
site are Plainfield fine sand (2-6 percent slope, eroded), a Regosol, and Sparta
loamy fine sand (0-2 percent slope) a Brunizem. These soils are very sandy, ex-
cessively drained, and have an average pH of 6.0. Crop yields are low even under
good management.

W. C. McKern estimated the actual areal extent of the site was considerably
smaller. A series of nineteen survey pits was dug to establish the limits of the
site. This survey indicated that the site covered at least eight and one-half acres
and possibly two and one-half additional acres in a general northeast direction.
This latter area is stippled on Map 1 .

The southern and eastern limits of the site are defined by a steep escarpment.
Halfway Creek runs along the bottom of the escarpment on the east and the bot-
tomlands begin at the foot of the esca; pment on the south. The field on which
the site is situated continues northward for about one hundred and fifty yards.
A maze of wind formed sand dunes define the western limit of the site and the
northern limit of the field. These dunes are now stabilized and form part of a
prairie setting that stretches northward. The site surface is fairly flat, except
for a gradual east-west rise reaching a mound-like peak some ten feet above the
rest of the site in the northeast corner.

EXCAVATION PROCEDURE

At the time of excavation, the northern three-fifths of the field were planted
in alfalfa. Two spacially distinct areas were selected for comparison in the Vil-
lage' area and designated Area I and Area II. A burial ground of undetermined
size in the extreme southern portion of the site was avoided (Map 1 ). The site
did not reveal any noticeable differences in cultural material as far as could be
determined from an analysis of an extensive surface collection.

Three hundred square feet of surface area were excavated in Area I and three
hundred and fifty square feet in Area II over a period of six weekends from
September through October in 1 964. In the excavation of both areas an attempt
was made to locate productive features. The plow zone, which was about eight
inches thick, was shoveled aside. Arbitrary four and six inch levels were main-
tained for vertical control within five by five or five by ten foot square units. All

82 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

of the soil removed from these levels was screened through a 1 A inch mesh. Mar-
ked soil horizons were absent in the undisturbed deposit. The depth of each level
from the surface is recorded in Tables 39 and 40.

Table 1. Depth and Surface Extent of Excavated Units

AREA I

AREA II

unit #

surface
extent

depth

unit //

surface
extent

depth

5 T by 10'

3'8"-

5' by 5'

1'8"

5' by 10'

2'8"

5' by 5'

empty

5' by 10'

2'Q"

5' by 5 f

l f 9"

5' by 5'

3'0"

5' by 5'

2'2 M

5' by 5'

2'3 M

5' by 5'

3'9"

5' by 5'

2'4"

15' by 5'

1'6"

5' by 5'

2'6"

5' by 5 f

1' 9"

5' by 5'

3'0"

5' by 5 f

1' 7"

5' by 5'

3'6"

5 T by 5'

3'2"

5' 'by 5 T

2'4"

5' by 5 1

4' 6"

5 f by 5'

2'10"

FEATURES

'Refuse' pits were the most common features encountered at the Midway site
The word refuse pit is used here to indicate black concentrations of refuse ma-
terial. These pits were not necessarily self-contained units, for similar materials
were found in the lighter colored soil around each feature at all depths. The lack
of integrity of these features can be illustrated by the scattered fragments of one
jar. Rim sections of this jar, all of which fit together, were found in square 9,
Feature 10, at a depth of 17-23" and in the continguous square 8 at levels four
(24-30") and five (30-36"). Other sections of the jar were recovered from square
2, level 1 (9-15") and square 20, level 2 (18-24"). Perhaps these refuse pits rep-
resent the differential oxidation of scattered organic matter and the consequent
reduction of the surrounding soil in areas of high food debris concentration.

No post molds or house trenches were found. A small mussel shell heap (F9),
a historic pig burial (F12), and fragments of a disturbed human burial (F13)
were the only other features discovered.

All of the refuse pits were roughly basin-like or flat lenticular in vertical out-
line. The smallest pit (F15) was approximately 13 inches long, 12 inches wide,
and 13 inches deep. The largest (F10) measured 48 inches in length, 44 inches in
width, and 13 inches in depth. Individual measurements and a short description
of each feature are given in Appendix A. A listing of the presence or absence of

The Midway Village Site 83

mussel shell fragments and the quantities of utilized flakes, unmodified flakes,
shell tempered body sherds, and grit tempered sherds found in each test square
is given by level in Tables 39 and 40. The remaining relevant data are contained
in the descriptive sections.

DATING THE SITE

A Carbon 14 date from a charcoal sample taken from Test square 1, feature
5 (36-40 inches below the surface) gave an age determination of A.D. 1420 +
70 (WIS 61). A second Carbon 14 date run on a charcoal sample from Test
square 20, level 5 (36-42 inches below the surface) gave an age determination of
A.D. 1630 + 60 (WIS-79). Both of these samples were taken from the deepest
levels containing adequate amounts of charcoal in each area excavated.

CERAMICS

Methodology

Various methods have been used in the past to describe the pottery found
at the Midway site. W. C. McKern utilized a method of descriptive classification
that was primarily concerned with elucidating certain fundamental characteris-
tics which differentiated one phase from another (1945:143-152). Standard
measurements, such as rim heights or percent of occurrence of design units,
were only infrequently mentioned.

Plate 1. View East Across the Midway Site

84 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Plate 2. View of the Mississippi Flood Plain from the Midway Site (looking
southwest)

During the years 1959-1961, three reports appeared which introduced the
typological concept to the site material. Mildred M. Wedel (1959) and Dale R.
Henning (1961) discussed the site ceramics in relation to Allamakee Trailed, a
pottery type defined by using material from the Upper Iowa River 'heartland'
of the Orr Phase. Both of these reports suggest that the Midway ceramics are
similar to Allamakee Trailed, but sufficiently different to warrant possible ex-
clusion from the phase (Wedel, 1959:121; Henning, 1961:35-36). Robert L.
Hall (1 960), while discussing Mississippian cultures in Wisconsin, indicated that
the Midway ceramics might be arranged in a number of types, the main type
being Midway Incised. When this report was prepared for publication (1962),
Hall included this type, however, within the limits of variation of Allamakee
Trailed (Hall, 1962:132).

The methodology used in this report for the purposes of ceramic analysis
consist^ of a blend of the typological method as used by Wedel (1959), Hall
(1962), and Henning (1961), and of an attribute analysis. This latter analysis,
conducted separately from the typological analysis, serves a twofold purpose: 1)
It makes possible a more detailed description of the potentially homogeneous
sherds from the two areas excavated and 2) It provides a measurement of the
integrity of the pottery types already in use. Grit tempered Woodland sherds
and all bowls are excluded from this analysis. The data have been presented as
they pertain to each 'area' excavated and then to the site as a single entity. Both
the excavated IN SITU material and that recovered from the surface have been

The Midway Village Site

Map 1. The Midway Site

included under the 'site total' columns.

The terminology used in this report to describe vessel form consists of the
lip, rim, neck, shoulder area, and body. The following definitions are slightly
modified from Henning (1961:11-12).

LIP: Juncture of the inner and outer surfaces
RIM: Section between the lip and neck
NECK: Point of maximum vessel constriction

86 WISCONSIN ARCHEOLOGIST Vol. 51, No. 3

SHOULDER AREA: Space between the neck and point of maximum body

diameter
BODY: Area between the neck and bottom of the vessel

Most of the attributes used in this report are non-continuous stylistic var-
iables. Several continuous variables (e.g. rim height) have been included, because
they appear to be important in distinguishing the 'Iowa Orr phase' sites from
the Wisconsin Orr phase' sites (Wedel, 1959:114-115, 131; Henning, 1961:35-
36). Other continuous variables such as thickness, hardness, or color are con-
sidered separately, as are surface finish and temper, which are homogeneous
for the sample. Information concerning general vessel shape, vessel height, ori-
fice diameter, and number of attached appendages is included as a supplemen-
tary part of the following general discussion. Because of the fragmented nature
of the sample, these latter attributes are rarely measureable.

Sample

The ceramic analysis is based upon the following sample:

Area I
Area II
Site Total

No complete or nearly complete vessels were recovered. Therefore, inferences
as to the characteristics of complete vessels must be based on large rim sections
and bodysherds. Each rim sherd probably represents a single vessel. The pottery
recovered from Areas I and II is predominantly a uniform shell tempered ware
which accounts for 98.3% of all sherds. The remaining 1 .7% is a grit tempered
Woodland pottery which will be discussed separately. Because of the basic simi-
larity of the shell tempered ware, the paste, surface finish, and general vessel
form are discussed together.

Paste

Temper: Temper consists of moderate to abundant amounts.of crushed shell
with few pieces exceeding 3 mm. in length. Most of the shell particles are aligned
parallel to the vessel surface. On many sherds the shell has leached away leaving
irregular pits or holes on the surface. Occasionally a small amount of grit is in-
cluded in the paste.

Texture: The consistency of the pottery ranges from compact to a more fre-
quent flakiness. The fracture of a sherd reveals a laminated cross section.

Hardness: Most of the sherds measure 23.5 in mineral hardness on Moh's
scale. A few sherds range as high as 4. When decorated these latter sherds display

bodysherds ,
1639

rim sherds
63

handles
12

1431

3283

143

The Midway Village Site .87

a greater neatness in design execution.

Color: Interior surfaces are generally grayer than the exterior colors, which
range from dull orange to gray and dull brown to black. Many bodysherds are
smoke blackened and some occasionally have a flaky black residue on the inner
surface.

Surface Finish: The surfaces of all sherds are smooth and dull. 'No polish or
glaze occurs on the ceramics. A red film or powder is found On 7 of the plain
bodysherds (5 from Arej I, 2 from Area II) and on 5 decorated shoulder sherds
(all from Area I). Evidence of scraping by a bundle of grass or similar object
frequently appears on both surfaces of the rim and on the interior surface of
bodysherds. Smoothing was usually carelessly executed.

Vessel Form: As far as was ascertainable from the fragments of vessels avail-
able, the ^predominant vessel form is globular or ellipsoidal witfefagbded rather
than flat or conoidal bases. The shoulder area ithins rapidly from the juncture
with the neck and ascribes an arc with the remainder of the vessel wall. No
sherds were found which would indicate an angular rather than curved shoulder.
The following table indicates the frequency in percent of the thickness of plain
bodysherds as estimated from a sample of each; area. -.-.. , ; f ? . ;* ,, ;,n

r">

Table ' i: Bodysheid thickness , . .-.,*.. . r-..V,

Max. Xhi'ckness r 'i-n "mm. ' : - ' Area' l' : ' ' Area II Site Frequency

( jr .3 to 4 ; , , ... ,; - -,ffi Q.5 .. i 9.0- > ! ,. ; : --^.0^"

A through- 5'" 577^ '' . 62. 59.0, ..-..,.

i 6: . L i. ,''"'. : '''7 '."L - ! J'-- si.^36.0 i'*->'24lO
1 8'" ""'- '*' 5.0 5.0

10 " 11 1.0 0.0

Lip Diameter and Vessel Capacity

The nature of the sample, pr-ecludes any exact rmiiisurernent of '^Sel sizes,
^ip.di^m.eters pro-bably range from SrrlO cm. on small jars amd;up td>at> least '45
cm. on larger vessels. . The . great majority ,ofi jars are 'scattered between r 20^3-5
cm., with a marked clustering around 30 cm. Estimates of lip diameters are given
in Tajble J.9 20. Vessel capacity ranges from small; jars capable of <ibrtlaining
somewhat m<" than a pint up to vessels with a 5-8 gallon capacity. No

88 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

estimates of vessel heights were attempted.

Attributes

Class I: Lip Profile

All lips can be separated into a rounded or a flattened grouping (rounded:
PI. 4, A and B; flattened: PI. 4, C2 and C3). The lip was usually thinned and
rounded before the application of decoration. Lip thickness varies from 41 1
'mm., with 67 mm. the most frequent measurement. The thickest lips are as-
sociated with flattened lip surfaces. Attribute associations for all classes are lis-
ted in Tables 19-20.

Table 3. Lip Profile

profile
rounded
flattened
totals

no
46
5

Area I

Area II
no. %

Site
no.

Total

90.2
9.8

94.
5.

125
9

93.
6.

100.0

100.

134

100.

Class II: Lip Decoration Technique

There are three techniques of lip decoration in the sample: tool-impressing
(PI. 3, nos. 2, 6, 7), fingertip punctating (PI, 3, nos. 1, 4), finger or broad tool-
impressing (PI. 3, nos. 3, 5). The latter group consists almost entirely of 'scal-
lops' and 'long ellipses' (discussed in Class III).

The term 'tool-impressed' refers to the impression produced by placing an
object other than part of the body parallel to or at an angle to the lip or vessel
surface and pressing downward. The term 'punctate' refers to the marks pro-
duced by pushing a tool or fingertip vertically into the vessel surface. Fingertip
punctate is characterized by the presence of a fingernail imprint in the punc-
tation.

Class III: Lip Design Elements

This class is composed of 7 basic designs. Plate 3 illustrates the design ele-
ments with their corresponding code numbers. Because this grouping depends in
part on method of execution, a brief explanation of each attribute is given
below.

1. simple fingertip punctates consecutive shallow fingertip punctations, us-
ually round, which seldom alter the entire lip surface. Punctations range from
4-6 mm. in diameter and are generally 5 mm. apart. The depth seldom exceeds
1 mm. and never 2 mm.

2. oblique tool impressions consecutive tool-impressions, always elliptical,

The Midway Village Site 89

Table 4. Lip Decoration Technique

Area I Area II Site Total

technique no. %_ no. _%_ no. _%_

1. tool-impressed 15 30.0 11 16.9 28 21.5

2. fingertip punctate 2 4.0 2 3.1 4 3.1

3. finger or broad 19 38.0 22 33.8 47 36.2
tool- impressed

4. plain 14 28.0 30 46.2 51 39.2
totals 50 100.0 65 100.0 130 100.0

that cross the lip at an approximate 45' angle. The width of the impressions and
the distance between them is 25 mm. The impression length ranges from 39
mm. and the depth seldom exceeds 3 mm.

3. scallops flattened impression^ 1520 mm. long, which are either continu-
ous or consistently about 5 mm. apart. They appear to have been made by pres-
sing a finger onto the lip surface.

4. deep fingertip punctates-essentially the ne as 1 above except th< punc-
tations are larger and deeper -ranging from 3-7 mm. in depth. The ptnctates
alter the entire lip.

5. long ellipses similar to 3 above except the impressions are about 35 mm.
in length.

6. transverse tool impressions essentially the same as 2 above except the
impressions cross the lip at an approximate 90 r angle.

7. inner lip tool impressions consecutive elliptical tool-impressions con-
tained within the lip surface. Impressions range from 71 1 mm. in length and
23 mm. in width.

Class IV: Rim Profile

Rim sections large enough to display a complete profile were grouped under
9 headings. Examples of each group are illustrated in Plates 4 and 5. The criteria
used in establishing these categories were height, thickness, and basic configura-
tion. Groups A C are of the same approximate height; A, however, consists of
rims with angular necks, while the neck sections of B and C are curved. Groups
B and C are separated by the greater thickness of both the rim and the lip of
specimens in the latter group. Rim heights in D and E are shorter, than in the
above groups. The relative thinness of the rims in E, in addition to the less pro-
nounced curvature of their necks, separate this group from the rims in D. Groups
F-l are composed of forms differing markedly from the other groups as well as

90 WISCONSIN ARCHEOLOGIST : Vol> 51, No, 3

Table 5. Lip Design Elements

...... ,.v* i* ; M 5 ^ ; '

Area I

Area II

Site Total

design

no. %

no . %

no. _% .-

2 5.6

3 8.6

6 7.6

12 33.3

5 14.3

18 22.8

16 44.4

12 34.3

i ( 31 39.2,

1 2.8

1 2.9

2 ,2.-5

3 8.3

10. 28.6

15 19.0

1 2.8

.: 1; : 2.9

2 2/5

1 2.8

3 8.6

5 6.3

* r ,;. .r' \ ' - '

; If-', ,$';! ' , ; ' -/ ; ' 'i'

1 totals

;i 36 i6o" f O; r .

35\ 100^0

79 \99.0 , .

from each other.

;r;rn v - ' "'' -i-

oL .Ov '. . Ir.'- . v*r

- Vessel rims generally

taper ifrorn the

neck to the lip.

Rims uniform w 'thick-

ness are in the minority .

The shapes of

the rim body vary from straight to out-

curving. Rim thickness varies from 51

2 mm. in both 1 Areas I and II. The most

frequent fnea.surement is

about 8 mrii.' :

.j ..: ." -'.. ' '.-r '-.:'

-.it:- in.?^*i'l .:'*.i.' .1

Table 6. Rim Profiles

.]: ilitf'r.- : d, tt i!

<nrrU;;joJ,::M '.&

>=-.v^ea!H;io-

..^- 'Area 11^:

/; 'Site- Tbtal^ ' ?

profile

nO. ; , _%_-:

, v no, : ^,l, :v

fi no ; rl-'. ' .'

.:-to--: J3.3'-

6rl'2tf;r-

"' 1^6' '; ; 25.0

; ro; - w&t'^'i'' " :;;

-""7 '%3 3'

4 'l3 8 '''

L 12 ! 18' 8 s '

;/-'., '!i-4,;-F ,;. ,in;r i ;

\ 1 i''>ii ijTi :: -

r;. >;/.. > ;.Pv, )fi1

., ; .. ; i

2 6.7

6 20.7

8 ' 12.5

4 13.3

4 13.8

''- ^^-14:1 '

. . i. t , , ......... ,, .. .

....... ../ '-I-::'} ,

. .:(. ':.; '> ': i/

-n : '. -. : i -;-: . J t' ' '"

'ft ^.

.3 ; 1Q. .,

.2. .6. 9

., 7 1 . -10,9 : .

t'.!"i ';' V* "**. - : *" > i'. i* '.

.' . . - . -''"..*' *-

. : ' /:' t . ( - i ' '-'

L , -

i .

;.' .'. : r : . f -!' y : , ' ,- "'

2'' "' ' o 7

3 10. 3

6 ' 9.4

i,, f r ,., f",< .-, 7 . .,- ..f ;:,

H, '; * t ;p|j{ : ' ....

nr l/-, ; ,3 v3 ...,

- Q. "0.0 ' '

* ! !' * lv&i- '

., ' ...,,,,-. .. ': -f.

: , i,

|.it \ .(,.;; ;' 'i

r r ; -, : ; ^ . ; . .^ :'.'.'

H 1>!

/'i'. ; ^i;'

3 10 . 3

4 .6. 3

'.MI < :! "iii!- / ,'..- '':' ~ ' '

.!'': . ' '

,-' :'. M.:.! :

;0 , ; <- I; i

' -iO =0v0'

1 3.4 : '

: ^ -1' 1.6:

;^<f; -0 ^J : : - ?fl-: l '- ; i !

", i si'

'totals" _ ,. ,., f

30 . .99^ ; ,,

.- 29 '...9Bj-Q:: {

,64 '100.0

The Midway Village Site 91 '

Class V: Rim Height

The boundaries used in this class are those suggested by Wedel (1959:85).
Measurements were taken from the lip to the neck interior. Rim height ranges
from 1352 mm. The mean rim height is 36 mm. in Area I and 35 mm. in Area
II. In general high rims appear to be associated with relatively large vessels. The
actual height of each rim is listed in Tables 19 and 20.

* ti*
Table 7. Rim Heights

height

1. 10 to 19 mm.

2. 20 to 29 mm.

3. 30 to 39 mm:-

4. 40 to 49 mm.

5. 50 to 59 mm.

totals 29 D(fc S 32- V X)0;0- ^ 69 99.9

Class VI: Inner Rim Upper Edge Design Technique

The decorative techniques used on the upper edge of the inner rim are the
same as those used in Class II, i.e. tool-impre?sin|'(Pl. 6, nos. 1-2), fingertip
punctating (Pi. 6, 3B), and finger or broad tool-impressing (PI . 6, 3 A).

Area I

Area II

Site

Tot

no.

'.JL-' ' fo.*- ' ftp

no.

20.

28.

26.

31.

*"^*

29.

48.

34.

42.

Table 8. Inner Rirn, -Upper fedge !M%n Techniques

Area I

Area II

Site Total

technique

no.

1
2

4
8

8.
17.

7
4

20
3

27
11

23.
9.

1
4

12.

plain
totals

69.

55.

100.

100

.0 1

inn

92 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

The Midway Village Site 93

I 2

1 2 3 ^^fc^ 4^^B ^^^

5 cm.

Plate 4. Rim Profiles: A-C (interior surface at left)

94 WISCONSIN ARCHEOLOGIST Vol. 5 i , No. 3

<<(( '//

I K <

Plate 5. Rim Profiles: D-I (interior surface at left)

Class VII : Inner Rim Upper Edge Design Elements

Three basic designs are found on the upper edge of the inner rim. They are
(PL 6):

1. long, oblique tool-impressions

2. long, vertical finger or broad tool-impressions

3. round, vertical fingertip impressions

The Midway Village Site : , .. -r" 95

Table 9. Inner Rim, Upper Edge Design Elements

., .. ^ : - :.{...*. -,,.;::

Area I

Area II

Site Total - , ,

designs

no .

no. %

ho. '%

I/-

..'. 3

,21,4

17 54.8

.24 46.2

14.3

5 16.1

7 13 .lf 5

-' 9

64.3

- 9 ; 29.0

21 40/4

totals

100.0

31 99;9-

. 52 10Q.1

Class VIII: Angle of Rim to Body

Angular divergence between the rim and a vertical line drawn through the per-
imeter of the vessel orifice ranges from 10-^5 degrees, with 30 degrees the most
frequent angle encountered. No vertical rims are present in the sample. All of the
rims, with one exception (PL 5, G), slant away from the vesselihterior.' "\

.; . * - . . : .... ':. ,'''- -i. '1*' ! .: -r

' . ._._ -;> -"i-"-! , il :in ' .-- :; ., . ' i* " '" ; *>>' ' ' -

Table 10. Angle of Rim to Body

; ;.

" ; - *>'*: :u

1 A

rea : I v '

Area II

'-Site TBtar
,'.*' ..

. angle,,. -..;.

; a^

-.Wlf:is:r

>02h

. !>' %

ins 1 ant.

j;i l.

2.3

rlfj-

2' r

outslant

97.7

100.

98.

totals

100.0

100.

Class IX: Shoulder-Neck Angle

Rim sections retaining an attached portion of the shoulder were segregated
into one of two groupings depending on whether the juncture of the rim and
shoulder was curved or angular. Rim profiles in PI. 4, A, for example, consist
entirely of rims exhibiting an angular juncture, whole grouping B is composed of
rims with a curved juncture.

Table 11. Shoulder-Neck Angles

Area I

Area II

Site Total

angle

no. %

no.

no . ?c

angular

15 34.

32.

28 32.

curved

29 65.

67.

53 67.

totals-

44 100.

100.

86 100.

96 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

Class X: Shoulder Decoration Technique

Eleven decorative techniques and combinations of techniques are present on
shoulders with rim sherds attached. These attributes are segregated in part by the
establishment of arbitrary boundaries defining widths of lines and in part by the
method of execution. The arbitrary boundaries used are those suggested by
Wedel (1959: 131). A brief explanation of each attribute is given below.

1. fine line, trailed (P 1.7) trailed lines, about 1 - 2.5 mm. wide and quite
shallow. Trailing in tjtis report refers to lines formed by pulling a tool or a finger
across the vessel surface while the clay is still damp (Wedel, 1959:88-89; Henning
1961:12). All of the trailed lines in attribute 1 were produced by the use of a
tool. Incised lines, herein interpreted as formed by cutting across the dry surface
of a fired vessel, are not present in the sample.

2. medium line, trailed-trailed lines produced by a tool, about 2.5-4.5 mm.
wide and deeper than 1 above.

3. wide line, trailed (PI. 8) trailed lines produced by a tool, about 4.58
mm. wide and ranging from 1 4 mm. in depth.

4. finger trailed (PI. 9) trailed lines produced by the use of a finger, about
1014 mm. wide. The countour of the lines appears on the inner surface of the
shoulder.

5. fine line, trailed and tool-impressed (PI. 9)-a combination of 1 above and
10 below.

6. medium line, trailed and tool-impressed (PI . 9) a combination of 2 above
and 10 below.

7. wide line, trailed and tool-impressed-a combination of 3 above and 1 be-

Plate 6. Rim Design Elements (2/3 natural size)

The Midway Village Site

Plate 7. Shoulder Decoration Technique: 1-fine line, trailed (3/4 natural size)

low.

8. finger trailed and tool-impressed (PI. 10) a combination of 4 above and
10 below.

9. fine line, trailed and punctate (PI . 1 0)-a combination of 1 above and 1 1
below.

Table 12. Shoulder Decoration T Unique Attributes

Area I

Area II

Site

Total

echnique

no.

3.50

3.60

133

4 . Ov.

141

8.70

5.20

225

6.20

3.40

2.00

2.70

0.80

0.40

0.60

0.20

0.30

0.70

0.20

0.50

0.10

0.07

0.10

0.00

0.07

0.03

0.00

0.07

0.06

0.10

0.20

0.06

0.10

plain

1327

82.30

1245

87.80

2815

84.50

totals

1613

99.86

1418

100.01

3319

99.29

98 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

10. tool-impressed (PI. 10) one side of the impression is usually deeper than
the other. If the angle is acute, a 'tear drop' effect is produced.

11. punctate (PI. 10)-the characteristic shape is round. Punctations at the
Midway site are usually smaller than tool-impressions.

All vessel shoulders in the sample are decorated. Most of the decoration is
poorly executed. A number of attributes, i.e. 58 and 1011, only occur on
shoulder sections detached from their rims. The following table indicates area
and site totals for these 1 1 attributes on all sherds.

Class XI: Shoulder Design Elements

Design elements consist of units of straight trailed lines, punctations, tool-
impressions, or of a combination of these decorative techniques. Dashes and
curvilinear lines are absent. The eleven basic design elements are illustrated in
Figure 1 and in Plates 11 and 12. Tables 39-40 record associations between
shoulder decoration techniques (Class X) and shoulder design elements (Class
XI). Attribute associations for shoulders with attached rims are listed in Tables
18-20.

Plate 8. Shoulder Decoration Technique: 3-wide line, trailed (1/2 natural size)

The Midway Village Site

Plate 9. Shoulder Decoration Technique: 4 finger, trailed; 5 fine lined, trailed
and tool-impressed; 6 medium lined, trailed and tool-impressed (1/2 natural
size)

100 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

HMIJ

Plate 10. Shoulder Decoration Technique: 8 finger, trailed and tool-impressed;
9 fine line, trailed and punctate; 10 tool-impressed; 11 punctate. (1/2 nat-
ural size)

The Midway Village Site

101

.00,
o.

o
o
o
o

oooo

A A

Figure 1. Shoulder Design Elements

1 02 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

Table 13. Shoulder Design Elements

Area I

Area II

Site Total

design

no. %_

no. %

no. %_

15 25.4

6 19.4

23 25.0

0.0

2 6.5

2 2.2

5 8.5

5 16.1

10 10.9

29 49.2

14 45.2

43 46.7

1 1.7

0.0

1 1.1

3 5.1

1 3.2

4 4.3

0.0

1 3.2

1 1.1

3 5.1

0.0

3 3.2

2 3.4

0.0

2 2.2

10
11

\A \ \ S

totals

... _.A. \f '

1 1.7
0.0

1 3.2
1 3.2

2 2.2

1 lei

59 100.1

31 100*0

: I

92 100.0

".'"----"' ,--,"'--'' /*

\ ../

% X

Class XII: Location

of Decoration

Decoration occurs on 3 main areas
edge of the inner rim. No undecorated
consist of the following combinations:

of the Inner Rim; 3. Shoulder, lip, and

; \ '

of the vessels: lip, shoulder, and upper
pots were found. Attributes in this class
1. Shoulder; 2. Shoulder and Upper Edge
Upper Edge of the Inner Rim.

'- '. 'i

; -. .

/ .,

Table 14. Location of Decoration

f / /. ft

f ' r
f location /

/I i

Area I
no. %
13 81.3

Area II
no. %_
7 41.2

Site Total
no.
19 61.3

3 18.8

9 52.9

11 35.5

3
totals

0.0

1 3.2

16 100.0

17 100.0

31 100.0

The Midway Village Site

103

Plate 11. Shoulder Design Elements: 1 (A-no. 34 in Table 19, B-no. 54 in
Table 19), 2 (no. 11 in Table 20), 3 (no. 57 in Table 19), 4 (no. 56 in Table
19) (1/2 natural size)

104 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Plate 12. Shoulder Design Elements: 6, 8, 9, 10, 11. (3/4*natural size)

The Midway Village Site

105

Class XIII: Appendages

This class is composed of 7 attributes defined by shape, method of attach-
ment, and location of attachment (PI. 14). The attributes are:

1. equilateral strap handle, welded to body and lip

2. flaring strap handle, welded to body and lip

3. equilateral strap handle, welded to body and rim below lip

4. detached equilateral strap handle

5. detached flaring strap handle

6. loop handle, welded to body and lip

7. detached loop handle

A-l

A- 2

A- 3

5 cm.

Plate 13. Handle Profiles.

(Letters indicate rim type in Class IV. A-l, A-3, B, and I are nos. 28, 67, 66,
and 8 in Table 20. A-2 is no. 15 in Table 19. Interior surfaces are at left.)

1 06 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

Two basic types of handles are represented in this class: strap and loop. These
were the only kinds of vessel appendages found. Loop handles are approximately
round in cross section and range from 914 mm. in diameter. The mean dia-
meter is 1 1 mm. The width of strap handles is usually constant along their
lengths, ranging from 22-44 mm. for the different handles. The average width
is 30 mm. and the average thickness is 11 mm. Most of these handles-flare from
25 mm. Three handles flare markedly and are placed in a separate grouping.
The difference between the width of the body of these handles and the end
attached to the rim ranges from 1117 mm.

The number of handles attached to a vessel is probably 2. No rims were found
which would indicate that four handled vessels were used at the site. Some
handled vessels were used at the site. Some handle profiles are illustrated in
Plate 13.

Table 15. Appendages %\

appendages
1
2
3
4
5
6
7
totals

Class XIV: Handle Decoration Technique

Two techniques of decorating handles are present in the sample: trailed
(PI. 15, 1-3) and grooved (PI. 15, 4).

Area I

Area II

, "
Site Total p^.

no.

1 no. . 1

25.

14.

22.

33.

14.

50.

11.

33.

22.

16.

11.

14.

100.

99.

100.

The Midway Village Site 107

Table 16. Handle Decoration Technique

Area I

Area II

Site Total

technique

no. %_

no. %

no.

trailed

3 25.0

3 37.5

30.0

grooved

2 16.7

0.0

10.0

plain

7 58.3

5 62.5

60.0

^totals

12 100.0

8 100.0

100.0

Class XV: Handle Design Elements

Four basic designs are found on the handles (PI. 15): 1-two vertical trailed
lines; 2-three vertical trailed lines; 3 -parallel diagonal trailed lines from upper
left to lower right; 4 a single vertical groove.

6 1

Plate 14. Appendages (1 and 2 are 1/2 natural size; 4-7 ^e natural size)

1 08 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

Area I

Area II

Site

Total

no.

0.0

66.7

25.0

40.0

33.3

37.5

20.0

0.0

12.5

40.0

0.0

25.0

12 34

Plate 15. Handle Design Elements (4/5 natural size)

Table 17. Handle Design Elements

designs

4
totals 5 100.0 3 100.0 8 100.0

Miniature Jars and Bowls

Rimsherd fragments of two decorated jars and eight plain bowls comprise the
sample. One jar from Feature 10, level 2, has narrow (1 mm.) interrupted tool
impressions on the lip. The impressions are grouped in clusters separated by
smooth and rounded lip sections at least 20 mm. apart. The second jar has con-
secutive ovoid tool impressions 5 mm. in diameter on the upper edge of the
inner rim. This sherd came from Square 19, level 2. The [,-oveniences of the
bowls are: Sq. 1, levels 1 and 3; Sq. 2, level 1; Sq. 13, level 2; Sq. 14, level 2;
Sq. 16, level 2; Sq. 19, level 2; Surface. Three rims are from Area I and four
from Area II. The bowl fragments vary in shoulder thickness from 2-7 mm. and
in lip thickness from 2-4 mm. All rims are too small to discern complete vessel
shape.

Pottery Types

All but two of the shell tempered Vessels' iii the sample fall within the range
of the proposed type ALLAMA EE TRAILED (Wedel, 1959: 91-92; Henning,
1961:28-29). Alternatively, a perusal of toll's thesis (1960) and published re-

The Midway Village Site 109

port (1962) suggests that the sample can be divided into three main types:
ALLAMAKEE TRAILED, RED BANKS PUNCTATE, KOSHKONONG BOLD.
The latter type includes those sherds having finger trailing as a shoulder decora-
tion technique (Gass X, no. 4) in a parallel vertical line design element (Gass
XI, no. 1). RED BANKS PUNCTATE includes those sherds with "areas of punc-
tates (tool-impressions in this report) between plats of vertical lines or in tri-
angular spaces between converging sets of diagonal lines" (Hall, 1960:177) (Qass
XI, no. 6). The remainder of the sherds with two possible exceptions would fall
within the range of ALLAMAKEE TRAILED. The two divergent sherds are an
inslanting rim (PI. 5, G; no. 34 in Table 18) and a shoulder fragment with an
unusual design element (PI . 1 2, no. 1 1 ; Qass XI, no. 11).

Both Midway site samples are distinct from the Upper Iowa River Orr Phase
ceramics by the preponderance of certain attributes and the presence or absence
of others. The more important differences, most of which were noted by Wedel
and Henning, are listed below.

1. iips-The Midway samples contain a greater proportion of undecorated lips
associated with decorations on the upper edge of the inner rim. A 'common'
Upper Iowa River lip decoration te^rinique of tiny impressions made in the lip
surface is absent. Except for one miniature pot, interrupted tool-impressions
are not found on the lips of Midway jars.

2. rims-The majority of the impressions OL he upper edge of the im r rim
of Midway jars are not straight but slant left or right. Vertical rim profits and
undecorated rim sherds (including the lip) are not present in the Miu,,, y
samples. The mean rim heights of the Midway samples are 6 mm. shorter, than
the means given for Upper Iowa River jars.

3. handles-Many of the lllidway vessels appear to lack handles. Most handles
are welded to the lip. Neither the Turkey foot' design, nor clay 'crests' are pre-
sent on Midway handles.

4. shoulder area-There is no evidence in the samples from the Midway site
of the use of dashes, parallel zigzag lines, circles, or curvilinear lines of any kind
in the composition of the shoulder area design elements. Most trailed lines are
3-4 mm. in width. A more intensive decoration is effected on Midway ceramics
by the use of broader lines placed closer togethe- Tool-impressions are used to
border trailed lines. There are no undecorated vessel shoulders among the Mid-
way samples.

A perusal of the archaeological reports mentioning the material recovered by
McKern in 1929 indicate that the full range of variation for the site is not
available in the two samples discussed here. Griffin (1943) mentions a few sherds
containing a grooved paddle surface finish (p. 288), and illustrates the use of
circular decorations in a shoulder design element (Pa. CXLI, fig. 22), the use of
tool-impressions to decorate the outer rim (PI. CXLII, fig. 2), and the presence

1 1 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

of the notched loop handle (PI. CXLII, fig. 5). McKern (1945) illustrates the use
of horizontal trailed lines on the exterior surface of a vertical rim possessing a
small loop handle welded onto its surface (PI. 55, fig. 8), and a 'cambered'
rim (PI. 55, fig. 2; PI. 70, fig. 2). He also mentions the recovery of an imper-
forate pottery disc (p. 138).

Grit Tempered Pottery

Grit .tempered sherds account for only 1.7% of the pottery from the site. All
sherds are too small to determine design motifs or vessel shapes. The grit tem-
pered pottery displays a preponderance of cord-marking, possibly of a cord-
wrapped paddle variety.Of the cord-marked sherds, 82% have impressions from a
s cord with a R} twist and 1 8% have impressions from a z^ cord with a L[ twist.
Percentages for Areas I and II are 76.5 arid 81.5 for the latter grouping and 23.5
and 18.5 for the former.

Table 1 8 lists the frequency of the sherds of the above two groups and all of
the remaining grit tempered sherds according to decoration technique. The fab-
ric impressions were the result of a simple under-and-over weave pattern (i.e.
'plain* weave) applied to the vessel surface. The individual strands in the weave
are z| with a L[ twist. Trailed lines in the 'Plain* categories are from 1-2 mm. in
width. The parallel oblique rows (20 to the right from a vertical position) on the
rim sherd with straight dentate stamped impressions are 17 mm. long and 4 mm.
wide. Each rectangular dentate is 4 mm. long, 1.5 mm. wide, and about 1 mm.
in depth. The lip of this rim is undecorated and flat (PI. 16, A). A second rim
fragment (Sq. 9, Feature 10) is cord roughened on both surfaces and has two
horizontal rows of wedge-shaped punctations separated by a band of 5 parallel
horizontal incised lines on the exterior surface of the rim (PI. 16, B). The lip is

decorated by V-shaped notches cut into the lip after it had dried. Letters in the
second column for cord-marked sherds refer to single specimens in each group.
A plain sherd with trailed lines and punctates is illustrated in Plate 16, C.

Although the rim sherds are small, enough attributes exist on three rims to
warrant their tentative placement into types. These types are: 1) HAVANNA
TRADITION, NAPLES STAMPED, VARIETY DENTATE (Griffin, 1952:110-
112) (Sq. 4, 8-36" below the surface; A in PI. 16); 2) ANGELO PUNCTATE
(William M. Hurley, personal communication) (Sq. 9, Feature 10, 16-20" below
the surface; B in PI. 16); 3) MADISON CORD IMPRESSED, VARIETY Z
(Baerreis, 1953:12-15; William M. Hurley, personal communication) (Sq. 12,
9-15" below the surface; letter C in Table 16). The placement of these three
rims into types must remain suspect due" to the small size of the sherds and the
limited number of observable attributes. Rims on these sherds are flat.

The Midway Village Site 1 1 ]

Plate 16. Grit Tempered Sherds

Miscellaneous Clay Object and Secondary Modification

A small piece (15x10x6 mm.) of untempered, fired clay was found in Square
20, level 2. Secondary modification occurs on a small shell tempered body sherd
in the form of a hole 5 mm. in diameter drilled from the exterior surface. This
sherd from Square 1 2, level 1 is decorated with trailed lines 3-4 mm. in width.

CHIPPED STONE ARTIFACTS

Scrapers

Scrapers were by far the most abundant stone artifact found at the Midway
site. Seventy-seven end scrapers were found IN SITU, for example, as compared
to eighteen projectile points. Scrapers are separated into end and side categories
depending on the placement of the main scraping edge.

End Scrapers

End scrapers are sorted into six groups depending in part on material com-
position and in part on the presence or absence of selected attributes such as
bulb of percussion, unmodified striking platform, or a ventral surface with 'a
single flake scar. Each of these groups, with one exception, is separated into two
sub-groups depending on whether the proximal end of the tool (the 'hafting'
end) is (A) less than half the maximum implement width or (B) more than half
the maximum width. Additional data, such as length, extent of secondary re-
touching, and color are included in Tables 22-25. Also included in the tables is a
column labeled "Cross section," which indicates if (A) the maximum thickness
adjoins the scraping or distal edge with the dorsal surface sloping toward the
proximal end, or (B) the thickness is uniform over at least half the scraper length
and the dorsal surface is flat or longitudinally ridged, or (C) the surface of the
tool is highly irregular. Other symbols used are: no secondary retouch, + sec-

1 1 2 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Table 18. Grit Tempered Pottery Decoration Techniques

cord

Plain with:

I 1

o cu

/ K

T3 s~*

H CO
14-1

twist

C CO

CO ^

U-l

G M

H 3

4-)

T3 4-1

4-1

. 0)

CU CO

4-1

5 e

a, o

O
G

H
CO

H O
CO G

CO
i-H

4-1

T3
H

CU >-l

Q *w

4-1

S-i 3
4-1 CX

CU
T3

15-19

19-24

6-10

10-16

8-36

ca.36

6-13

11-18

11-15

15-20

2a,d

20-28

F10

16-20

6-10

9-15

15-21

18-26

10-16

16-24

24-33

12-18

7-13

25-31

9-15

15-21

10-18

24-30

30-36

42-48

48-54

10-18

24-34

a - exterior surface plain with narrow trailed lines and punctates;
interior surface displays the cord-marking recorded in the table

b - similar cord-marking on both surfaces; two rolls of wedge shaped

punctates

sz
z

c - contains a single cord impression with a Z cord and a L ^ twist

sz ]

L|
d - contains cord impressions with a S cord and a R twist

zl Lr

s r

e - exterior surface contains trailed lines

The Midway Village Site

113

Table 19. Attribute Associations of Shell Tempered Jars in Area I

l-c 0)

CLASS

O U-t

M
td

CU QJ

t_^

i i

"O E
U

no.

(U 0)

i i

1 1

f";

i i

1 1

> i

IH -H

4/45

2/23

3 -

2/1

2/4

4/44

2/4

3/38

3/4

2/25

3/4

4/4

6/5

4/46

6/5

7/5

4/43

3/39

_ .

2/29

all

4/40

- -

3/31

, _

1/7

4/47

9 R

_ "

4/41

2/25

3/31

_ .

4/47

_ _

*pleces of this jar also found in level 3

1 1 4 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Table 19. (cont.)

CO
M (U

U-l

LASS

no.

CD CD

(U 0)

rH rH

i i

M
M

i i

M
I I

i i

M
M

i i
i i
i i

h-i

M
M
X

i i

CX
H C

i 1 'H

41
42

6
7

43
44

7
7

3/36

45
46
47
48
49
50
51
50

7
8
8
8
8
8
8

3
2
3
3
3
4
4

1
1

3
3
4

3
3

C
A
F

4/46
3/37
3/31

3'
3

2
2
2
2

2
2

1
2
1

2
2

;

53
54*
55
56
57
58'
59
fin

8
8
9
9
9
9
9

5
5
1
2
3
3
2/10

9 /I fi

1
1

1
1
1

3
3

1
1
4

5
3
2
2

B
B
A
A
A

4/41
4/42
4/40
4/41
4/42

3
3
2

2
2
2
2
2
2
2

2
2
2

1
1
1
2

4
4

1
4

1
1

4
3
1-

1
1
2

;

27
38
40

9 /i n

62
63
64
65
66
67
68
69
70
71

9
9
9
9
9
9
9
1
1

Z/1U

3/10
3/10
3/10
3/10
3/10
3/10
3/10
1
4

9 //

2
1
1
1
1
1

1
3
4
4
4
4

7
3

E
F
D

2/28
3/37
2/22

3/31

3
3
1
1
4
2

1
1
3
3

2
2
2

2
2

2
2
2

1
2

1
3

5
4

7
7

3
3

72
73
74
75
76

77
70

1
5
6
6
8

9
q

2/4
2/1

1
1
2
2

4
a /i n

4
4
4
4
4
5

3
3
2
3
3
3

- .

3/1C

* pieces of this jar also found in: Area I - t.s.2, L#l ; t.s.8, L//4 ;
t.s.9, L#3, F#10. Area II - t.s.20, L#2.

1 pieces of this jar also found in: Area I - t.s.6, L//2. Area II -
t.s.20, L//2.

The Midway Village Site

115

Table 20. Attribute Associations of Shell Tempered Jars in Area II

no.

I square

1 level or
level/feat. 1

CLASS

i i

i <

M
H

i i

M
M

h- 1

X
M

i i
X

M
1 I
X

M
M
( 1
X

2/27

2/22

1/14

4/42

3/38

2/27

-39'

3/32

2/27

4/41

2/25

3/38

4/42

2/25

3/39

4/41

1 1 6 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Table 20. (cont.)

t-
M a

CL,

\SS

4-l

QJ
M

i-H tH

0) (U

i i

no.

0) 01

i i

H- i
I 1

i i
i i

i i

iH iH

| |

I I

t i

4/41

42*

4/41

43'

4/42

4/44

4/42

3/37

-2

4/48

: _

_ '

3/39

J
1

3/32

_ ,

5/52

1 _

3/33

2/27

- _

2/26

3/39

' 1

4/42

I. _

Fill

* pieces of this jar also found in: t.s.2, L#l; t.s.8, L//4+5; t.s.9,
L//3, F#10.

pieces of this jar also found in: t.s.6, L#2; t.s.9, L#3.

The Midway Village Site

117

Table 21. Attribute Associations of Shell Tempered Jars in Surface Sample

no.

SURFACE 1

CLASS

1 1
l-l

i i

M
M

i i

h- 1

M
H

X
M

i i
X

1 1
M

i i
i i
i i

4/49

2/28

2/25

3/38

4/42

2/28

ondary retouching along the entire edge, +* secondary retouching along the dis-
tal half only, .+* retouching executed from the ventral surface. These latter sym-
bols refer to the presence or absence and extent of lateral retouch.

A definition of each group is given below. These groupings are a reflection of
the research problem for which this report was initiated, i.e. to test the cultural
homogeneity of the two areas excavated. If the scrapers recovered from the site
had demonstrated greater area diversity, the minutiae with which these group-
ings are defined would have been unnecessary. Group means and the relative
frequency of scraper groups by Area are listed in Tables 22-26.

Group 1 -Plano-convex chert flakes with a ventral face consisting of a single
flake scar containing a pronounced bulb of percussion and an unmodified
striking platform at the proximal end. Sides are usually partially retouched. Dor-
sal surfaces consist of a ridge or a flat surface formed by the removal of a series
of flakes. Group 1 is defined by the possession of three main attributes, i.e. a
pronounced bulb of percussion, an unmodified striking platform, and a ventral
surface consisting of a single unaltered flake scar. Each of the next three groups
differ from Group 1 by the absence of one, two, or all of these attributes res-
pectively. (PI. 17, A and B; Table 22)

Group 2 Similar to Group 1, except a pronounced bulb of percussion is
absent. (PI. 17, C and D; Table 23)

Group 3 Similar to Group 1, but both the pronounced bulb of percussion
and the striking platform are absent (PI. 18, A and B; Table 23)

Group 4 Similar to Group 1 , except the ventral face consists of a number of
primary retouching flake scars in addition to the absence of both a pronounced

1 1 8 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

bulb of percussion and a visible striking platform (PI . 1 8, C and D; Table 24)

Group 5 The ventral face of these plano-convex chert end scrapers consists
of a single flake scar. A pronounced bulb of percussion and a visible striking plat-
fbrm are absent. The dorsal surface has been flattened by the removal of one or
more flakes. Steep retouching occurs around the lateral and proximal edges of
the implements in such a fashion that they have the appearance of flat topped,

f f

^ijf

Plate 17. End Scrapers: A-Group 1A, B-Group IB, C-Group 2A, D-Group 2B

four-sided pyramids. This symmetrical shape is characteristic for the grouping.
All proximal ends are at least half as wide as the maximum width. (PI. 19, A
Table 24)

Group 6-Similar to Group 3, except the stone is quartzite rather than chert.
(PI. 1 9, A and B; Table 24)

Miscellaneous End Scrapers-Eight end scrapers do not conform to any of the
above groups, because they are broken or have an unusual combination of at-
tributes. These scrapers differ in the following manner:

nos. 1, 2, 7 in Table 25: dorsal surface formed by a single flake scar; striking
platform absence; ventral surface modified by the removal of several flakes;
chert.

no. 3: scraping edges on opposite ends of the dorsal and ventral surfaces;

The Midway Village Site

119

5 Cm

Plate 18. End Scrapers: A-Group 3A, B-Group 3B, C-Group 4A, D-Group 4B

5 Cm

Plate 19. End Scrapers: A-Group 5, B-Group 6 A, C-Group 6B

120 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

lateral

dimensions

provenience

retouch

in mm.

i 1

W -H

4-1

color

ctf

W 4-1

toO

4-1

M-l

$-t

>-i 0)

Cl)

f-j

no.

U CO

Group 1 - A

gray

+ '

tan with white

white + brown

light orange

gray

gray + white

+ *

gray

white + pink

light orange

gray

light brown

gray

light brown

gray

pink

Group 1 - B

Fl-4

white + brown

F5-6

gray

F5-6

+ '

white

gray

+ '

white

dark brown

gray

Table 22. Group 1 End Scrapers

The Midway Village Site

121

lateral

dimensions

provenience

retouch

in mm.

*-.. M

H 3

CO -H

color

0) -ui

CO 4-1

4-1

(!)

> CO

M-t

cr 1

<U OJ

M QJ

no.

Cfl

.-4 Pn

C_) C/5

,-J

Group 2 - A

gray

NVC

white

light tan

+ '

white

gray

light tan

gray + pink

Group 2 - B

F5-6

+ *

7 ^

white

gray

light tan

Group - A

gray

tan

gray

white

light orange

white 4- gray

Group 3 - B

gray + brown

yellow

gray

+ *

light tan

gray

Table 23. End Scrapers: Groups 2 and 3

122 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

lateral

dimensions

provenience

retouch

in mm.

M" 3

CO -H

4-1

j-.

^ >

color

<D 4-1

cn 4-i

4-1

jjj

4-1

> CO

*4-l

^ cu

.-J PH

QJ CO

Group 4 - A

gray

white

dark red

white

gray

pink

Group 4 - B

F4/2

gray

-36"

gray + white

brown

gray + white

brown

gray

gray + white

gray

white

Group 5

F5/5

gray

Group 6 - A

F4/2

dull yellow

F7/1

white

Fll

white + red

dull yellow

Group 6 - B

dull yellow

F14

white

Table 24. End Scrapers: Groups 46

The Midway Village Site

123

lateral

dimensions

provenience

retouch

in mm.

\
cu

color

3 t-H

CO H

4-1

4-1 QJ

CO 4-1

4-1

cfl >

<4-l

>-l

0) 0)

>-i CU

fe ,-4

O CO

Fll

gray

white

gray

white/orange

- no secondary retouch

+ secondary retouch along the entire edge

+* secondary retouch along the distal half only

A maximum thickness adjoins the scraping edge; sloping dorsal

surface

B thickness uniform over at least half the scraper length;
doral surface flat or longitudinally ridged
C highly irregular surf ace (s)

Table 25. Miscellaneous End Scrapers

nos. 4-6; broken transversely near the proximal end; chert,
no. 8: markedly biconvex; pronounced bulb of percussion; other attributes
similar to Group 6B; quartzite.

Side Scrapers

For purposes of this analysis, a side scraper is defined as a flake that has
steep unifacial retouch on one or both lateral edges. These tools have been man-
ufactured with less care than most end scrapers. This is apparent from both their
highly irregular shapes and sizes, antf from the limited extent of secondary re-
touching. Side scrapers have been divided into two groups on the basis of ma-
terial composition.

124 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Dimensions in Millimeters

Group

Area +

Site Total

Length

Width

Thickness

22.5

21.0

7.3

1 a

26.9

20.0

8.1

site

25.8

20.0

7.9

22.3

18.8

7.5

1 b

19.0

19.7

5.7

site

21. D

18.8

6.5

35.5

19.5

7.8

2 a

22.8

17.5

7.0

site

29.6

19.1

7.6

22,5

15.5

5.5

2 b

22.0

16.0

7.0

site

22.8

15.8

5.8

17.5

15.0

4.5

3 a

26.0

21.0

7.5

site

23.4

19.6

6.3

28.3

18.3

8.3

3 b

23.0

18.0

6.6

site

23.8

17.4

6.9

17.0

15.0

4.0

4 a

17.0

19.0

6.0

site

22.7

18.0

7.2

38.6

23.2

8.2

4 b

38.2

22.5

9.2

site

38.4

22.8

8.7

15.0

18.0

6.0

site

15.0

18.0

6.0

50.5

34.0

10.0

6 a

31.3

21.0

8.0

site

38.0

28.1

8.9

34.5

26.0

7.5

site

34.5

26.0

7.5

Table 26. Mean Dimensions of End Scraper Groups in Millimeters

The Midway Village Site

125

Group 1 Chert (6 specimens illustrated in PI. 20)

These scrapers are irregular in shape and pattern of manufacutre. However,
all possess a plano-convex cross section, all lack a pronounced bulb of percussion,
and, with the exception of the surface find, all lack a striking platform. Dorsal
and ventral surfaces of scrapers 1 , 2, 4, and 5 in Table 28 have been modified by
the removal of a number of flakes. The ventral surface of the other two scrapers
contains a single flake scar. Both lateral sides of scrapers 2, 4, and 5 have scrap-
ing edges. The column labeled 'length of scraping edge' in Table 28 contains the
measurement of the longest edge for these latter scrapers. Scraping edges are
either straight (1, 4, 6) or convex (2, 3, 5).

Group 2 Quartzite (10 specimens; several illustrated in PI. 21)

All of these implements are more or less plano-convex in cross section and ir-
regular in shape. Ventral surfaces of all scrapers except nos. 10 and 14 in Table
28 contain a single flake scar. Scrapers nos. 10, 11, and 13 retain both a pro-
nounced bulb of percussion and an unmodified striking platform. Scraper no. 7
has a bulb of percussion, but lacks a striking platform; no. 12 has a striking plat-
form, but lacks a bulb of percussion. The remaining scrapers lack both of these
attributes. Scraper no. 8 consists almost entirely of cortex, except for the nar-
row scraping edge (PI. 21, A-left). Scraping edges are convex (7, 13, 16), con-
cave (12), or straight (8-11, 14).

Group 1

Group 2

Group 3

Group 4

Group 5

Group 6

i i
nJ

13.8

6.9

10.3

3.4

17.2

6.9

99.9

27.5

7.5

10.0

2.5

10.0

12.5

2.5

15.0

7.5

5.0

100.0

Table 27. Relative Frequencies in Percentages of End Scraper Groups Fro
Area I and Area II

Drills

Two broken drills were found in Area I. Both are carefully retouched and
worked on all surfaces. They are long and narrow with the point of maximum
width near the basal end. The points of both tools are broken. One of the drills

126 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

(Sq. 1 , level 4) is bi-convex in cross section and bears smoothing from use on
both surfaces. It is of white chert and has maximum dimensions of 35x11
x5 mm. The other drill (Sq. 6, level 2) is made from white quartzite and is also
bi-convex in cross section. The broken point was partially reshaped and bears
some signs of use. Both drills are illustrated in Plate 22, A.

Miscellaneous Perforators

Two 'perforators' were found in Area I. One of these tools (Sq. 1 , Feature 4,
level 3) is a symmetrical, bi-convex ovate form with bifacial retouch along the
sides of the distal or pointed half. The remainder of the implement is shaped by
primary retouch over its entirety. The 'other perforator (Sq. 6, level 1) has the
shape of a scalene triangle and a concave-convex cross section. This tool is re-
touched only on the point. The maximum dimensions of the first perforator are
48x27x19 mm. and of the second 34x19x7 mm. Both of these tools are illustra-
ted in Hate 22, B.

Plate 20. Group 1 Side Scrapers (A-no. 6 in Table 27; B-nos. 1 and 2 in Table
28; C-nos. 3-5 in Table 28)

The Midway Village Site 1 27

$ Cm

Plate 21. Group 2 Side Scrapers (A-nos. 7-9 in Table 28; B-nos. 10-12 in Table
28; C-no. 14 and 16 in Table 28)

Projectile Points

Small, unnotched, triangular points common to the Mississippian Pattern in
Wisconsin are the most common projectile points at the Midway site. Four of
the twenty-four points collected from the site are stemmed. These are forms not
ordinarily associated with an Upper Mississippian complex. Triangular points are
divided into two categories on the basis of material composition.

128 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

provenience

0)
00

lateral
retouch

dimensions
in mm.

'"H pq

^-s^,

3 .H

4-> -H

4-1

color

4- 0)

tUj ^

4-1

03 >

G l-i

(4-1

no.

<U 0)

-J 3

Grou

pi 1

F8/2

white, pink

gray + white

gray

gray + red

gray

gray + white

Group 2

F4/4

pink

dull yellow

F10

white

dull yellow

white

+ *

white

pink

- no secondary retouch

+ secondary retouch along the entire edge

+* secondary retouch along the distal half- only

Table 28. Side Scrapers

The Midway Village Site 1 29

5 Cm

Plate 22. A-Drills B- Miscellaneous Perforators

Group 1-Chert (7 specimens; PL. 23, A and C)

Most of these unnotched triangular points h 'e their greatest length air ig the
medial line (from the apex to the mid-point of the base), but some irregul -rities
are apparent. The lateral sides are usually longer than the width of the imple-
ment. All of the broken points in the sample have snapped transversely. The lat-
eral sides are either convex and symmetrical (2-4 and 7 in Table 29), or irregular
and asymmetrical (5 and 6 in Table 29). The bases of points 2-4 (Table 29) are
straight and the base of no. 7 in convex. Point 1 is broken transversely. Asym-
metrical points are very crudely fashioned, while symmetrical forms are re-
touched over both surfaces.

Group 2-Quartzite (13 specimens; PI. 24)

Most of these points are symmetrical in outline and crudely worked. At-
tributes for this group are included in Table 29. ' .ie lateral sides of the points
are either convex (1, 7, 10, and 12 in Table 29), straight (6, 11, 13), or irregular
(5, 9). Bases are convex (1, 6, 10, 12) or stra^it (7-9, 11, 13). Retouching oc-
curs on all surfaces of a few of these implerrents (1, 3, 8, 11, 13), but the ma-
jority are crudely fashioned and lack extensive retouch (4-7, 9, 10, 12). Speci-
mens 2, 4, 7, and 1 1 are broken transversely. A cortex remnant adheres to the
surface of point no. 12.

The four stemmed projectile points of chert are listed by provenience in
Table 29. Points 1 and 4 in Table 29 consist only of the stem and base. Both
bases are retouched from both surfaces, but grinding or rubbing is absent.

130 WISCONSIN ARCHEOLOGtST

Vol. 51, No. 3

Grinding is quite distinct along the lateral edges of the stem of point 3. The
edges of the blade of this point are slightly convex and are neither beveled or
serrated. This point is similar to specimens recovered by William M. Hurley in
the Kickapoo Valley (1965: Fig. 4, nos. 6 and 19; Fig. 8, no. 24; Fig. 13, no.
20; Fig. 16, no. 20).

Knives

(13 specimens)

A flat, thin, light brown quartzite flake knife retouched on both surfaces is
the only unbroken knife found at the site (second from left in A, PI. 25). This
tool came from Square 10, level 1, and has maximum measurements of 51x24x7
mm. Attributes of knife fragments are- listed in Table 30. Six of the fragments
are blade tips (1-6 in Table 30); three are midsections (7-9); and three are bases.

provenience

dimensions in mm.

(!)

color

4-1

i i

4-1

no.

Group 1 - Chert

(18)

white

brown , gray

gray

white

brown

light brown

brown

Group 2 - Quartzite

purple

yellow

(11)

yellow

pink

white

(18)

white

(13)

white

yellow

Stemmed Projectile Points

white

white, pink

(51)

pink, red

(33)

white, pink

Table 29. Projectile Points

The Midway Village Site

131

4 *

Plate 7 J. Projectile Points. A-Group 1 (nos. 1,-, 4, 3, 2 in Table 28), B-
Stemmed Points (nos. 2, 1, 3 in Table 28), C-Group 1 (nos. 6 and 5 in Table
28).

Plate 24. Projectile Points: Group 2. (A-Surface, nos. 10 and 9 in top row and
11 and 12 in bottom row in Table 28; B-Area II, top row: nos. 6, S, 4 in
Table 28, bottom row: nos. 8 and 7 in Table 28, C-Area I, nos. 2 and 3 in
Table 28)

132 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

(10-11). All are chert except two quartzite tips (2, 6) and a quartzite midsection
(8). All fragments except no. 10 have been shaped by primary bifacial retouch
and secondary retouch along the edges. Knives are illustrated in Plates
25 and 26 (B and C).

Table 30. Knive Fragments

provenience

dimensions in mm.

color

p-t

no.

tip

gray

white

gray.

, - - '"

.*'_ '.

white

-S

..-

, -;

: '

. . - ,

_- -

gray

1 *s

white

mid-

gray

section

dull purple

gray

base

(31)

gray

(32)

gray

S-

(29)

gray, orange

Core Tools

Eight irregularly shaped dull yellow quartzite implements are classified as
core tools (PI. 29; Table 31). It is difficult to determine whether all of these
were actually used as tools, or if some are discarded cores. The jaggedness of the
Vorking' edges and signs of crushing suggest that they are general purpose tools
employed for chopping, cutting, or possibly scraping. No standardized size or
pattern of manufacture is apparent. Three are irregularly shaped and have a
single working edge (3, 5, 6 in Table 31). The other five have bi-convex bodies
and two fairly parallel working edges. All have been shaped by bold percussion
flaking.

The Midway Village Site 133

5 Cm

Plate 25. Knives. (A-nos. 3, -, 11, 10 in Table 30; B-no. 1 in Table 30; C
nos. 5, 4, 12, 6 in Table 30)

134 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

5 Cm

Plate 26. Cores and Knives. A-Cores (nos. 1-3 in Table 32), B and C-Core
Knives (nos. 7 and 8 in Table 30).

Table 31. Core Tools

no.

Area

Square

Level/
Feature

Length
ram.

Width
mm.

Thickness
mm.

F5 L7

The Midway Village Site 135

Cores

Nine flake cores were found at the site (PL 26, A; Table 32). Small 'ribbon'
flakes were removed from the smaller cores.

Hammerstones

A few pieces of stone are roughly sperical and have irregular, pitted or
crushed surfaces (PL 28, A; Table 32).

C
Plate 27. Core Tools

1 36 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

ttaie ^8. Hammerstones (A) and Miscellaneous Ground Stone (B). (A nos.
3, 2, 1 in Table 32; B-no. 1) (B is 12 cm. in length.)

provenience

dimensions in mm.

color

3 r 1

*7j

4-1 QJ

txO

CO >

no.

QJ QJ

cores

F5/8

chert

dull yellow

gray

quart-

dull yellow

zite

dull yellow

F10

dull yellow

F10

white

chert

gray

hammerstones

1
2

I
I

9
9

3
F10/3

quart,
chert

63
66

47
42

dull yellow
dull yellow

F10/3

quart .

dull yellow

Table 32. Cores and Hammerstones

The Midway Village Site

137

GROUND STONE ARTIFACTS

Abrading Stones

Four ferruginous sandstone abraders were ground into similar shapes before
use (PI. 29; Table 33). Except for no. 3 in Table 33, all of these tools possess at
least one groove that could have been used in smoothing arrowshafts. A brief
description of each abrader is given below:

Abraders 1 and 3 in Table 33 contain multiple grooves (PI. 29, B and A
right). The latter tool has three parallel longitudinal grooves, semicircular in
cross section, running along the length of one surface. The grooves are all 4 mm.
in width, but range from 1-2 mm. in depth. Two longitudinal grooves run the
length of the tool on the reverse surface. One groove is 2 mm. wide and barely
perceptible. The other is irregular, varying from 3-8 mm. in width and 1-2 mm.
in width. Both sides contain short irregular gashes, which have-*been worn into
their surfaces. The gashes are 1 mm. wide, 1 mm. deep, and vary from 8-13
mm. in length. Except for the grooves and three gashes, all four sides are
smooth. Both ends appear broken. Abrader no. 1 has a single, semicircular, lon-
gitudinal groove, 11 mm. in width ard 5 mm. in depth, running the length of
one surface. Two smaller shallow >, joves, 2 and 3 mm. in width, run nearly
parallel along the length of the reverse side. A similar groove parallels these
along one narrow side. 1J>is, groove,, is 2onm. wide, quite >. -shallow,. an4, distinct. *>
Both ends are broken. ft >M< 1 tA >

Both abraders 2 and 4 in Table 33 have a single central longitudinal roove,
semicircular in cross section (PI. 29, A left and C). These grooves run the It- ^h
of the tools and measure 7 mm. in width and 4 mm. in depth for the former and
8 mm. and 1 mm. for the latter. The remaining surfaces are smooth and rounded
except for the ends, which are all broken with the exception of one rounded
end on no. 4.

:: .r. i:: ... : t^vjtx*' :-..

- N.'ti ' ; JlViit

' .. f.

? ?*:

,i'-;.c ir. .".;. - .;=-.> of* I .^- r .. U:. -;;" -t .- C- ;- -^ .

Table 33. Abrading Stones :; 7 , : ^ : ,; r , , :n - ., ,--, .. i, 5 ?

Test

Feature

Level

Length

Width

PhTcfciiess '*&

. ., , - - ' - ' '

'' . . 1 1_ i

- r- ' , .

no.

Square

" ' ft "

. cm.

cm.

! .;;::v

4 : ' t<

-,. ~,'

Hti -

3.6

2.*' bni

2.3 .

'3i,r

i>- -r ?*,

' "; 4

;. . .- ''

; 4 . 7 '

V 4.1- :

2.*2-

, ; l->, r

'j>:;,'-- ' !

- 2 : r .'

-,'4.9:.!

4. 8 ->.

2. '2 '

138 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

5 C

Plate 29. Abrading Stones. (A-nos. 2 and 3 in Table 33; B-no. 1 in Table 33;
C-no. 4 in Table 33)

Grinding Stone

A fine textured sandstone implement has parallel sides and rounded ends of
different sizes. Grinding has rounded the larger end into a curved butt, the sur-
face of which is smooth and flat in the arch it perscribes. The stone was clearly
employed with a rocking motion when used on this end. The smaller end is
pitted and may have been used for 'hammering.' The tool was found in Square
1 , Feature 4, and has maximum dimensions of 9.5x7.3x6.8 cm.

Catlinite Fragments

Three pieces of ground and sawed catlinite were found at the site. One piece
(Sq. 8, level 4) has maximum dimensions of 2.2x0.4x0.3 cm. (PL 30, E right).
A second (Sq. 19, level 2) has maximum dimensions of 2.7x1.7x0.6 cm. (PL
30, E left). Neither piece has an apparent function or design. The third piece
(Sq. 20, level 6) is part of a small disc with a small semicircular perforation in
the center (PI. 30, D). Dimensions of the broken disc are 1.9x0.7x0.3 cm.

The Midway Village Site

139

I *

5 Cm

Plate 30. Copper and Catlinite Artifacts

Miscellaneous Ground Stone

Five worked stones are either too small to identify with any functional cate-
gory or do not fit into any of the above categories. They are: l)a large rounded
pebble (12x3x1.5 cm.) from Square 3, level 2, with pecking marks on both flat
sides and on one rounded end; 2) a small fragment of what may be a sandstone
abrader from Square 8, level 5, has a semicircular groove 25 mm. wide and 2
mm. deep; 3) a small fragment of a ground stone disc less than 5 mm. thick
from Square 20, level 1 ; 4) two large stones of quartzite from Square 6, level 2
and Feature 14 are rounded on one end. No. 1 is illustrated in Rate 28, B. The
quartzite stones in no. 4 may be shaped by weathering or other non-cultural
factors (H. 31).

140 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Plate 31. Miscellaneous Ground Stone: no. 4 (A-Sq. 19, Feature 14; B-Sq. 6,
level 2)

Utilized Flakes

Of the 3207 flakes found IN SITU at the site 387 or 12.1% are either bi-
facially or unifacially retouched along part of one or more edges. The mean
length of these flakes is about 44 mm. and the maximum length is 74 mm. 1.9%
of the 3207 flakes have bifacial retouching along the margins, while 10.1% are
unifacially retouched. 44% of the 62 bifacially retouched flakes are from Area I
and 56% are from Area II. Of the 325 unifacially retouched flakes, 45% are
from Area I and 55% are from Area II. These computations and those in Table
34 include only those flakes found IN SITU. The distribution of utilized flakes
is listed in Tables 39-40.

The Midway Village Site

141

Area

bifacial
retouching

unifacial
retouching

% of flakes
utilized

chert
no. %

quartz,
no. %

chert
no. %

quartz,
no. %

100

13.1

118

11.4

site
total

218

107

12.1

Table 34. Numbers and Percentages of Utilized Flakes by Material Composition
(quartz = quartzite)

Unmodified Flakes

Of the flakes found IN SITU 2820 or 87.9% bear no signs of retouching or
use. Forty-one per cent or 1159 waste flakes are from Area I and 59% or 1661
are from Area II. The above computations and those in Table 35 include only
those flakes found IN SITU. Table 39 and 40 include a listing of these flakes by
provenience unit. Approximately 80% of the chert refuse consists of a light gray
material with white blotches throughdut its matrix. Sources for this chert are
local. The largest unmodified flakes are of quartzite. In the computations in

4-1

.c
o

quartzite
no. / %

quartz

Area

no./ %

red

yellow

white

misc.

total

no. / %

no./ %

450
38.8

74
6.4

336
29.0

237
20.5

57
- 4.9

704
-60.8

4
0.3

1
0.09

697

119

491

326

954

t?l

42.0

.7.2

29.6

19.6

1.1

57.5

'0.5

0.06'

OJ cfl

1147

193

827

563 "'

1658

2 '

H
W 4J

40.7.

6.8

29.3

20.0

2.7

58.8

' 0.5

0.0.7.

Table 35. Numbers and Percentages of Unmodified Flakes by Material Compo-
sition and by Color for Quartzite

142 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3 .

Table 35, quartzite flakes are separated by color. Even though the range of
colors can be found in a single deposit or even on a large core, a preferential
selection of the colors might reflect differences between the areas tested. Of
the chert flakes 39% are from Area I and 61% from Area II. 41.5% of the un-
modified quartzite flakes are from Area I and 57.5% from Area II.

Less than 5% of all flakes from the site are small 'blades.' The remaining
-flakes are irregular in form.

BONE IMPLEMENTS

* - ' r -

The head of a bison scapula hoe and a small portion of its posterior border
(PL 32, D and E) were found at a depth of 14 inches in Feature 10. Polish along
one border of the jagged neck region suggests the tool may have continued in
use after breakage. Characteristic signs of wear are present along the lateral
edges of the glenoid cavity. Wear on the anterior side of the shattered sec-
tion of the posterior border indicates that a hole was present in the infraspinous
fossa. Two other bone implements came from Feature 10. A fragment with one
blunt spatulate end was also at a depth of 14 inches (PL 32, B). It measures
62 mm. in length and 19 mm. in width. A worked swan numerous was .23 inches
below the surface (P 1. 32, C). A gash had been made across one surface and a
section of bone 20 mm. wide and at least 44 mm. long removed. Shine on the
bone suggests that it was used as a tool. A small fragment of bone from a depth
of 40 inches in Feature 5 is cut and worked on one end and may be part of a
tube.

A worked deer metapodial from a depth of 48 inches in Test Square 20
is the only worked bone from Area II. The bone had been split longitudinally
and signs of wear are visible along the edges formed by the split. This imple-
ment is 65 mm. in length and 20 mm. in width.

COPPER ARTIFACTS

Worked copper was found only in Area I. A ring-shaped artifact 16 mm. in
diameter came from 15 inches below the surface in Feature 10; the ends of the
object overlap (PI. 30, B). A 73 mm. long double-pointed perforator somewhat
rectangular in cross-section came from Square 9 at a depth of 10 inches (PI.
30, C). Five partially disintegrated fragments of sheet copper measuring from
10-27 mm. in length and 7-17 mm. in width also came from Area I (PI. 30, A).
One fragment was found in Feature 1 and the other four in levels 3, 4, and 5 of
Square 8.

The Midway Village Site 143

Plate 32. Bone Implements. (B is 62 mm. in length)

FAUNAL REMAINS

Mammal and fish bones were identified by Charles E. Cleland of the Michigan
State Museum. Mollusk specimens were identified by the writer. Table 36 lists
the identifiable mammal and fish by species and the suggested percentage used
in the meat diet. Table 37 below contains a list of the identifiable molluscs.
The distribution of unidentifiable mollusk remains is recorded in Tables 39
and 40.

Single specimens of AMBLEMA PERUVIAN A were found in test squares:
1 (levels 4 and 7; Features 2, 4, and 5), 5 (levei 3), 6 (levels 3 and 4), 9 (Feature
10), 14 (levels 2 and 3), 18 (levels 4 and 5), 20 (levels 4, 6, and 7), 21 (levels 1
3). A single QUADRULA PUSTULOSA was in level 4 of Square 20. The STRO-
PHITUS RUGOSUS specimen came from Feature 2. All the remaining shells
came from a tightly packed concentration in Square 5 designated Feature 9.

144 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Table 36. Identifiable Fauna by species and suggested percentage used in meat
diet (excluding mollusks)

SECTION A SECTION B

SPECIES

05
0)

i i

h- 1
M

Ibs.
usable

-0

meat

X 'O

iH
r

per

6 4_l

C
S-i -H

individ.

species

C -H

H T3

fr^

Deer

170

1700

47.3

Dog

135

3.7

Beaver

31.5

258

7.2

Elk

700

350

9.7

Bison

1200

600

16.7

Bear

300

210

420

11.7

Muskrat

2.1

8.4

Woodchuck

5.6

11.2

1.1

Marten

' 1

' 2.1

2.1

Raccoon

17.5

Prairie pocket

gopher

Eastern mole

Snapper

Soft shell turtle

0.9

Painted turtle

Passenger pigeon
Swan

1
1

70
70

0.7
1.0

0.7
7.0

Common mallard

2.'

1.75

0.2

Wood duck

i. 1 .

1.05

Channel catfish
Catfish family

3
93

1
14

80
80

3.2
0.4

3.2
5.6

Longnose gar

1.6

3.2

Bowfin

2.'

2.0

18.0

0.9

Stizostodion spp.

OJ:

0.4

2.0

R. black bullhead

0.4

0.8

Sucker spp.

0.4

TOTAL

3576

100

1 Table 36 is slightly modified in presentation from the original data provided
by Charles E. Cleland.

The Midway Village Site

145

Table 37. Frequency of Identifiable Mollusks by Species

Amblema peruviana (Lamarck) 102

Quadrula pustulosa (Lea) 14

Quadrula quadrula (Raf) 1 1

Fusconaia ebenus (Lea) 1 1

Fusconaia undata (Barnes) 6
Ligumia recta (Lamarck) 1
Strophitus rugosus (Swainsom) 1

FLORAL REMAINS

Rant remains from the Midway site were identified by Richard A. Yarnell.
Table 38 lists the identifiable remains. Yarnell identified most of the wood char-
coal sent to him as oak. The plum pits have been identified as PRUNUS AM-
ERICANA (Marsh), the hickory nut as CARYA OVATA (K. Koch), the acorn as
QUERCUS sp. and the bean cotyledons as PHASEOLUS VULGARIS (Lam.).

Table 38. Plant Remains from the Midway Site (k=kernel)

square

level

feat

corn
cob

5k, cob
Ik

Ik, cob
Ik, cob

beans

other

2 plum pits

acorn shell

hickory shell

146 WISCONSIN ARCHEOLOGIST

Vol. Sly No, 3

Table 39. Listing of Unidentifiable Shell, Utilized Flakes, Unmodified Flakes,
Shell Tempered Body Sherds (Plain and Decorated), and Grit Tempered Sherds
By Provenience Unit and Level for Area I

i
1

LEVEL or FEATURE ]

DEPTH IN INCHES
FROM SURFACE

! unidentifiable shell

Utilized
Flakes

Unmodified
Flakes

Shell Tempered Pottery
Class X Code
Y Uji^ \

iH
(X

i grit tempered sherds

^"M
0)

quart zite

4-1

quar'tzite

''?'

1
2
3
4
5
6
7
8

Fl
M

F2
F3

F4
ii

F5
ir

a- 14

;4-20
20-24
24-28
28-32
32-36
36-40
40-44
12-18
18-24
24-30
9-22
8-23
11-20
20-24
24-30
32-36
36-40

> X
X

X
X
X
X
X
X
X
X
X
X

i 5'

1 -

' ^

' *i

34
'$

10;

12
5

s; ~'

2
1

. 7,

9
2

7
10
11

1
1

2
4

' -

s. ,-

1
4
2
3
6

j T*
1

2
3

5
3
2
-1
3

13
2
3
9
2
4

11
4
4
.,3.
4
4
14
9
11
12
28
37

1
2

, 1
1
1

1
1
3

14
8
9
' 6
8
4
27
12
18
12
27
46

'"-

r*.i

.;-

5
4

1
2
3
4

9-15
15-19
19-24
24-32

X
X

6
2

5
4
3

15
12

36
19
16

- 2

1
3

34
23
15

1
3

1
2
3
F6

6-10
10-16
16-24
16-18

1
4

19
3

15
11

3
2

2
2

37
34
7

1
1

8-36
ca.36

The Midway Village Site

147

Table 39. (cont. )

w
p

LEVEL or FEATURE |

DEPTH IN INCHES
FROM SURFACE

, i

0)
, |

Utilized
Flakes

Unmodified
Flakes

Shell Tempered Pottery
Class X Code

>-i

0)
en

toO

unidentif i

chert

quartzite

4-1
O

quartzite

1
2
3
F7
F8
F9

6-13
13-20
20-27
17-22
19-24
11-18

4
1

19
2

3
2

23
3

1
1

9
3

4
2

5
1

34
10
5

1
2
3

6-11
11-15
15-20
20-28

9
6
6
2

4
2

14
23
9
10

26
31
17
11

4
1
2

6
14
4

2
6

91
90
2

1
2

1
2
3

6-12
12-18
18-30

X
X
X

4
5

10
1
5

1
3

4
1

1
1

40
23
53

2
3
4
5

5-12
12-18
18-24
24-30
30-36

X
X

4
2
3
3

1
2

15
22
35
47
13

26
11
16
16
16

4
1

7
c

1
1
2
3
7

1
1
2

23
43
35
57
38

1
2

F10
it

6-12
12-18
18-24
24-29
29-34
34-38
38-42
12-16
16-20
20-25

1
3

3
2

4
9

23
4
6
3C

2
1

40
68
18
17

X
X

1
4
3

4
13
14

1
3

1
1

24
44
35

148 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

Table 40. Listing of Unidentifiable Shell, Utilized Flakes, Unmodified Flakes,
Shell Tempered Body Sherds (Plain and Decorated), and Grit Tempered Sherds
By Provenience Unit and Level for Area II

,-H

i-H

QJ
N CO

H cn

Shell Tempered Pottery

)-l
a;

H 0)

T3 C

iH r*i

O **?

w
W

0)
i I

H CO
4-1 i 1

S c
d i i

Class X Code

c_> w

p Pn

P f^

2 U

cti

>-i

P-.

14-4

M ;=>

[

4-1

,-J

w
>

g S

(X, O

4-1
J-i

4-1

Cfl

4-1
QJ

4-1

w
,-J

W 03
Q Pn

3
cr

i i
ex

6-10

10-16

16-20

steri]

9-15

15-21

21 ?S

8-13

13-18

18 26

-i

10-16

16-24

24-33

-11

32-45

6-12

12 18

"12

14-23

"13

ca.22

5-9

- _

915

15-21

7-13

13-19

The Midway Village Site

149

Table 40. (cont.)

r 1

i |
CU

l-l

Shell Tempered Pottery

N CO

H CO

H CU

T3 CU

I 1

H CO

-3

Class X Code

CJ W

4-1 rH

C rH

2 O

3 fa

M fa

14-1

M P

'jj

4-1

fa O

Q fa

7-13

13-19

19-25

25-31

31-38

9-15

15-21

21-28

-*i /

U9o

i ft

?1 S

Z j
-10 o -|

J.C

,
_L

* j-j

J-O J-L

10-18

18-24

24-30

30-36

O C. / O

A Q

42-48

X
X

M 1 -7

100

.
2

48-54

10-18

18-24

4 I -

24-34

1 50 WISCONSIN ARCHEOLOGIST Vol. 5 1 ; , ;^o, 3

INTERPRETATIONS
Intra-Site Homogeneity

Obviously the fragmented nature of the grit tempered potter and the ab-
sence of stratification of components in the site debris limit the extent to which
interpretations may be carried. The tentative placement of some specimens into
types suggests that the grit tempered sherds are part of earlier Middle and early
Late .Woodland components and,- therefore, not'difeetry^sbcIafedTwifK theOir
Prpsei material. Corroborating evidence, at the ilidway site for this interpretation
isj the juxtaposition of extended burials containing shell temperesd pottery qn
circular burials containing sherds of the Woodland type (McKern, 1945U67-
168); Possible grit tempered inclusions in the Mississippian component are khos'e
sherds displaying trailing. However, the appearance of ~punctates^ on sprjie qf
these ? sherds i evidence^ag|inlt "tffis association] i

; Tr|e ocfcupati|)n perfod^ofkhejsite by elirlier non-Mississlppian peoples is dif-
fidultito dptermike.|A spedulajive; date of ca. AiD. 00 can ^e assigned to atjleasV
somesof-tfce Middle] Wc|od}arrd ceramics, A)9^^^1^lQfDlS^^Dr son.!
ofithi material i| also suggested by jhe 'prlsenpe qf obsidian flakes kt the site, a
st6ne;type probably iniro^ucpd into- the Middje A^est at this time in association
with ^HopeweU 5 (Grjffin, L96^:ll9)J THe date^of the termination cjf Woodland
occupancy cahnqt t>e dete.'miriecffrqm 6eramics alone.; However, the early |Late

s I * 5 '* ' '' ' . -.- -^- -. .-:>**--- * ---*- *

WoodkndjEffigy. Mounjt Quitliieiprqbably ierrriinated"at varying periods in dift ,
ferent regions'of Wiscorfcin!beiwe|n A.D; 1000-1200,

!Radipcairb6n (Jateis ojf X.D.' 14^0 t --7p ^ AJX 1630 +:60 were! determine^
for the Orr Phas6 materials. -jkildred We^er(1^59:i21 j wa? the; firs/t to suggest
that the Orr component at the Midway sije is earlier Un time than the Upper
lo^a -River sltesj wkic|i^ cbntain hisitoric^ tfiaterials". Dp 'tfte basis of inte f-site
cerpmc dejsign ejements, Robert Hall suggested that the Midway Component
should havb a^ da^e of ca. iA.D. 150Q (1960:17^-175). Supporting evidence for 1
thd arlproxjmate ^sixteenth; century rilacement '<xf the 0rr phase component a4
the Midway site is found in two site repprts published fn trie ILLINOIS ARCH-
: f <r rtie Anker Sitl" mi TTh? Oalc Forest Site." ifhd

AEOLOGICAL SURVEY: f <r ie Anker Sit" mi
suggestion is pfofTered in the 1'attejr report that a. greater percentage of plain yims?
(including the lip^, -a^i irjcreasad use of narrbw linei 4 a gner4l simpHfication
in design elements ar|pears in the ceramic traditi related to thi Oil Phase, anrf
to fhe "Bluej Island cultule ds the historic period is j r yrQached (%ihjm am^pen-
ner, -1961: 161): These suggestions are Based upon a comparison of the ceramics
recovered from the Anker site, which has an estimated age determination of A.
D. 1400-1500 (Bluhm and Liss, 1961:135), and the Oak Forest site, which has
an estimated date of ca. A.D. 1680 (Bluhm and Fenner, 1961:161).

Chi-square tests were conducted in shell tempered ceramic classes containing
relatively large sample sizes. The null hypotheses that the samples composing

The Midway Village Site

Classes I (x 2 = 0.71; df = 1), II (x 2 = 4.76; df = 3), V (x z = 1.04; df=4), VII
(x 2? = 5.41; df = 2), IX{x 2 = (X021; df = 1) are independent of the .01 level of
significance ; were hot accepted. These tests suggest, that there are no real differ
ences between t the jDroportions being compared in each of the above classes. The
null hypothesis was accepted for Class VI. All of the above classes except IX
have been rearranged (V) or contain one or more expected frequencies, less than
5 (I, II, Vj, VII). Chi-square computations, for Qass VI are recorded in Appendix
B. the most '.marfe'dly divergent classes (rim profile, shoulder design element, ap- , (
pendages, handle design elements) are also those classes containing very small
samples. ,-.-. ; ...

^ comparison of the range and frequency of lip, rim, and body sherd thick-
ness (Table 2) (Joes not demonstrate any marked disconformities between the
two areas. ^e samples are homogeneous" in paste, surface finish, and general^
vessel form., A study of in.iividual attributes and a limited number of Chi-square
tests suggest^ therefore, an in|:erpretatiQn of general homogeneity between the ...
two ceramic assemblages. Pottery 'type rcUstribuitiqns and the presence : pf pieces.-;
of particular jars^in Ibptn areas ^cprrpboraite a .suggestion. of essential .homo-
geneity. However, the fragmentary nature, of the re.rnains could mask differences
existing in any proportions ^epencfcnt ..cm whoje or nearly complete vessels.
What can be stated with some assurance is that the two ceramic samples demon-
strate a greater degree of similarity . betweep themselves- than either, does with
any other component described in the archaeok ^al literature.

Scraper groupings exhibit a general uniformity in size and shape betwe n the
two areas excavated. A possibility that must be considered is that^sqrrje scrapqrA ,.q
and other stone tools belong to earlier Woodland components. Evidence supppr-
ting an interpretation of general similarity between the two scraper samples is,. ,
provide'd'by two Chi-squiare testsl End scraper,s were' separated into: 1) speci-
mens possessing proximal ends rriore than half of their maximum width ; 2) speci-
mens with of without pronounced bulbs o'f percussion. The null hypotheses that , ,
the proportions of scrapers in each class 1 are dissimilar between Areas I and U
were not accepted at the' .01 level of significance (x 2 =1.05 arid 0.38 at df = !,)

The rerrtaining chipped stone tools and the ground stone tools appear in such -
small quantities that their omission in one or the other area may be the result
of sampling' error. However, none of the tools d'ffer from the expected lithic
content- of an Upper Mississippi, Orr Phase component. The omission of a
number of expected implements should be me tione^. Flat stone mortars, hand
mullers, stone celts, disc-shaped pipe bowls, * nt and biconical 'clubheads', are .
some of the artifacts previously recovered trom the site. Whether these imple- ,
ments were found with burials or in trash pits is not recorded and, therefore,,
the significance of their omission is difficult to assess.

Although the sample size of mbst of 'the stone tools is too small to serve as a
convincing indicator of the homogeneity or heterogeneity of the two areas ex- , -
cavated, a study of the horizontal and vertical distribution of 'utilized and^un-

1 52 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

modified flakes is persuasive corroborating evidence of an essential intra-site
homogeneity (Tables 39-40). Table 35 lists almost exact percentages of unmodi-
fied flakes of chert, red, yellow, and white quartzite, and quartz for the two
areas. The possibility of this occurrence between dissimilar archaeological cul-
tures would seem highly improbable. A study of Tables 39 and 40 also indicates
that the vertical distribution of chert and quartzite proportions in each area re-
main essentially the same, with a general numerical decrease in each by descend-
ing levels. The use of quartzite in each area as a minority material in the compo-
sition of scrapers, knives, drills, and projectile points is also suggestive of the
cultural homogeneity between the area's excavated.

Chi-square tests on two different samples of waste flakes produced dif-
ferent results. Waste flakes were separated in each area into: 1) chert, quart-
zite, and 'other'; 2) red, yellow, white and 'other' quartzite. The null hypo-
thesis that samples from the two areas within the classifications are independent
was accepted in the second case and rejected in the first (x^ = 3.31; df = 2).
This can be interpreted as conflicting evidence concerning the homogeneity of
the waste flake samples from each area. A glance at the calculations involved in
the determination of the former Chi-square value suggests, however, that the
'other' category contributes most numerically to the Chi-square value, while
there is only a slight difference between the remaining categories. Chi-square
computations for this test are recorded in Appendix B.

Economic Patterns

The occupational refuse from the Oneota component at the Midway site re-
flects an economic pattern characterized by hunting, fishing, gathering, and gar-
dening. Domestic dogs may have been eaten. Abundant evidence of hunting is
found in the large numbers of scrapers and in the projectile points. Robert Hall
has hypothesized with the aid of an "end scraper index" that "the proportion of
end scrapers relative to triangular points increases toward recent times and to-
wards the Plains area" among components of the Oneota Aspect (1962:121-
122). The "end scraper index," which is determined by multiplying the number
of end scrapers divided by the number of triangular projectile points by one
hundred, is 417.8 for the combined Midway site samples. Although this index
seems disproportionately high in relation to some other Upper Mississippi sites
(102.3 and 47.8 are recorded for the Lane and OReagan sites respectively), it
does emphasize the abundance of end scrapers at the site as compared to Aztalan
and to the 'transitional' Carcajou Point site. Only five end scrapers are recorded
for Carcajou Point and two for Aztalan (IBID). This has been interpreted as
evidence supporting an hypothesis of greater dependency upon hunting in the
'Classic' Oneota economic pattern as compared to the patterns apparent in a
Middle Mississippi site (e.g. Aztalan) and in an early 'transitional' site (e.g. Car-
cajou Point).

The Midway Village Site 1 53

The percentage of deer bone compared to ther mammal bone is exactly 88
per cent of the total at Cahokia, Aztalan and Carcajou Point, and 65.2 per cent
at Lasley's Point, a Lake Winnebago Phase Oneota site (Cleland, 1965:3).
Charles Cleland has interpreted the above percentages and the percentage of 36.2
per cent at the historic period Bell site (a non-Mississippian site) as indicative of
an increasing emphasis upon smaller animals such as raccoons, muskrats, and
beavers ca. A.D. 1 200 and later (IBID). Since the quantity of faunal remains
found at the Midway site is very small, it would probably be misleading to re-
late the percentages of animal species to the above hypothesis for comparative
purposes.

The appearance of corn and beans in both areas of the site excavated is ev-
dence for gardening, as well as additional evidence for the close similarity be-
tween the material content of the areas. Beans, which were available perhaps by
A.D. 1000 in the Middle West, have been reported from the related Upper Iowa
River sites and the Utz site (Yarnell, 1964: 110, 118-119; Wedel, 1959:35).
Yarnell (1965) suggests that the shortage of nut remains from the Midway site
may be explained by the presence of productive corn, bean, and, possibly,
squash cultivation.

Length and pattern of occupancy of the Oneota component at the site is
difficult to determine because of the small and differentially preserved sample.
The Whistling Swan is a transient visitant of the area, primarily from March to
April and from October to November (Gromme, 1963:13), but the single bone
of the species is part of a tool and, therefore, possibly not indicative of spring
or fall occupancy. The Wood Duck is a common summer visitant and the com-
mon Mallard is an abundant visitant during both summer and winter (IBID.:
14, 18). The value of the bird species as indicators of periods of site occupancy
is diminished by their limited numbers. Each species is represented by a single
specimen, which could be a stray individual captured at any time during the
year.

Mature plum fruits are available from August to October, acorns from Sep-
tember to November, and hickory nuts in October (Yarnell, 1964:62, 68-70).
All of these food plants were commonly stored, so their presence may repre-
sent more than just a fall occupancy (IBID.: 75). This is also true for the culti-
vated plant foods, but the presence of the scapula hoe indicates that the site
was probably occupied during the growing season too. Although many of the
species of fish found in the refuse are considered game fish and difficult to cap-
ture (Harlan and Speaker, 1956:49, 51, 69), they could have been captured at
any season of the year and, therefore, do not provide evidence of seasonal oc-
cupation.

The economic pattern of the Upper Iowa River complex of the Orr Phase
sites conforms closely to the pattern depicted for the Midway site. Evidence of
gardening is found in the presence of scapulae hoes, beans, and corn (Wedel,
1959:35). Large amounts of mammal, fish, and turtle bone, and of mollusk

154 WISCONSIN ARCHEOLOGIST Vol. 51, No. 3

shell, indicate that hunting and gathering were: important aspects of the
economic pattern (IBID.:35-36). Most of the larger mammals mentioned are
also forest or forest-edge animals. However, numbers of bison bones at the Lane
village site on the Upper Iowa River are suggestive of a greater orientation to-
wards the prairie areas to the west and, possibly, of an increasing Plains in-
fluence in the later Upper Iowa River sites. Early historic records also affirm the
Plains area influence that characterized the historic loway (Mott, 1938:241).

The concept of types of community patterning is an attempt to construct
integrated models of socio-economic patterns as structural systems, in contrast
to a description of an agglomeration of concepts of its constituent parts. Size of
population and community pattern are difficult to determine from' the features
and refuse material at the Midway site. If the Oneota component does represent
a short term occupation by a single community, it is probably representative of
the primary type- of community patterning termed 'Semi-Permanent Sedentary'
by Beardsley et al .'(1956:129-157).

External Relationships *

Radiocarbon determinations for4-he TvTidway site indicate occupation, by
Oneota peoples approximately 500-700 years- after the' first emergence of thi
life- way (Barreis arid Brysori, 1965). Clues as to the origin of the Oneota life-
way caniiot be expected from -the Midway site, filsewhere^the writer has sug-
gested, that the 'Oneota Cultures are most likely Woodland peoples acculturated
to a Mississippian pattern Involving significant agriculture, new ceramic styles,
and other socio-cuitural changes (Gibbon, 1969). In this ''acculturation' hy :
pothesis early Oneota Phases' (e.g. Koshkonong, (irarid River) are considered
'Lake-^Voods' ecotypes having a ; n intense focus upbn local riyerihe-lacustrinere :
sources (e.g. fish, rnollusks). A new type, f the 'Grassland,' probably developed
from J:he J northern Lake- Woods Oneota ecotypes during the subsequent drying
period which saw the expansion of the Prairie Peninsula. This new ecotype is
commonly called "Gassic" Oneota' and is associated \yith the expanding Prairie
Peninsula. i ,i "

The Orr Phase is here interpreted as a regional Cfaeoia 'Grassland ^^
apted to^'he northern limits of the Opper' Austral Bibtjc'Zone in a prairie: fpres^
setting along the Upper Mississippf River Valley. This prairie-forest province was
an area 1 of close'd 'deciduous forests with oak savannas and prairies (Curtis, 1 959).
The terraces were covered l iri general with sparse growths, of oak in a prairie set-
ting 'as fafr north as St. Croix; In y contrast, the Uplarips were heavily 7 timbered.

"Life Zone" studies 'attenipt' to recognize environmental and ecological pre-
ferences of plant and animal species. Although life zones transcend.the valley,
and, therefore 1 ,' do not differentiate Between the animal life, of the uplands an^l
the valley bottoms; 'theif^tudy is of interest in rehtion to the geographical dis-
tribution of the Orr Phase as a unit. C. Hart Merriani divided the tlnited States

f I t - - < - ' , > . : i ( - .- r : *.-"; i ' " ,

into> three great regions, each' characterized fey particular associations of animals

The Midway Village Site 1 55

and plants (1898). The three great regions Boreal, Austral, and Tropical, were
each subdivided into a number of minor belts or areas. A map depicting the ex-
tent of these zones, areas, and regions appears on the frontispiece of Merriam's
report.

The Orr Phase cluster of sites is included within Merriam's Upper Austral
Zone of an eastern humid or Carolian Area of the Austral Region. The northern
boundary of the Upper Austral Zone is delimited by the isotherm showing a
sum of normal positive temperature of 6,400 C, while its southern boundary
coincides closely with the isotherm of26 J C/or the six hottest weeks (IBID.:
55). More recent authors have attempted to refine the fringes of Merriam's life
zones in their own specific geographical areas of interest. Hartley Jackson
(1961), in Wisconsin, has retained the original terminology, but refined the geo-
graphical extent of the Upper Austral Zone, the Canadian Zone of the Boreal
Region, and the Transition Zone between them. All Orr Phase sites in Wisconsin,
Minnesota, and eastern Iowa are confined to the Upper Austral Zone, which
hugs the Mississippi River as far north as St. Croix County. This is slightly north
of the northernmost possible Orr phase site, Diamond Bluff. Other discussions
of life zones also indicate that the sites comprising the Orr Phase (with the pos-
sible exception of the eighteenth ce-itury northwest Iowa sites) are confined to
the northern fringes of a distinctive prairie-forest biotic province or life-zone
generally designated the Upper Austral Zone or Province (Dice, 1938; Scott,
1942; Roberts, 1945; Buss and Mattison, 1955

The age determinations of the Midway site indicate that the site is ea . 'ier in
time than the Upper Iowa River sites. The problem of the degree of relation^
between them is, therefore, one that must be considered within the context of
the implications of this temporal difference. Subsequently, two possibilities may
be presented to explain the similarities between the two areas: 1) while the in-
habitants of the Midway site were in some way related to the peoples who later
populated the 'Iowa Orr Phase,' they were not involved directly in the cultural
line that eventually moved into the Upper Iowa River; 2) the earlier culture rep-
resented at the Midway site eventually moved across the Mississippi River into
Upper Iowa River Valley and a definite genetic connection existed between the
two populations. If the latter alternative is correct, then the Midway and Upper
Iowa River ceramics and other material items are representative of different time
periods on a spectrum stretching through a limited time span.

Admittedly the Upper Iowa River Valley contains a multitude of sites from
which the greatest amount of Orr Phase material has been recovered. But to des-
ignate it as the cultural 'heartland' of the Orr Phase on this basis alone may be a
distortion of the actual circumstances. A survey of northwestern Illinois, south-
western Wisconsin, southeastern Minnesota, and northeastern Iowa would most
likely demonstrate that the Orr Phase is a more widely spread phenomenon, ex-
hibiting a greater range of variation, than is currently recorded (e.g. Me Kern,
1945:126). The heartland of the Orr Phase might more accurately be placed in

156 WISCONSIN ARCHEOLOGIST Vol. 51, No. 3

the general Upper Mississippi Valley south of St. Croix, with an earlier concen-
tration in the tributaries to the east of the river and a later concentration in the
western tributaries. Certain discrepancies between the two areas in ceramic mo-
tifs or projectile point measurements (e.g. Wedel, 1959:1 10), might be explained
by this temporal difference and the predominance of quartzite on the east side
of the river.

Keyes, who excavated many of the sites along the Upper Iowa River, was of
the opinion that the sites near the mouth of the Upper Iowa River were older
than the sites further up the stream (reported in Wedel, 1959:38). A westward
movement across the Mississippi River is suggested by tradition (Mott, 1938:
241) and by the material content of the sites. The presence of historic trade
goods among the Orr Phase sites along the Root River and its tributaries in Min-
nesota, and the similarity of the pottery from these sites with the proto-historic
Upper Iowa River sites, supports this interpretation. If the inhabitants on the
eastern side of the Mississippi River were forced across the river by pressure ex-
erted by disrupted tribes moving through Wisconsin or Illinois, or for some as
yet unknown reason, a greater concentration of sites would naturally occur as
these people were funneled into the fewer and smaller western tributaries.

CONCLUSIONS

The Midway site is an eight-and-one half to eleven acre multi-component site
situated on a terrace in the Gorge of the Upper Mississippi River. The gorge is
within the Oak Subclimax of the North Central United States and the northern
fringe of the Upper Austral Life Zone or Biotic Province. An Oneota Aspect
component is dominant at the site, with at least one Middle Woodland and one
early Late Woodland component present. The components are not stratified ex-
cept in the burial ground.

The Oneota Aspect component is part of an early concentration of Orr Phase
sites on the eastern side ^f the Mississippi River. Carbon 14 dating has resulted
in age determinations 01 ^.D. 1420 + 70 (WIS-61) and A.D. 1630 + 60 (WIS-
79) for the component. Testing of two widely spaced areas on the site suggest
that the component was the result of occupation by the same archaeological cul-
ture over a relatively brief period of time in the sixteenth century.

The occupational refuse from the Oneota component suggests an economic
pattern characterized by hunting, fishing, gathering, gardening, and, possibly, by
the inclusion of dogs in the diet. Food remains were too scanty to reconstruct
the relative emphasis upon these differenct activities. If the component does rep-
resent a short term occupation by a single community, it is probably representa-
tive of the primary type of community patterning termed "Semi-Permanent Sed-
entary."

The Midway Village Site 1 57

BIBLIOGRAPHY

Barreis, David A. 1953. The Blackhawk Village Site (Da 5), Dane County, Wis-
consin. JOURNAL OF THE IOWA ARCHEOLOGICAL SOCIETY, Vol. 2,
No. 4, pp. 5-20. McGregor.

Baerreis, David A. and Reid A. Bryson. 1965. Climatic Episodes and the Dating
of the Mississippian Cultures. THE WISCONSIN ARCHEOLOGIST, Vol. 45,
No. 4, pp. 203-220.

Beardsley, Richard, et al. 1956, Functional and Evolutionary Implications of
Community Patterning In SEMINARS IN ARCHAEOLOGY: 1955. Edited
by Robert Wauchope. Mem. Soc. Amer. Archaeol., No. 1 1. Salt Lake City.

Bluhm, Elaine A. and Gloria J. Fenner. 1961. The Oak Forest Site. In "Chicago
Area Archaeology," edited by Elaine A. Bluhm, pp. 89-137. ILLINOIS
ARCHAEOLOGICAL SURVEY, BULLETIN NO. 3, Urbana.

: .. . i : : i

Buss, Irvin O. and Helmer Mattison. 1955. A Half Century of Change in Bird
Populations of the Lower Chippewa River, Wisconsin, MILWAUKEE PUBLIC
MUSEUM, PUBLICATIONS IN ORNITHOLOGY, No. 1.

Cleland, Charles E. 1965. Oneota Adaptations to the Environmental Resources
of Prehistoric Wisconsin. Paper read at the 1965 meeting of the Society for
American Archeology.

Curtis, John T. 1959. THE VEGETATION OF WISCONSIN. University of Wis-
consin Press, Madison.

Dice, L.R. 1938. The Canadian Biotic Province with Special Reference to the
Mammals. ECOLOGY, Vol. 19, No. 4, pp. 503-514.

Gibbon, Guy. 1966. The Midway Village Site: an intra-site analysis. M.S. thesis.
University of Wisconsin. Madison.

Gibbon, Guy. 1969. The Walker-Hooper and Bornick Sites. Ph. D. thesis. Univer-
sity of Wisconsin, Madison.

Griffin, James B. 1943. THE FORT ANCIENT ASPECT. University of Michigan
Press.

Griffin, James B. 1952. Some Early and Middle Woodland Pottery Types in Il-
linois. In HOPEWELLJAN Communities in Illinois. Edited by Thome Deuel.

1 58 WISCONSIN ARCHEOLOGIST Vol. 5 1 , No. 3

ILLINOIS STATE MUSEUM SCIENTIFIC PAPERS, Vol. 5, pp. 95-129.

Griffin, James B. 1960. A Hypothesis for the Prehistory of the Winnebago, pp.
809-865. In CULTURE IN HISTORY, edited by Stanley Diamond. Columbia
University Press, N.Y.

Griffin, James B. 1961. Some Correlations of Climatic and Cultural Change in
Eastern North American Prehistory. ANNALS OF THE NEW YORK ACAr
DEMY OF SCIENCES, Vol. 95, Art. 1, pp. 710-717.

Griffin, James B. 1965. Hopewell and the Dark Black Glass. Papers in Honor of
Emerson F. Greenman, THE MICHIGAN ARCHAEOLOGIST, Vol. 11, Nos.
3-4, pp. 115-155.

Gromme, Owen J. 1963. BIRDS OF WISCONSIN. Madison, Published for the
Milwaukee Public Museum by the University of Wisconsin Press.

Hall, Robert L. 1960. The Carcajou Site (Fe2) and Oneota Development in Wis-
consin. Doctoral dissertation, University of Wisconsin, Madison.

Hall, Robert L. 1962. THE ARCHEOLOGY OF CARCAJOU POINT: WITH
AN INTERPRETATION OF THE DEVELOPMENT OF ONEOTA CULr
TURE IN WISCONSIN. 2 Vols. The University of Wisconsin Press ,
Madison.

Harlan, James R. and Everett Speaker. 1956. IOWA FISH AND FISHING. 3d
ed. Des Moines, State Conservation Commission.

Henning, Dale R. 1961. Oneota Ceramics in Iowa. JOURNAL 6t THE IOWA
ARCHEOLOGICAL SOCIETY, Vol. 11, No. 2.

Hurley, William M. 1965. Archeological Research in the Projected Kickapoo Re-
servoir, Vernpn County, Wisconsin. THE WISCONSIN ARCHEOLOGIST,
" V01.-46, Nb! I; i^'T-ll^

Jackson, Hartley H. 1961. MAMMALS OF WISCONSIN. Madison, University
of Wisconsin Press.

Me Kern, Will C. 1931. Wisconsin Pottery. AMERICAN ANTHROPOLOGIST,

Vol. 33. No. 3; pp. 383-390.

Me Kern, Will C. 1945. Preliminary Report on the Upper Mississippi Phase in
Wisconsin. BULLETIN OF THE PUBLIC MUSEUM OF THE CITY OF

The Midway Village Site 1 59

MILWAUKEE. Vol. 16, No. 3 pp. 109-285. Milwaukee.

Merriam, C. Hart. 1898. Life Zones and Crop Zones of the United States. U.S.
DEFT AGRIC., DIV. BIOL. SURV., Bull. 10.

Mott, Mildred. 1938. The Relation of Historic Indian Tribes to Archaeological
Manifestations in Iowa. THE IOWA JOURNAL OF HISTORY AND
POLITICS, Voi; 36, No. 3, pp. 227-314. Iowa City.

Roberts, T.S. 1945. The Vanished Mammals of Minnesota: a Retrospect. In.
"The Mammals of Minnesota," edited by Gustav Swanson, T. Surber, T.
S. Roberts. MINNESOTA DEPARTMENT OF CONSERVATION TECH-
NICAL BULLETIN NO. 2.

Scott, W. 1942. General Life Zones of Wisconsin Avifauna. In "Wisconsin
Birds," edited by N.R. Barger, Bussewitz, Loyster, Robbins, Scott. WIS-
CONSIN SOCIETY OF ORNITHOLOGY, Madison.

Wedel, Mildred. 1959. Oneota Sites on the Upper Iowa River. MISSOURI
ARCHAEOLOGIST, Vol. 21, Nos. 2-4.

Yarnell, Richard A. 1964. Aboriginal Relationships Between Culture and Plant
Life in the Upper Great Lakes Region. ANTHROPOLOGICAL PAPERS,
No. 23, Museum of Anthropology, University of Michigan.

Yarnell, Richard A. 1965. Archaeological Plant Food Remains From Wisconsin.
Unpublished Manuscript.

APPENDIX A: FEATURE DESCRIPTIONS
AREA I

FEATURE 1 : Part of roughly circular, basin-shaped pit in SW corner of unit
1; the observed portion measured 38" from the S wall and 20" from the W
wall; Feature 1 began 12" below the surface and had a depth of 18"; 10 fire
burned rocks ranging in size from 4.8x4.6x6.3 cm. to 10.8x8.6x3. 2 cm. were
scattered through a concentration of charcoal, sherds, and shell; not completely
excavated.

160 WISCONSIN ARCHEOLOGIST Vol. 51, No. 3

FEATURE 2: Part of roughly circular, basin-shaped pit in SE corner of unit 1;
observed portion 30" in diameter on N-S line and 20" from E wall; the pit began
immediately below plow zone (9") and had a depth of 13"; similar contents to
feature 1 with 5 fire burned rocks ranging in size from 33x5.6x7.7 cm. to
4.5x5.5x8.1 cm.; not completely excavated.

FEATURE 3 : Part of roughly elliptical, basin-shaped pit in NE corner of unit
1; observed portion 20" from north wall and 22" from E wall; pit began im-
mediately below the plow zone (8") and had a depth of 15"; concentration of
sherds, bone, and shell with 7 scattered fire burned rocks ranging in size from
2.1x3.7x4.0 cm. to 4.9x5.3x7.7 cm.; not completely excavated.

FEATURE 4: Irregular basin-shaped pit in center of unit 1; this pit was 34"
long, 34" wide, and 19" in depth; the pit began 11" below the surface; con-
centration of charcoal, sherds, shell and 13 fire burned rocks ranging in size from
4.1x4.2x6.5 cm. to 4.2x5.0x7.9 cm.; a small concentration of shells in the
center of the pit was apparently discarded together as refuse, for they were
stacked back to back; the pit splayed irregularly at its base, measuring 10" N-S
and 52" E-W; completely excavated.

FEATURE 5: Roughly elliptical, basin-shaped pit in N-W corner of unit 1;
pit orifice measured 48" E-W and 28" N-S; the pit began 32" below the surface
and had a depth of 8"; feature 5 contained a concentration of sherds, shell, bone
and 27 fire burned rocks ranging in size from 2.4x4.3x6.6 cm. to 5.7x13.8x14.3
cm.; 19 of these rocks were clustered on the E side of the pit and 8 on the W;
completely excavated.

FEATURE 6: An approximately circular, flat lenticular pit 2" thick in unit 3;
the observed portion measure 14" from the N wall and 16" from the W wall; the
pit began 14" below the surface and contained 1 fire burned rock measuring
4.3x4.7x8.8 cm.; like the other features so far described, this pit was probably a
refuse concentration; not completely excavated.

FEATURE 7: A pit approximately circular in plan and flat lenticular in cross
section in the SW corner of unit 5; the pit measure 5" in depth, 34" from the W
wall, and 17" from the S wall; feature 7 began 17" below the surface and con-
tained a concentration of charcoal, sherds, and shell; not completely excavated.

FEATURE 8: A pit approximately circular in plan and flat lenticular in cross
section 19" below the surface in the SE corner of unit 5; the pit measured 22"
from the E wall and 17" from the S wall; it had a depth of 5"; similar contents
as other features, but like feature 7 no fire burned rocks; not completely exca-
vated.

The Midway Village Site 1 6 1

FEATURE 9: Concentration of mollusk shells 11" below the surface in center
of unit 5; the feature was roughly elliptical in outline, 7" i n depth, and ovoid in
cross section; it measured 36" E-W and 18" N-S; only a few flecks of charcoal
and a few chert flakes were in association with it; completely excavated.

FEATURE 10: Comparatively large feature overlapping portions of units 6, 8
and 9; the large black area defining this feature began immediately below the
surface and measured 44" in length (E-W), 48" in width (N-S), and 13" in depth;
26 scattered fire burned rocks ranging in size from 1.9x2.0x3.3 cm. to 6.9x13.
5x15.4 cm. were found; this feature was completely excavated, but not very
productive in refuse material.

AREA II

FEATURE 11: A roughly elliptical pit with a fairly flat bottom was found
32" below the surface in the center of unit 14; a concentration of charcoal,
sherds, seeds, and shell defined an area 26" N-S, 19" E-W, and 13" in depth; the
pit contained 14 scattered fire burned rocks ranging in size from 2.0x3.1x4.0
cm. to 5.1x5.3x7.1 cm.; completely excavated.

FEATURE 12: unit 15; intrusive historic pig burial; no artifacts.

FEATURE 13: Badly eroded fragments of cranium, vertebrae, and one upper
arm in unit 15; extremely eroded; disturbed by feature 12; no cultural material
in association.

FEATURE 14: A pit roughly elliptical in outline and basin-shaped in vertical
cross section found 14" below the surface in the north-central area of unit 19;
the feature measured 18" E-W, 1 1 " N-S, and 9" in depth; concentration of char-
coal , sherds, and shell; 5 fire burned rocks ranged in size from 1.2x2.6x3.9 cm.
to 2.4x3.3x6.1 cm.; completely excavated.

FEATURE 15: A pit approximately circular in outline and basin-shaped in ver-
tical cross section 18" below the surface in the SE corner of unit 19; sherds,
shell, and bone were concentrated in a blackened area 13" N-S, 12" E-W, and
13" in depth; no fire burned rocks; completely excavated.

162 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 3

APPENDIX B: CHI-SQUARE COMPUTATIONS

I. Chi-square test for the significance of the distribution of varieties
of inner rim, upper edge design techniques (Class VI) between Areas
I and II at the Midway site.

1
ob/ex

technique

2 3

ob/ex ob/ex

4
ob/ex

Chi-square test for significance of the distribution of colors of
quartzite (1. red, 2. yellow, 3. white, 4. 'other') between Areas
I and II at the Midway site.

color

red yellow white 'other'
ob/ex ob/ex ob/ex ob/ex

COMMITTEE ASSIGNMENTS
WISCONSIN ARCHEOLOGICAL SOCIETY

MEMBERSHIP COMMITTEE: Phillip H. Wiegand, Chairman.
Herman Zander, J. K. Whaley.

PUBLICITY: Tom Jackland.

FRAUDULENT ARTIFACTS: J. K. Whaley, Chairman. Martin
Greenwald, Robert Hruska, Neil Ostberg, Dr. Robert Ritz-
enthaler, Paul Scholz, Phillip H. Wiegand.

SURVEYING AND CODIFICATION: Paul Koeppler, Chairman.
Darryl Beland, Elmer Daalmann, Wayne Hazlett, Ernest
Schug.

PRESERVATION OF SITES: Robert Hruska, Chairman. Dr.
Richard Peske, Dr. Robert Ritzenthaler. (Chairman has op-
tion to make further appointments.)

EDITORIAL: Dr. Robert Ritzenthaler, Chairman. Dr. David A.
Baerreis, Dr. Joan Freeman, Dr. Lee Parsons, Dr. Melvin
Fowler.

PROGRAM: Dr. Diehard Peske, Chairman. Paul Turney, Mrs.
P. H. Wiegand, Dr. Robert Ritzenthaler.

LAPHAM RESEARCH MEDAL. Dr. Robert Ritzenthaler,
Chairman, Dr. David A. Baerreis, Phillip H. Wiegand.

THE CHARLES E. BROWN CHAPTER
Madison

(Meets second Tuesday, 7:45 P. M., State Historical Society,
October thru May)

President: James Stoltman
Secretary: Peter Storck
Treasurer: Lathel Duffield

THE FOX VALLEY CHAPTER

Oshkosh

(Meets second Monday, 7:30 P. M., Oshkosh Public Museum,
September thru May)

President: Elmer Daalmann

Vice President: Richard Mason

Secretary: Robert R. Jones

Treasurer: Nancy L. Hoist

THE WISCONSIN
ARCHEOLGOIfT

KANSAS CITY, MCX
PUBLIC LIBRARY

MAR1 1971

SEED ANALYSIS IN ARCHEOLOGY
by Janet Doris Spector
I

163

THE ONEOTA COMPONENT AT THE PORTE DES
MORTS SITE, DOOR COUNTY, WISCONSIN 191

by Carol Mason
University of Wisconsin-Green Bay

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee, Wisconsin

Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. M., Milwaukee Public
Museum, September thru May

OFFICERS
PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Ostberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Greenwald.

TREASURER

Wayne J. Hazlett
376^ S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct, Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr. Robert E. Ritzenthaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL,

MEMBERSHIP FEES

The Wisconsin Archeologist is distributed to members

as part of their dues.

Annual Members, $5.50

All communications in regard to the Wisconsin Archeological Society
and contributions to the Wiscorsin Archeologist should be addressed to
Mrs. Edwai Flaherty, Secretary, N62-W15127 Tepee Ct., Menomonee
Falls, Wis. 53Ur>l. Entered as Second Class Matter at the Post Office at
Lake Mills, Wis., under the Act of August 21, 1912. Office of Publica-
tion, 316 N. Main St., Lake Mills Wis. 53551.

THE WISCONSIN ARCHEOLOGIST

New Series
MILWAUKEE, WISCONSIN - DECEMBER, 1970

Published Quarterly by The Wisconsin Archeological Society

SEED ANALYSIS IN ARCHEOLOGY
Janet Doris Spector

PREFACE

The seeds reported on in this study were recovered through water flotation
and screening of soil samples collected during excavation of the Harvey site,
Waukesha County, Wisconsin. Excavation of this site was conducted during the
summer of 1968, under the auspices of the Wisconsin State Historical Society's
highway salvage program. During the eight week field season, a crew of 8 stu-
dents, under the direction of the author excavated an area of over 1,175 square
feet to an average depth of 1.5' below the surface within the highway right-of-
way established for the relocation of State Highway 15. : ,

Analysis of the material remains recovered at the Harvey site was handicapped
in that the site yielded only a small number of diagnostic artifacts, most of
which were disturbed from their original provenience by modern plowing. Typo-
logical comparison of the ceramics and projectile points found with known local
types indicated that the site was occupied by at least two different groups dur-
ing the Woodland period. A more provocative complication in analysis of the
Harvey site arose in the attempt to interpret the nature of the occupations at the
site. The artifact inventory which was extremely small and typologically mixed,
suggested that the site was never inhabited for prolonged periods of time. In
contrast to this evidence however, was the presence at the site of 15 features
which indicate rather intensive occupation. These features varied in outline and
profile from small, shallow, regularly shaped basins, to rather large, deep ellip-
soid pits. Functional interpretation on the basis of feature content was difficult.
With the exception of three features (1 fire pit, 2 storage or refuse pits), the pits
contained little other than charcoal flecks and occasionally scattered lithic debi-
tage (see site report, Spector, 1970).

The most reasonable hypothesis concerning the site occupation in view of the
somewhat conflicting evidence, is that the site was the location of some special-
ized activity which did not necessitate the use or manufacture of many lithic or
ceramic implements but which did require features for either the storage or pro-
cessing of some kind of resource. These resources or materials were either sub-
sequently removed from the pits; too minute to be noticed during excavation, or
perhaps, perishable. To carry this interpretation further, it might be suggested

164 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

that the site was visited, periodically, during the Woodlan'd period by small
groups exploiting locally available plant resources. Ethnographic accounts of
historic Indians in the Midwest indicate that plants such as wild rice could be
harvested and prepared without the use of lithic or ceramic artifacts. The thresh-
ing of wild rice, however, often required the construction and use of subsurface
pits (Jenks, 1902). The Harvey site, situated along the bank of the Fox River
and adjacent to a marsh would have been an ideal spot for the collection of
vegetation. This hypothesis about the site, however, could be supported only if
plant remains, such as seeds (which are too small to be observed during exca-
vation) were present in the features. In fact, seeds were the only remains found
in the soil samples taken from features at the site. Interpretation of these re-
mains became the next problem faced in analysis of the site.

SEED ANALYSIS

Water flotation, a relatively new technique in archeology, has made possible
the recovery of small floral and faunal remains which formerly went un-noticed
during the routine excavation of sites. Vegetal remains, the only type present in
the soil samples from the Harvey site, can potentially provide two kinds of in-
formation about a site: Ethnobotanical data relating to the economy, subsis-
tence, and diet of prehistoric groups; and environmental information about the
habitat in which the groups lived. If a site is stratified or if several sites in one
area are being investigated, seeds can theoretically add a new dimension to our
study and understanding of cultural and environmental change. In order to ex-
amine the specific nature of the information provided by vegetal remains re-
covered archeologically, seeds from a number of sites must be subjected to sys-
tematic analysis. The following represents one attempt to test the potential and
limitations of seed analysis in archeology. The primary purpose of this study is
to investigate the approach as it applies to wild plant remains found in the Mid-
west. Many other such studies must be undertaken before the value of this ap-
proach can be judged. Since the published material on the subject is so limited,
the following report includes a summary of the techniques used to date for the
recovery of seeds; a survey of several representative Midwestern studies in which
seed analysis has been attempted; and a general discussion of the principles and
problems involved in using the approach for cultural and environmental infer-
ence.

Until very recently the prevalent attitude among archeologists has been that
plant remains, because of their perishable nature, are not usually present in arch-
eological contexts (Meighan 1958:15). Those plant remains which have occa-
sionally been recovered in the routine excavation of features, caves, etc. have
generally been identified, tabulated and then presented with little interpretive
comment in the appendices of site reports. In the Midwest, the paucity of vegetal

Seed Analysis In Archeology 165

remains recovered archeologically has been clearly shown by Richard Yarnell
(1964). Yarnell surveyed midwestern site reports in search of records of wild
plant remains from Michigan, Wisconsin, Ontario, Ohio, northeast Illinois, nor-
thern Indiana, northwest Pennsylvania, western New York, and the northeastern
edge of Minnesota. In a total of 59 sites, only 168 occurrences were recorded.
This included various types of fruits, berries, nuts and tubers, with nuts, the
most conspicuous plant remains, accounting for the majority of the occur-
rences. Yarnell 's more recent summary of Wisconsin plant remains included in
site reports adds data from 3 more sites (1966). Even with these additions, it is
clear that the seeds reported do not reflect the vegetal remains which actually
existed at the sites but rather the techniques used for the recovery of such ma-
terial.

Elso Barghoorn, a botanist, published an article in AMERICAN ANTIQUITY
in 1943, urging archeologists to become aware of the potentials of recovering
seeds during excavations. He argued that plant remains can reveal- the former
vegetation of an area and help in understanding previously existing climates.
However, he noted that ". . .in order that the greatest value may be obtained
from a study of botanical materials. . .it is essential that certain precautions and
techniques be used in collecting and preserving them. Unfortunately, much bo-
tanical information available in an archeological site is either discarded or over-
looked in the process of digging and exposing the site" (Barghoorn 1943:289).
It was not until the present decade, that such techniques were developed by
archeologists, when it finally became evident that the assumption that plant re-
mains are not often present in archeological contexts is a false one. Instead, the
fact is, that because of their small size and fragile nature, plant remains, es-
pecially seeds, cannot be recovered by conventional field methods (Struever
1968:353).

In surveying the literature on this subject, two different techniques for seed
recovery have been encountered. The most successful has been Stuart Struever's
water flotation. The "water-separation" technique works on the principle that
different substances have different porosities and therefore settle in water at
different rates (Struever 1968). Thus, plant remains and small bones, because of
their slow setting rates can be separated easily from heavier materials when soil
samples are placed in water and the floating materials extracted. To further
separate the plant remains from small bones, Struever recommends a chemical
flotation process using a zinc chloride solution with a specific gravity of 1 .62, in
which the heavier bone will sink, while vegetal materials float. Watts and Winters
in their analysis of plant "macrofossils" from a stratified lake bed in Minnesota
used a similar water-separation method. Soil cores were taken from the site using
a 3.8 cm. diameter Livingstone sampler. Core segments, 10 cm. long were
individually scraped with a spatula to remove contamination and then stirred in
water, where most samples broke up satisfactorily. They found that soaking in a
hot solution of detergent in water helped to disaggregate cohesive sediments.

166 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

The disaggregated sediment is next washed with water jetted through no. 40 and
no. 140-mesh screens. Residues from both screens are then set in shallow water
on a white plate and examined with a low powered binocular microscope (Watts
& Winters 1966:1341-1343).

A second technique for recovering seeds is that used by Richard Yarnell in
his study of materials from the Feeheley and Juntunen sites in Michigan (Yar-
nell 1964:34). This technique involves the dry-screening of soil samples. Al-
though Yarnell 's discussion of procedures used is cursory, he does report that
". . .most of the larger plant remains were found by low power microscopic ex-
amination of samples of charcoal remaining after gross sifting in the field. Most
of the smaller seeds were located by low-power microscopic examination of the
mixed detritus that remained after fine sifting of cultural deposits in the labora-
tory" (Ibid). The dry screening technique has several limitations and is less sat-
isfactory than water flotation. It is a time consuming operation and does not
lend itself well to the processing of large samples; the mechanical abrasion can
easily damage seeds; it separates material only by size, not by class; and finally,
it cannot be used if the soil is at all cohesive since seeds will remain lodged in
the matrix.

Four studies have been selected as representative of the ways in which seeds
have been used to date for interpretive purposes in mid western archeology. The
most comprehensive use of vegetal remains encountered in the archeological lit-
erature was made by Richard Yarnell in his 1964 study of the "Aboriginal Re-
lationships between Culture and Plant Life in the Upper Great Lakes Region."
He concentrates on three major topics: The utilization of native plant products
as determined from ethnographic records and the results of archeological re-
search; aboriginal agriculture with respect to its historical aspects and the dis-
tribution of length of frostless season; and the effects of aboriginal activities on
the natural flora and vegetation of the region (Yarnell 1964:iii). Perhaps the ma-
jor contribution made by Yarnell in this study is his synthesis of known infor-
mation about the plants utilized by historic Indians in the Great Lake region.
His tables providing data on the use of plants, their season of availability and
habitat preference and geographic range can be used as a basic reference for arch-
eological interpretation of plant remains found at sites in the region. Archeolo-
gists, who have usually concluded that seeds recovered at sites represent the food
resources of aboriginal inhabitants, will find Yarnell's work illuminating. He re-
cords that the historic Indians used plants in a number of ways other than food
and that, depending on the specific purpose, various parts of the plant may have
been utilized. This information should be considered before offering what may
be rather naive conclusions about archeological plant remains, especially seeds.
Most relevant to the present report is YarnelTs analysis of seeds found after
dry-screening cultural deposits at the Feeheley and Juntunen ;sites in Michigan.
On the basis of his identifications, Yarnell suggests both ethnobotanical and en-
vironmental interpretations about the sites. Typical of the kind of conclusions

Seed Analysis In Archeology 167

he draws is the following: "The season of availability of hickory nut, walnut, and
butternut is October. For acorn, September-November, and for grape, August
to autumn. Thus, it appears that the occupation or occupations of the site rep-
resented by the plant food resources took place in early autumn, if not other
seasons" (Ibid.: 26). He also makes use of seeds to suggest that one of the sites
was well cleared at the time of occupation (Ibid.: 40). At both sites, Yarnell
employs ethnographic analogy to infer that various wild species found were used,
prehistorically, for food. This method is useful only if applied with caution, as
Yarnell is careful to point out (I bid.: 47). Most ethnographies in the Great Lakes
region are based on research conducted in the 20th century. By that time, the
Indians had been under the influence of whites for a considerable number of
years. Many plants used for food, medicine, etc. by 20th century Indians may
have been introduced by whites. Likewise, other plant resources used aborig-
inally, had been replaced by the products of wlu'te traders. Generally, Yarnell is
conservative in drawing conclusions about the plant remains found. It is note-
worthy that he always mentions the condition of the seeds, the material with
which they were associated and the certainty of his identifications. This makes
it possible to evaluate his conclusions. The major criticism of the seed analysis
done by Yarnell is that he fails to discuss sampling procedures. No information
is given concerning sample size, or how areas were selected to be sample. Thus,
we have no way of judging whether the seeds found are representative of the na-
tive plant communities present at the sites or only those species culturally sel-
ected by the inhabitants.

Stuart Struever (1962) and Olaf Prufer (1965) have both used vegetal re-
mains from Hopewellian sites as the basis of their hypotheses regarding the con-
troversial subject of Hopewell subsistence. Both suggest on the basis of plant
remains recovered, that the people were horticulturalists. After analysis of seeds
from an Illinois Hopewell site, Struever postulates, as others had earlier, that
there might have been a center of plant domestication in the eastern U.S. where
local plant forms were cultivated prior to the diffusion into the area of the
maize-bean-squash complex (Struever 1962:584-587). He uses as evidence the
large number of Chenopodium seeds found at the Synders site and the context
in which these were found. Seeds identified by Hugh Cutler as CHENOPODIUM
SP. were found only in refuse pits (though we are not told what other areas
were sampled), and were always associated with other plant foods such as hic-
kory nuts, acorns and grasses. Struever believes this tends to support the hypo-
thesis of Volney Jones, of an eastern agricultural complex of local plants ". . .in-
cluding goosefoot or lamb's-quarter (CHENOPODIUM SP.), pigweed (AMAR-
ANTHUS SP.), giant ragweed (AMBROSIA TRIFIDA L.), sunflower (HELIAN-
THUS* ANNUS L.), and marsh elder (IVA SP.) (Ibid: 584). Struever is cautious
in this support of the theory. Before such a suggestion can be accepted, he feels
that many more plant remains must be recovered and methods designed whereby
domesticated plants can be clearly distinguished from wild forms. It might also

168 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

be added that only if seeds can be determined to the species level can such an
hypothesis be tested. Many variants of the genus Chenopodium, for example,
were introduced to this country from Europe. Such species found archeologi-
cally, would indicate contamination, even if the seeds were found in features.
This problem will be further discussed later in this report.

Olaf Prufer, in his analysis of the McGraw site in Ohio, is not as cautious in
his conclusions as is Struever. Prufer suggests that this site was a Hopewellian
farmstead and ". . .that the most important economic activity of the inhabitants
seems to have been farming (Prufer 1965:136). This conclusion is in part based
on the positive identification of one medium-sized, 12 row ear of corn and one
distorted kerneL In spite of the fact that the wild plant remains recovered were
more numerous than domesticates (44 pieces of walnut, 1 26 of hickory, 3 pieces
of acorn, 37 fragments of hackberry, 2 fragments of plum), Prufer goes on to
state the "collecting of wild plants was of least importance" in the economy of
the Indians (Ibid). Prufer's use of plant remains found at the site is clearly in-
consistent. His theory is further weakened by the fact that he fails to mention
what techniques were used to recover plant remains. Although the presence of
corn in Hopewellian contexts is of considerable importance in solving the ques-
tion of Hopewell subsistence, poor techniques and mis-use of available evidence
limits the value of his report.

The three studies so far discussed, generally represent the manner in which
archeologists have used plant remains for interpretive purposes. Although it is
now evident that small materials, such as seeds, are preserved in archeological
contexts and can provide information about sites, no attempt has been made
by archeologists to standardize procedures or outline the principles of using seed
analysis for either ethnobotanical or environmental inference. Without such
work, conclusions drawn will always be subject to question. Fortunately, pale-
oecologists are also interested in seed analysis. The most comprehensive study of
the general principles and procedures of plant macrofossil (seed and fruit) an-
alysis encountered in the present survey of the literature, is that by Watts and
Winters (1966). Although, the major focus of their report is environmental, their
findings have relejvance for those interested in the approach for other interpre-
tive purposes as well. For that reason a rather extensive summary of their re-
search is presented here. Essentially, Watts and Winters were interested in com-
paring the results of seed analysis to those of pollen analysis.

The authors suggest three fundamental requirements necessary for produc-
tive macrofossil analysis: 1) Quantitative procedures must be used and presented
clearly in any report. Frequency symbols such as "cc" (very common), and "r v
(rare) are inadequate. Rather, frequencies must be related to the total number
of seeds in a stated volume of sediments. 2) Large numbers of fossils should be
collected to enhance the value of the quantitative approach. 3) The stratigraphic
position of finds should be accurately related to a detailed pollen diagram (for
archeology, 'we might add or related to other environmental data recovered).

Seed Analysis In Archeology 169

The findings of Watts and Winters in their comparison of seed to pollen analysis
can be summarized as follows:

1. The total pollen that falls on the surface of a site is referred to as the
"pollen rain." Pollen counts are suitable for statistical treatment based on
the practical assumption that pollen is randomly mixed in the pollen rain.
Seeds also fall as a "seed rain" on various sites though they are not mixed as
efficiently as pollen grains nor produced in so large a quantity. These differ-
ences, however, are only a matter of degree, not kind, and do not vitiate
statistical analysis for interpretation.

2. Dispersal distance for seeds is more limited than for pollen. Therefore, the
seed rain gives more information about local vegetation, whereas pollen an-
alysis determines the nature of regional vegetation.

3. Both pollen and seed diagrams show "regional parallelism." That is, dia-
grams from one reasonably homogeneous vegetational unit are very similar.
Therefore, if vegetation changes in response to a change in climate, seed rain
changes at the same time pollen rain does.

4. The recurrence of macroflora sequences in different localities suggests that
each major plant formation or community has a distinctive seed rain that re-
flects the abundance of each big seed producer.

5. There are difficulties using both pollen and seeds for environmental recon-
struction. With pollen there are problems of certain species being transported
over long distances and significant differences between plants which are wind-
pollinated versus insect-pollinated. The- problem of dispersal over long dis-
tances is not encountered in seed analysis, but seed analysis is complicated by
the fact that heavy-seeded plants have much lower dispersal powers than light-
seeded plants do. There is also the problem of differential production of seeds
in annual and ephemeral plants compared to long-lived species. The former
have a high production rate of small seeds; the latter have a low production
rate of heavy seeds.

6. Seed statistics and pollen statistics have much the same kind of validity,
and are subject to many of the same errors.

7. In general, pollen and seed analysis produce complementary, not contra-
dictory results. Seeds tend to give a better (fuller) indication of the kinds of
plant communities that were present at a given site.

8. Fossil seeds are usually determinable to species. Insect-pollinated species

170 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

with low counts in pollen diagrams may be common among the macrotbssils.
A rich macroflora, therefore, adds substantially to the information obtained
from a pollen diagram.

9. Seed analysis may result in the reassessment of much pollen data recorded
to date. Long distance transport of pollen often biases the pollen diagram,
especially if local plants are low pollen producers.

10. Quantitative seed study becomes especially important in elaborating pro-
cesses not apparent in a pollen diagram.

The Watts and Winters' study resulted in a number of findings relevant to the
use of seed analysis in archeology. First, their work indicates that seeds found at
sites can provide data on local environment. For the archeologist, this means
that sampling in the field must be done with great care. In order to get a repre-
sentative sample of the "seed rain," soil samples must be taken from areas all
over the site and outside of the actual area of occupation. If this is not done, the
seeds recovered may reflect only those species selected for use by the aboriginal
inhabitants. Even if the investigator's main interest is in the subsistence or econ-
omy of the groups occupying a site, sampling should not be limited to features,
as is often the case. Seeds found in features must be compared to those found in
other areas of a site to ascertain if they occurred there naturally, as a part of the
seed rain, or because they were used by the inhabitants. Secondly, for seed anal-
ysis to be meaningful, the results must be quantified. This procedure is critical
for environmental interpretation since such information is available only if the
plant community at a site can be reconstructed. These communities, which re-
flect local environmental conditions, are defined on the basis of the proportion
of species present (Curtis 1959:3). Again, even if the major focus of a study is
ethnobotanical, the quantitative approach is necessary. To state merely that
certain species are present at a site, reveals nothing about the role they played
in the subsistence-economy of prehistoric groups. Finally, it should be noted,
that seeds, unlike pollen, can usually be determined to the species level. Plant
communities can 'be defined only if the species present can be identified.
Furthermore, seeds identified only to genus level, can easily bias cultural inter-
pretations. Different species of a single genus can be either native or non-native.
If the latter species are present in archeological contexts, contamination is indi-
cated. This kind of information is unknown without species identification.

Archeologists interested in utilizing plant remains must consider a number of
variables which are not especially important to the paleoecologists who work at
non-habitation sites. The problem of selecting areas to be sampled has already
been discussed. Poor sampling techniques can invalidate both cultural and envir-
onmental interpretation. After seeds have been recovered from a site and identi-
fied, the next problem confronting the archeologist is to determine what pro-

Seed Analysis In Archeology 171

cesses might account for their presence there. These processes may be either
natural or cultural. Plants are extremely portable and can be transported by man
from distant areas. This can complicate environmental interpretation. Con-
versely, for ethnobotanical interpretation, many seeds recovered may be those
of plants which flourish in the disturbed habitat created by man's presence at a
site. To assume that these same plants were exploited by man simply on the basis
of their presence at a site is an over-simplification. Environmental interpretation
is also complicated by the fact that man both intentionally and unintentionally
disturbs the native vegetation of an area. Many human activities such as disposal
of waste, clearing of sites, burning areas, and collecting plants, all affect soil
conditions and micro-environments at a site. To avoid mis-interpretation, these
variables must be carefully considered, and sampling procedures should be de-
signed to control them as much as possible. Non-human disturbance is another
problem encountered in archeological sites. Rodents and other vermin are at-
tracted to are.as inhabited by man. Burrowing animals are especially relevant,
since they often store various nuts and seeds in their sub-surface nests. Natural
phenomena, such as flooding, can also be responsible for introducing foreign
seeds to a site. Again, only carefully controlled sampling can help to avoid
serious mis-interpretations.

The natural processes affecting the plant community at any site must be
understood prior to drawing any inferences regarding climatic change on the bas-
is of seeds recovered archeologically. Soils, water, temperature, light, atmos-
pheric conditions, natural fires and the general biotic conditions all influence
vegetation and are not of equal importance at any one time (Daubenmire 1959:
2-3). Each factor assumes increasingly greater importance whenever and
wherever it begins to tax the ability of the organism either to tolerate it in
greater intensity or to survive under lower intensity. Each species has a minimum
and maximum tolerance for each of the various factors (Ibid). Clearly, these i
factors are all significant and variable at archeological sites and can be sub-
stantially altered by man's activities.

The above survey of the published material relating to seed analysis in arch-
eology suggests several steps necessary for the successful application of the ap-
proach. First, sampling in the field must be done with great care to insure that a
representative nmple of the seeds present are collected. This should include tes-
ting within and outside of the area of habitation. Second, laboratory techniques
should be employed which allow for maximum recovery of seeds from the soil
samples. Third seeds should be identified to the species level and quantified in
order to reconstruct the native plant community and to check for contamina-
tion in the samples. Next, cultural and environmental inferences can be drawn.
Finally, these interpretations should be compared to those based on other kinds
of evidence re< >vored at a site. All of these steps should be discussed in the pub-
lished report. Only if this is done, can the results of the seed analysis be evalu-
ated and compared to findings at other sites. An attempt was made in analysis

172 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

of the seeds from the Harvey site to follow these steps. Unfortunately, there
were no remains at the site which could be used to supplement or compare to
the results of the seed analysis. Thus, the investigation was more useful as a
study of the methods and problems of seed analysis that as a procedure which
yielded conclusive information about the site.

SAMPLING PROCEDURES

A basic pre-requisite for meaningful seed analysis is adequate sampling pro-
cedures in the field. At the Harvey site, soil samples were taken at various loci
in each feature, placed in plastic bags appropriately labelled and transported to
the laboratory for processing. In addition to sampling features, soil samples were
also taken from each of four test columns. Two of these were located within the
site proper, and two were located south of the occupation zone, toward the
Fox River where there was no evidence of aboriginal activity or modern plowing
(Map 1). Each test column was excavated down to water table and soil samples
were collected at 0.2' intervals. As has been mentioned, it is vital to sample areas
outside the site. There is no other way to determine if the seeds recovered
within the site are a part of the native vegetation of the area or only those
selected for use by the inhabitants or a mixture of the two. It is also necessary
to sample areas other than features within the site to enhance cultural
and environmental interpretations. One of the major criticisms of the published
archeological reports that include seed analyses is the failure of investigators to
provide information on sampling procedures. Stuart Struever, for example, bases
some of his conclusions of the presence of Chenopodium seeds in features
(Struever 1962). Unfortunately, he does not report what areas of the site had
been sampled. At the Harvey site Chenopodium seeds were also found in com-
paratively large numbers in the features. However, these seeds were also present
in large quantity, in the test columns inside and outside of the site. The argu-
ment that seeds found heavily concentrated in features reflect food supplies of
the inhabitants loses its persuasion if the same seeds are found in equally high
proportion in other areas tested, particularly if they are found in samples taken
outside of the immediate area of occupation.

Two recommendations with regard to sampling in the field can be suggested
on the basis of work done at the Harvey site. First, in testing outside of the site,
it is necessary to take samples from a number of different locations. The sam-
ples taken from the test columns near the Fox River represented only one micro-
environment near the site and it appears, as will be shown, that the area may
have been disturbed by human activity even though this was not apparent during
excavation. Secondly, a collection of the modern flora should be taken while in
the field. This could be helpful in identifying seeds found archeologically and
could also provide some comparative information concerning past and present
environmental conditions at the site.

Seed Analysis In Archeology 173

LABORATORY TECHNIQUES FOR SEED RECOVERY

One-liter samples from each feature and from each 0.2' level of one test col-
umn outside and one inside the site were processed in the laboratory. The sam-
ple size was arbitrarily chosen after several samples had been tested and it was
found that this unit produced a fairly large number of seeds. Sample size is ob-
viously to be determined by the nature of the material at any particular site. Use
of a standard sample size is not demanded in seed analysis. The number of seeds
per unit will vary depending on the area from which the soil samples were taken
and it may be necessary to process different volumes of sediment for different
samples in order to recover a significant number of seeds. If this is done, it is
critical that seed frequencies be related to the total number of seeds in a stated
volume of sediment in the published report.

The major concern in processing soil samples should be to recover enough
seeds so that the proportions of different species will be reflected. Seeds from
the Harvey site were extracted from soil samples using Stuart Struever's water-
separation and chemical flotation techniques with minor modifications. Al-
though Struever suggests processing samples in the field, all of the flotation in
the present study was done in the laboratory. It was felt that greater control of
the material could be maintained in a laboratory setting where the chances of
contamination and seed loss are minimized.

Initially, each one liter sample was placed in a container which was then fil-
led with water. Seeds which floated to the surface were poured into fine-mesh
crinoline. More water was addeo* and the solution gently stirred until material no
longer floated to the top of the container. The residue left after this process was
completed, when necessary, was boiled in a sodium bicarbonate solution to dis-
aggregate cohesive sediments and tjiereby release any seeds which may have been
lodged in the matrix. Each sample .was then washed through a no. 40-mesh
screen, which is fine enough to ^prevent the loss of any seeds. Larger material,
such as pebbles can be separated out by placing a larger-mesh screen on top of
the no. 40 screen. Seeds which had floated up initially, and those recovered in
the water screening process were carefully extracted with a small artists' brush,
using a low-powered binocular microscope to scan the residue of each sample.
The seeds were then placed in appropriately labelled glass vials.

Chemical flotation was employed in processing some samples. The purpose
here was to see if this technique would be an expedient way to separate seeds
from the silty residue which remained after flotation and screening. Struever
had used this technique to separate bones and seeds. Residues containing both
floral and faunal material were placed in a zinc chloride solution which has a
specific gravity of 1.62, in which the light plant remains float while the heavier
bone is suspended at a lower level and silt sinks to the bottom (Struever 1968:
355-357). With the Harvey site material, since there were no bones, the tech-
nique was used to separate seeds from silt. To test the method, two liters of soil

174 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

from the 0.4-0.6* level of TPB (test pit B) were processed, one using chemical
flotation and one just water flotation and screening. The results can be seen in
Table 6. It will be noted that chemical flotation allowed for the recovery of
many more seeds than the water-flotation and screening process. However, the
proportions of various species were not substantially different in the samples
processed. Only in the case of the minute seed type, CAPSELLA BURSA-PAS-
TORI, did the chemically processed sample differ from the other. In TPC (test
pit C), several samples were again subjected to chemical flotation after water
flotation and screening. In these samples no additional seeds were recovered
through the use of the chemical procedure. Generally, in samples where the silt
is fine enough to pass through the no. 40-mesh screen, the technique of chemical
flotation does not affect the proportion of seeds recovered. For each different
site the success or necessity of chemical flotation will vary depending on the na-
ture of the soil. It is advisable to at least experiment with the chemical process
to see if it is worthwhile for separating seeds from silt. The solution is made by
mixing two pounds of granular zinc chloride in one quart of water. It can be
used several times and should be stored in acid resistant containers. Since the
solution is extremely caustic, the chemical flotation procedure must be used
with great care.

SEED IDENTIFICATION

The value of seed analysis is greatly dependent upon the accuracy of the
identifications made. Ideally, someone familiar with seed taxonomy should
be responsible for identifications made of archeological specimens. How-
ever, such specialists are not often available for this work and for an approach
to be of general applicability in archeology, the archeologist must, of necessity,
attempt to acquire skills normally outside of his immediate field. Seed identifi-
cation can be undertaken by the archeologist if a reference collection is available
and if his identifications can at least be checked by specialists in the field of seed
taxonomy.

The initial process of separating seeds into distinguishable groups is neither
difficult nor time consuming. The most useful characteristics for this preliminary
separation are size, shape, coloring and peculiarities of the outer surface of the
seeds (Martin and Barkley 1961:5). Seeds from the Harvey site were separated
into types by examining each specimen under low-power magnification and
placing those seeds sharing similar characteristics of external morphology into
small vials. These were then compared to the seeds illustrated in the Martin and
Barkley seed identification manual, and tentatively identified to the genus level.
Next the seeds were compared to the reference collection in the seed laboratory
of the State Department of Agriculture, and when possible, identified to the
species level. All identifications were checked by Mr. Russell Marx, director of

Seed Analysis In Archeology 175

the State seed lab. As a further control, seeds also compared to the collection of
the University of Wisconsin, Department of Botany.

The archeol jist employing seed analysis is confronted with several problems
with regard to plant identification of the basis of seeds only. Identifications
would be much more secure if other parts of the plant were also present in arch-
eological contexts since in some cases the seeds of different species are almost in-
distinguishable. Another problem for the archeologist is that species determina-
tion for some plants is possible only by studying the internal morphology of the
seed. Unfortunately, archeological specimens are unlikely to have internal struc-
ture intact. Seeds of different genera may be very similar in external morphology
and can be identified only on the basis of internal characteristics (Martin and
Barkley 1961:6). These problems are somewhat balanced by the fact that many
plants can be identified to the species level on the basis of external morphology
of the seed. The identifications made of the Harvey site specimens are pre-
sented in Tables 5-7. It should be noted that not all seeds could be determined
to the species level. Those which could not be so identified are of limited value
for either environmental or ethno botanical interpretation.

Prior to reconstructing the plant community represented at the site for en-
vironmental interpretations or utilizing the identifications for ethnobotanical in-
ference it is useful to ascertain the ranges and habitats of the species present.
Such information is available in GRAY'S MANUAL OF BOTANY (Fernald
1950). All of the species present at the Harvey site are found in the area today.

176 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

Table 5. Seeds from Features - WK 236
( 1 - 1 i ter samp les)

CHENOPODIACEAE
Chenopodium album
C. hybridum gig-

antospe rnum
PORTULACEAE
Portulaca oleracea
CARYOPHYLLACEAE
Lychnis alba
POLYGONACEAE
Rumex acetose i la
Polygonum erectum
P. convolvulus
Polygonum sp.
ERICACEAE
Vaccinium sp.
OXALIDACEAE
Oxa 1 i s st r icta
GRAMINAE
Setaria sp.
Panicum sp.
Unknown sp.
CRUCIFERACE
Capse 1 1 a bursa-

pastor i
ROSACEAE

Rubus occidental is
AMARANTHACEAE
Arnaranthus retroflexus
JUGLANDACEAE
Carya sp.*
UNI DENTI FIABLE

Totals:

1 1

103

1 1

152

'fragments only (hickory nut)

Seed Analysis In Archeology

177

Table 6. Seeds from TPB - WK 236
( 1 - 1 i ter samp 1 es)

ENOPODIACEAE

enopodium album
hyb r i dum g i g-
antospe rnum

RTULACACEAE

rtulaco oleracea

RYOPHYLLACEAE

chnis alba

LYGONACEAE

mex acetose i 1 a

lygonum erectum
ca rey i
convo 1 vu 1 us
Pennsylvani cum

LANACEAE

ysa

is hetero-

phy 1 1 a
ICACEAE

cc i n i urn sp.

ALIDACEAE

al i s st r i c ta

PHORBIACEAE

a 1 pha v i rg i n i ca

AMINAE

taria sp.

nicum sp.

known sp.

UCIFERACAE

pse 1 1 a bursa-

pastor i
SACEAE

bus occ i dental i s
ARANTHACEAE
aranthus retro-

f lexus
RBENACEAE
rbena sp.
IDENTIFIABLE

TALS:

o
i

CM
CM

1 1

1 15

136

145

390

260

544

116

.. 2

178 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

Table 7. Seeds from TPC - WK 236
(1-1 i ter samples)

000

O CM J-

CHENOPODIACEAE
Chenopodium album
C. hybridum gig-

antospernum
PORTULACACEAE
Portulaca oleracea
CARYOPHYLLACEAE
Lychnis alba
POLYGONACEAE
Rumex acetosei la
Polygonum careyi
P. convolvulus
P. pennsylvanicum
SOLANACEAE
Physal is hetero-

phyl la
ERICACEAE
Vacc i n i urn sp.
OXALIDACEAE
Oxa 1 i s str icta
EUPHORBIACEAE
Alca 1 pha vi rgi n i ca
GRAMINAE
Setaria sp.
Panicum sp.
Unknown sp.
CRUCI FERAE
Capsel la bursa-

pastori
ROSACEAE

Rubus occidental is
AMARANTHACEAE'
Amaranthus retro-

f lexus
LEGUMINOSAE
Med icao 1 upu 1 i na
COMPOS I TAE
Ambrosia sp.
UNIDENTIFIABLE

TOTALS:

VO CO O CM _3"

CM CM CM

I I I

O CM

o CM' CM'

o
i i i i i i i

vD CO O CM ^J- \O CO

139

172

189

2\k

120

109

607

606

^82

597

Seed Analysis In Archeology 179

It is significant, however, that many of the species are not native to the United
States. Information concerning the habitat of each species found at the site is
presented below.

NON-NATIVE SPECIES

CHENOPODIUM ALBUM (Pigweed, lambsquarter>weed of cultivated and
waste ground; naturalized from Europe (Fernald 1950:594).

PORTULACA OLERACEA (Pursland>Cultivated and waste ground; natural-
ized from Europe (Ibid.: 608).

MEDICAGO LUPULINA (Black Medick>Roadsides and waste places; natur-
alixed from Europe (Ibid.: 895).

CAPSELLA BURSA-PASTORI (Shepherd's pouch>Weed of cultivated and
roadsides, cultivated ground etc.; naturalized from Europe (Ibid.: 704).

LYCHNIS ALBA (White cockle, campion>Roadsides, borders of fields and
waste places; naturalized from Eurasia (Ibid.:630).

POLYGONUM CONVOLVULUS (Black bindweed>Weed of cultivated and
waste ground; naturalized from Europe (Ibid.: 588).

RUMEX ACETOSEILA (Sheep sorrel>Ubiguitous weed of worn-out fields or
sour soils; naturalized from Europe (Ibid.: 571).

NATIVE^PECIES

CHENOPODIUM HUBRIDUM GIGANTOSPERNUM (Maple leafed goose
foot)- Rocky woods, thickets, clearing and waste places (Ibid.: 594).

POLYGONUM PENNSYLVANICUM (Pinkweed>damp shores, thickets, cl-
earings and disturbed or cultivated ground (Ibid.: 581).

POLYGONUM CAREYI-low thickets, swamps, recent burns, clearings and
cultivated ground (Ibid.: 584).

POLYGONUM ERECTUM-disturbed soil, waste places (Ibid.: 579).

AMARANTHUS RETROFLEXUS-(Wild beet, pigweed>waste or cultivated
ground, semicosmopolitan weed (Ibid.: 603).

180 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

RUBUS OCCIDENTALS (Black Raspberry > Rich thickets, ravines and bor-
ders of woods (Ibid.: 822).

OXALIS STRICTA (Wood sorrel>dry, open soil (Ibid.: 944).

ACALPHA VIRGINICA (Copperleaf>dry soils (Ibid.: 961).

PHYSALIS HETEROPHYLLA (Ground cherry>dry open wootfs, clearings,
(Ibid.: 1258).

The presence of non-native species in the plow zone and the humus of Test
Pits B and C is to be expected since the seeds there represent the modern flora at
the site. However, the occurrence of such seeds, those either directly or indir-
ectly introduced to the U.S. by man, in the features and below plow and humus
levies in test columns is suspicious.

Features TPB TPC

(below plow & humus)
Non-native Species

Chenopodium album 133 7 6

Portulaca oleracea 12 13 2

Capsella bursa-pastor i 1

Lychnis alba

Rumex acetosei la 1

Medicago lupulina 2

Totals: 159 2k 13

Native Species

Chenopodium hybridum gi gantospernum ]Q 2

Amaranthus retroflexus 18

Rubus occidental is 123

Oxal is stricta 2 3

Acalpha virginica 25

Totals: 162 2? .3

There are three possible explanations to account for the presence of the non-
native species in these contexts. First, it may be that the species determinations
are inaccurate, and that in fact, the seeds identified as non-native are either in-
distinguishable from related native species, or simply incorrectly identified. How-
ever, since the seeds from the Harvey site were carefully compared to two ref-
erence collections and all identifications checked by a specialist in n-,ed identifi
cation, it seems more reasonable to seek anoiher explanation for the presence oi

Seed Analysis In Archeology 181

the non-native species. One possibility is that Botanists have erred in designating
these species as non-native. The supposedly non-native CHENOPODIUM
ALBUM found in large numbers at the Harvey site is particularly relevant here.
Yarnell, who found this species at the Juntunen site at a depth of about 8
inches, reports that although the species is generally thought to have been nat-
uralized from Europe, it is probably also native- to Alberta, Canada (Yarnell
1964:41). If this is so, then its wide distribution throughout the Midwest might
be due to the fact that it was frequently used by Indians for food and spread
through the area by them. Although this explanation may be adequate for the
controversial CHENOPODIUM ALBUM, it would be unlikely that the Botanists
were also mistaken about the origin of the five other species (non-native) found
in the features and the lower levels of TPB and TPC. Probably the most plausible
explanation is that the seeds identified as non-native are in fact, recent seeds and
that the soil samples from the features and the test pits were contaminated.

The problem of contamination or disturbance is frequently encountered at
sites by archeologists. As has been mentioned, when attempting to use seeds
found archeologically, the various mechanisms which could account for the pre-
sence of the plant remains must be considered carefully and critically. At the
Harvey site, much of the occupation zone was disturbed by modern plowing. It
is feasible that recent seeds could have become mixed into the features after
many years of cultivation, even though soil samples were intentionally taken
from the lower depths of features to avoid such contamination. Plowing, how-
ever, could not account for the presence of non-native plant remains from the
deeper zones of the test columns. A more likely source of this contamination is
the activity of small animals which live in soils, such as earthworms, insects, spi-
ders, crustaceans, etc. The effects of such animals could easily escape notice dur-
ing excavation in contrast to the obvious signs of rodent disturbance which can
be avoided in sampling. In cultivated soils, earthworms are known to burrow two
to three feet below the surface. "Incidental to carrying organic matter to deep
horizons. . .they leave tunnels behind them that facilitate movement of water
through soils. . ." (Thorp 1967:192). Small seeds could easily move to consider-
able depths through these tunnels. Likewise, ants also move materials in soils
through their tunnelling operations and reportedly store masses of seeds in the
sub-surface "bins" of anthills (Ibid.: 199). A further source of contamination,
especially in southeastern Wisconsin would be soil disturbance caused by natural
phenomena such as annual freezing and thawing which could produce cracks in
the earth through which seeds could filter downward from the upper levels of
the site.

The problem of contamination limits the kinds of interpretations which can
be made on the basis of seeds from the Harvey site. One solution to the prob-
lem would be establishing a method whereby the intrusive, recent seeds could be
distinguished from those which were actually associated with earlier deposits.
In some cases, the condition of the seeds might be such that the two types could

182 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

be separated. It might be expected that seeds found in features would be char-
red or somehow altered after remaining in the earth for a prolonged period of
time. At the Harvey site, unfortunately, only the hickory nut fragments, the
black raspberry seeds and a few of the unidentifiable seeds were conspicuously
altered (charred). Most of the seeds were in extremely good condition and
showed little sign of alteration.

The problem of disturbance has been a recurrent one at the Harvey site,
both with respect to the cultural remains and the seeds. Reconstruction of the
native plant community demands that samples be taken from undisturbed areas
where the natural seed rain would be represented. If, in such areas soil samples
have been taken at various depths in vertical columns, stratigraphic changes in
the plant community can ideally, provide data on environmental change. At a
stratified site, this information could be of great value in interpreting culture
change. At the Harvey site, the vertical columns both inside and outside the im-
mediate area of occupation were sampled at 0.2' intervals. Although there was
no cultural stratigraphy at the site, the column samples could be used to test the
potential of seed analysis as an indicator of environmental change. Every plant
community is characterized by a particular structure and appearance imparted
by the numerical proportions of the species which compose it. All species have
certain limits to environmental variables and those with similar limits tend to
go together (Curtis 1959:3). Thus, if changes in the proportion of various so-
ecies are noted in a stratigraphic column, environmental change is indicated.
Unfortunately, the soil samples at the Harvey site, appear to have been con-
taminated by the movement of recent seeds from the upper levels to the lower
levels of the test columns. Consequently, even as a test of some of the potential
of seed analysis, the stratigraphic samples were of little value.

Taken as a whole, the sample of seeds from the Harvey site does little more
than substantiate the conclusion that the site was disturbed. John Curtis, in his
The VEGETATION OF WISCONSIN, compiled and synthesized all of the ecolo-
gical information on Wisconsin plant communities available as of 1959. 34 types
of plant communities are defined by Curtis on the basis of species composition.
Most important in these communities are those species which are dominant, i.e.
those well adapted to the climate and soil groups of an area. These in turn, ex-
ert a controlling influence on the communities where they occur either because
of their size or their high population density. Generally, the dominant species
are trees. The interaction of the dominant species with general climate and re-
gional soil groups produce a series of microenvironments which differ according
to the biological characteristics of the dominants concerned (Curtis 1959:47-52).
Most of the remaining species must grow in these modified conditions and they
tend to be sorted out in groups aligned with the particular dominants. A com-
parison of the complex of plant remains found at the Harvey site with the type
of species prevalent in the various communities defined by Curtis, results in the
conclusion that most of the seeds are from those plants common in Wisconsin

Seed Analysis In Archeology 183

weed communities (Ibid.: 59 1-592).

When a disturbed site is relieved from the disturbing agents and allowed to
develop by itself, a secondary succession of plants is initiated. The pioneer sp-
ecies adapted to the disturbed conditions are generally called weeds. Weeds, as a
plant type, have been only loosely defined but such plants do share certain char-
acteristics. They are plants which are very vigorous; can withstand and surmount
high intra-specific competition; have great tolerance of soil disturbance; and gen-
erally possess a high reproductive potential. The seeds of such plants germinate
under extreme fluctuations of soil temperature and seedlings are tolerant of ex-
treme fluctuations of soil temperature and soil moisture typical of open, dis-
turbed sites (Ibid.:416). Prevalent species of weed communities in southern Wis-
consin, present in the samples from the Harvey site include: AMARANTHUS
RETROFLEXUS, CAPSELLA BURSA-PASTORI, CHENOPODIUM ALBUM,
LYCHNIS ALBA, OCALIS STRICTA, POLYGONUM CONVOLVULUS, POLY-
GONUM PENNSYLVANICUM, PORTULACA OLERACEA, and MEDICAGO
LUPULINA (Ibid.:591-592). In addition to these, RUMEX ACETOSEILA, CH-
ENOPODIUM HYBRIDUM GIGANTOSPERNUM, POLYGONUM CAREYI,
and POLYGONUM ERECTUM are also plants which thrive in disturbed habi-
tats.

Both aboriginal activity and modern farming would account for the dis-
turbance of the native plant community at the Harvey site. This explains the
presence of weed seeds in the soil samples taken within the site. Those seeds
found in the features may be from plants which invaded the site during abori-
ginal occupation or after the site was abandoned, since the disturbed conditions
would provide an excellent habitat for weed species. Apparently the area tested
south of the site was not free from disturbance, judging from the seeds recov-
ered since the species present are again those typical of disturbed sites. The
fact that the samples from the lower levels of the test column in this area appear
to have been contaminated with seeds from the upper levels, and that the area
tested appears to have been disturbed, precluded the possibility of recon-
structing the native plant community at the Harvey site. As a consequence of
these problems environmental interpretations could not be made on the basis of
the seeds recovered.

The seeds from the features do give some information about the possible cul-
tural activities undertaken by the inhabitants of the site. Charred black raspberry
seeds (RUBUS OCCIDENT ALIS) were found in seven of the twelve features at
the site. In feature 5, 103 such seeds vvere found in a Miter sample. Black rasp-
berry is a plant native to the area, which grows in rich thickets, ravines and the
borders of woods (Fernald 1950:822). The high frequency of this type of plant
in the features, especially when compared to the low frequency of the species in
the test columns, suggests the aboriginal inhabitants collected the plant for some
purpose. Yarnell reports that several Historic tribes including the Objibwa, Men-
omini, Sauk, Fox and Iroquois collected the berries of this plant and used them

184 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

for both food and medicine (Yarnell 1964:58). Black raspberries are available
from July-August, suggesting that the site was occupied during the summer, if
not at other times of the year. Another seed type found in large numbers in the
features, is CHENOPODIUM ALBUM. Again there are records of this plant be-
ing used by Historic Indians. Chenopodium seeds were collected by the Ojibwa,
while the greens of the plant were used by the Ojibwa, Potawatami, and Iro
quois (Ibid.: 55). The conclusion that Chenopodium was intentionally collected
and utilized by the inhabitants of the Harvey site is, however, unwarranged. It
will be noted that CHENOPODIUM ALBUM was found in large numbers in both
of the test columns. This plant, which produces many seeds, grows prolifically
in disturbed habitats. The high frequency of the seeds in both the features and
test columns indicates that it was probably just one of many weeds which flour-
ished in the disturbed soils of the site. Since the seeds in the features showed no
signs of alteration there is no reason to assume that they were exploited by the
prehistoric inhabitants of the site. The burned hickory nut fragments found in
Feature 3, undoubtedly represent a food source of the Indians at the site al-
though it is doubtful that the Indians came to the Harvey site specifically to col-
lect this resource, since the remains of hickory are not present in large nu.iibers.

The artifact inventory and features at the Harvey site suggested that the or-
iginal occupants of the site may have come to the area to collect some natural
resources. This hypothesis can be neither confirmed nor rejected on the basis of
the seed analysis. Seeds were recovered from most of the features, but with the
exception of the black raspberry and Chenopods, they were not found in large
numbers. If the inhabitants did come to the site to collect plants, very few re-
mains were left behind in the features. The seed analysis was also unproductive
at this site for environmental interpretation, largely due to the fact that the soil
samples came from contexts in which the native plant community was disturbed
and replaced by pioneer weed species.

COMPARISONS

Contamination of soil samples was a serious impediment to interpretation
based on seeds at the Harvey site. Seeds from three additional sites were identi-
fied in an attempt to discover if contamination is a fairly pervasive phenomena
or if it was one specific to the Harvey site. The criterion for determining con-
tamination was the presence or absence of non-native species in samples taken
below the plow zone or humus. Two of the sites chosen for this investigation
are located in Iowa. Rock Run Shelter (13CD10), was excavated during the sum-
mer of 1967 by Robert Alex of the University of Wisconsin. The site, a strati-
fied Woodland Rockshelter, is located in Cedar County, Iowa. Soil samples were
subjected to water flotation and water screening and the seeds were abstracted

Seed Analysis In Archeology 185

from the residues by Mr. Alex. Another Iowa site, the A.C. Banks site (13PM40),
was tested by Dr. David Baerreis during the summers of 1967 and 1968. Soil
samples at this site were taken in columns at two inch intervals to a depth of
over 1 40 inches. Dr. Baerreis' major interest was in investigating the potentiali-
ties of gastropod analysis in the study of minor climatic episodes that are in-
volved in the span of time of the late prehistoric period (Baerreis 1969:2). Seeds
found in the floated soil samples were separated from the residue by Dr. Baer-
reis and identified by myself (as were those of all three of these sites). The third
site considered in the present study is Aztalan, a Mississippian village site lo-
cated in Jefferson County, Wisconsin. Soil samples were collected from features
at this site by Mr. Manfred Jaehnig of the University of Wisconsin during excava-
tions conducted in the summer of 1967 by the State Historical Society of Wis-
consin. Seeds were extracted and identified from the residues of floated soil
samples from five features at the site.

The results of the study of these three sites are interesting though not meant
to be conclusive (see Tables 8, 9, 1 0). The samples collected from the two strat-
ified sites in Iowa appear to be free from contaminants. At both sites, CHEN-
OPODIUM ALBUM was present in the upper inches while in the lower levels
only the native Chenopodium species were found. This suggests that seeds did
not filter downward from the upper levels of the sites as was the case at the
Harvey site. At the A.C. Banks site, where seeds other than Chenopodium were
recovered, none of the species identified from the lower levels of the site were
non-native, although many of the specimens could not be identified due to the
fragmentary nature of the seeds. In contrast to the two stratified sites, are the
samples from Aztalan. The features at that site are located just below the plow
zone. Four of the five features tested in this study, contained non-native species
(see p. 40) and if we assume that CHENOPODIUM ALBUM is non-native, all
five of the features were contaminated with recent seeds. It might be tentatively
suggested on the basis of the Harvey/site ^as well as the three other sites inves-
tigated, that contamination is most likely in shallow sites where the occupation
zones are directly below a plowed surface. Furthermore, since non-native species
were not present in the lower levels of the two stratified Iowa sites, it may be
that contamination of the Harvey site test columns resulted from conditions
which are not usually present at archeological sites.

186 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

Table 8. Seeds from 13CD10

~ O "<M co

CM I I

I I CO CO OO OO

O CO <N -3" U"\

CHENOPODIACEAE

Chenopodium album

C. hybridum

1 ]

Table 9. Seeds from JE1-6? (Aztalan)

CHENOPODIACEAE

Chenopodium album

PORTULACACEAE

Portulaca oleracea

CARYOPHYLLACfAE

Lychnis alba

POLYGONACEAE

Polygonum careyi

P. convovulus

P. sp.

OXAL1DACEAE

Oxal is stricta

AMARANTHACEAE

Amaranthus retroflexus

LEGUMINOSAE

Tr ifol ium sp.

UNIDENTIFIABLE

Seed Analysis In Archeology

187

Table 10. Seeds from

CHENOPODIACEAE
Chenopodium album
C. hybridum
C. boscianum
POLYGONACEAE
Polygonum aviculare
AMARANTHACEAE
Amaranthus retro-

f lexus
GRAMINAE
Setaria sp.
LEGUMINOSAE
Tr i fol ium sp
VERBENACEAE
Verbena sp.
ULHACEAE
Celtis sp.
(presence/
absence)
UNIDENTIFIABLE

-3- 00

-3-

C-J CM

r^ oo

i i

O v>

r-> r--

oo
oo

till

OO O CM J-

CHENOPODIACEAE
Chenopodium album
C. hybridum
C. boscianum
POLYGONACEAE
Polygonum aviculare
AMARANTHACEAE
Amaranthus retro-

f lexus
GRAMINAE
Setaria sp.
LEGUMINOSAE
Tr i fol ium sp.
VERBENACEAE
Verbena sp.
ULMACEAE
Celtis sp.
(presence/
absence)
UNIDENTIFIABLE

188 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

Table 10 cont

CHENOPODIACEAE

1
1

Chenopodium album
C. hybridum
C. boscianum
POLYGONACEAE

Polygonum aviculare
AMARANTHACEAE

Amaranthus retro-
f lexus
GRAMINAE
Setaria sp.
LEGUMINOSAE

Tr i fol i urn sp.
VERBENACEAE

Verbena sp.
ULMACEAE
Celtis sp.
(presence/
absence)
UNIDENTIFIABLE

CONCLUSIONS

Although the seed analysis undertaken was of limited value in interpreting
the rather enigmatic Harvey site, it was useful in a broader sense as a test of the
potential and problems of the approach in archeology. This study confirmed the
fact that small seeds, normally overlooked in routine excavation can be collected
in the field and recovered through the use of water flotation and screening, pro-
cedures which do not require elaborate equipment or specialized skills. Thus, the
approach is not an impractical one for archeologists. Secondly, seeds can be
identified to the species level by the archeologist, if a reference collection is
available, so that it is not necessary to employ specialists from another field to
use the approach. Once the seeds have been identified, ethnobotanical and en-
vironmental information otherwise unavailable to the archeologist can poten-
tially be revealed. The potentials of the method as applied to the Harvey site ma-
terial were limited by certain problems. Sampling procedures were inadequate
for using the approach environmentally. In order to reconstruct the native plant
community, a step necessary for environmental inference, samples must be taken
from undisturbed areas. Only one area outside the site proper was tested and the

Seed Analysis In Archeology

189

area proved to have been disturbed. This problem could have been avoided by
taking samples from a number of different locations outside of the site. A more
serious problem encountered in this analysis is that of contamination. Samples
from both the features and the test columns indicate that recent seeds can move
from the upper levels of a site to considerable depths. Furthermore, at the Har-
vey site, there was no way to distinguish between the intrusive seeds and others
on the basis of seed condition. A final judgment as to the applicability of seed
analysis in archeology must await further investigation. The potentials of the ap-
proach for both cultural and environmental interpretation are demonstrable.
Other critical and systematic studies will determine if the problems encountered
at the Harvey site are widespread enough to inhibit the successful application of
seed analysis in archeology.

REFERENCES CITED

Baerreis, David

1969 "Environmental Reconstruction through Molluscan Remains:

A Preliminary Report on the A.C. Banks Site," unpublished
manuscript.

Barghoorn, Elso

1943 "Collecting and Preserving Botanical Materials of Archeo-

logical Interest," AMERICAN ANTIQUITY, Volume 9, 289-
294.

Curtis, John

1959 THE VEGETATION OF WISCONSIN.

Daubenmire, R.F.

1959 PLANTS AND ENVIRONMENT (2nd Edition).

Fernald, Merritt

1 950 GRAY'S MANUAL OF BOTANY (8th Edition).

Jenks, Albert

1902 "The Wild Rice Gatherers from the Upper Great Lakes,"

19TH ANNUAL REPORT OF THE BUREAU OF AMER-
ICAN ETHNOLOGY, 1013-1160.

Johnson, El den

1969 "Archeologicai Evidence for Utilization of Wild Rice," SCI-

ENCE, Volume 163, No. 3864, 276-277.

190 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

Martin, Alexander and Barkley, William

1 96 1 SEED IDENTIFICATION MANUAL.

Meighan, C.W., et. al.

1958 "Ecological Interpretation in Archeology, Part 1," AMERI-

CAN ANTIQUITY, Volume 24, No. 1, 1-21.

Prufer, Olaf

1966 "The McGraw Site: A Study in Hopewellian Dynamics,"

SCIENTIFIC PUBLICATIONS OF THE CLEVELAND MU-
SEUM OF NATURAL HISTORY, n.s. Volume 4, No. 1.

Spector, Janet

1970 THE HARVEY SITE AND SEED ANALYSIS, unpublished

M. A. thesis, University of Wisconsin.

Struever, Stuart

1962 "Implications of Vegetal Remains from an Illinois Hopewell

Site," AMERICAN ANTIQUITY, Volume 27, No. 4, 584-587.

1968 "Flotation Techniques for the Recovery of Small Scale Arch-

eological Remains," AMERICAN ANTIQUITY, Volume 33,
No. 3, 353-362.

Thorp, James

1967 "Effects of Certain Animals that live in Soils," SELECTED

PAPERS IN SOIL FORMATION AND CLASSIFICATION,
191-208.

Watts, W.A. and Winters, T.C.

1966 "Plant Macrofossils from Kirchner Marsh, Minnesota- A Paleo-

ecological Study," GEOLOGICAL SOCIETY OF AMERICA,
BuUetin:77, 1339-1355.

Yarnell, Richard

1964 "Aboriginal Relationships between Culture and Plant Life in

the Upper Great Lakes Region," ANTHROPOLOGICAL PA-
PERS, Museum of Anthropology, University of Michigan,

No. 23.

1966 "Archeological Plant Food Remains from Wisconsin," WIS-

CONSIN ARCHEOLOGIST, Volume 47, No. 4, 196-202.

The Oneota Component 191

THE ONEOTA COMPONENT AT THE PORTE DES MORTS

SITE, DOOR COUNTY, WISCONSIN

by Carol Mason

University of Wisconsin-Green Bay

The Porte des Morts site is a multi- component stratified
site on the mainland side of the Porte des Morts Straits
between the tip of the Door Peninsula and Detroit Island.
In a previous report on this site, Ronald J. Mason (1967)
presented information on the North Bay component; the
present report is limited to a discussion of the Oneota oc--
cupation while a third projected paper will cover Late
Woodland materials. For a description of the site, general
information on physical stratigraphy, data on land owner-
ship, etc., the reader is referred to Mason 1967. Such in-
formation will be repeated here only when pertinent to the
Oneota occupation.*

Archaeologically, the Oneota deposits at Porte des Morts
were very much like those at other sites in Northeastern
Wisconsin. Oneota sherds were present over a very wide
area, particularly on the surface and often with little depth.
Back from the shore, material was present at least in small
scatterings far into the woods and all along the dirt roads.
As far as formal excavation units are concerned, the major
concentration of Oneota artifacts was in Units E, F, H,
and I(Fig. 1). Elsewhere on the site, there were always
some Oneota sherds in the test pits and large excavation
units, but not in numbers great enough to indicate any
intensive settlement there.

Excavation Unit E was the point farthest down the shore
to yield any appreciable amount of Oneota material. Of the
444 Oneota sherds to come from this unit, the great bulk
of them were grit-tempered (333) and plain (324). From
Excavation Unit E, a series of test pits and one other ex-
cavation unit (G) failed to produce any signs of greater
concentration in that direction. However, in the woods beyond
these test pits, cottage construction may have destroyed
any evidence that formerly might have been present. At the

*I would like to express here our appreciation to Mrs. George Apfelbach of
Fish Creek, whose information originally drew our attention to the Oneota
occupation at Porte des Morts, and to the University of Wisconsin-Green
Bay for a Chancellor's grant which aided in preparation for publication.

192 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

NORTH PORT
FERRY

LANDING

too 1

Figure 1 The archaeological site at the Porte des Morts Straits.

far end of the Porte des Morts Straits, a series of small
limestone cliffs delimits the sandy area, a situation typical
of the Door Peninsula and incidentally serving to pinpoint
places of aboriginal settlement there. Beyond these cliffs,
no aboriginal artifacts were found. The situation of Porte
des Morts as the nearest sandy area facing the islands in
northern Lake Michigan suggests that it was not only a
sheltered living area with convenient access to fishing
grounds but also a jumping-off place for island-bound traf-
fic.

In all units, the vertical distribution of grit and shell-
tempered Oneota sherds was checked in order to see if the
same relative frequency through time obtains here as the
Mero site farther south (Mason 1966:160-161). There, grit-
tempered Oneota sherds were most frequent in deeper de-
posits, declining in the upper, more recent levels, with a
reverse distribution for shell-tempered sherds. The total
ceramic sample at Porte des Morts scarcely reaches half
that at the Mero site, however, and there is a clear confir-
mation of the trend in only two or perhaps three areas.

The Oneota Component 193

One of these, Excavation Unit F, was excavated in ar-
bitrary four-inch levels (as were all units where natural
stratification was not present). In this unit, a total of 323
Oneota sherds indicates a gradual increase in the fre-
quency of shell tempering through time and a concommi-
tant decline in the popularity of grit-tempering. Grit-
tempered sherds reach a peak in numbers earlier than their
shell-tempered counterparts and decline from that point
on. The largest sample of Oneota pottery (1827 plain and
decorated sherds) came from Excavation Unit F (see Table
1). In Excavation Unit II, where excellent natural strata
were preserved, there is a more ambiguous distribution of

Table 1. The distribution of shell and grit-tempered Oneota sherds
by unit and level. The percentages from the Mero site have been
recomputed to include only Oneota sherds.

shell grit

level no. % no. % totals

Excavation Unit E

0-4" 25 37.9 41 62.1 66

4-8" 90 27.3 240 72.7 330

8-12" 7 19.4 29 80.6 36

12-bottom 4 33.3 66.6 12

Excavation Unit P

0-4" 108 85.4 20 15.6 128

4-8" 80 5*.l 68 45.9 1*8

8-12" 11 28.9 27 71.1 38

12-bottom 45 9

Excavation Unit I

0-4" 330 78.8 89 21.2 419

4-8" 512 61.0 327 39.0 839

8-12" 154 47.7 169 52.3 323

12- bottom 63 34.1 122 65.9 185

Excavation Unit H

humus 79 92.9 6 7.1 85

upper midden 162 48.9 169 51.1 331

buried beach 4 4

Mero site Oneota totals (from Mason 1966, Table VI, 176)

1 150 75.8 48 24.2 198

2 478 72.1 185 28.0 663

3 143 49.1 148 50.9 291

194 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

grit and shell tempering. The percentages as they stand
(Table 1) confirm the temper shift, but a good bit of the
absolute increase in grit-tempered Oneota material in the
upper midden is owing to the presence of 44 sherds from a
single vessel.

When compared directly with the percentages from the
Mero site, the figures from Excavation Units F and I be-
have in a comparable manner allowing for the fact that the
deposits at the Mero site were more distrubed than at Porte
des Morts. Excavation Unit H, which has a truncated dis-
tribution in the ground as far as Oneota artifacts are con-
cerned, provides evidence of a time (upper midden) when
shell and grit-tempered sherds were equally plentiful.
Taking into account the top-heavy character of the grit-
tempered component in the upper midden, perhaps the
most valuable comment from Excavation Unit H is the abrupt
decline of grit temper in the humus. The early and def-
inite preference for grit tempering is certainly a remarkable
one given the overwhelming commitment to shell tempering
observable in the latest deposits.

Excavation Unit I, so rich in Oneota artifacts, was the
only one to produce any historic objects: a single opaque
blue bead and a piece of brass kettle (?) scrap from the
humus. The Porte des Morts Straits, at least in local folk-
lore, has been the scene of recent Indian settlements, bat-
tles, and maritime disasters. Archaelogically, in any event,
there is nect to nothing to demonstrate an occupation of the
immediate area in the historic period. It may well be, how-
ever, that Excavation Unit I was placed on the very fringe
of historic Oneota or Late Woodland settlements and that
evidence for their presence was removed and destroyed
during the preparation of the township park, the Northport
Ferry Landing, and the building of many cottages along the
shore to the north of the ferry pier.

As far as physical evidence recovered from the site,
the Oneota occupation is represented by very few features
of any kind. Although an occasional small trash pit was
found, there were no large storage pits, burials, house re-
mains, or even fire pits. The dense forest covering the site
is a young one and presumably was preceded by the less
scrubby, more mature pine forest destroyed during lum-
bering operations of the last century. The presence of this
forest may be a partial explanation for the absence of fea-

The Oneota Component 195

tures, but a cultural explanation is at least as tenable. The
conspicuously unforested (but unfortunately plowed) Mero
site to the south was similarly lacking in many definable
features, and it has been suggested that it might have rep-
resented a seasonal, perhaps a summer, camp (Mason
1966: 178). Over most of the Porte des Morts site, the
Oneota materials were most abundant in a gray-black mid-
den just beneath the thin humus cover. This gray-black
midden was 12 inches thick in places and faded in an often
indistinguishable manner irito the lower middens. Occas-
ionally, as in Excavation Unit H, clearly separable strata
were present, but this was the exception rather than the
rule.

POTTERY

Nost of the artifacts that could be unhesitatingly identi-
fied as belonging to the Oneota occupation were potsherds.
Only rarely were other artifacts associated so closely with
these sherds as to be most likely originally deposited with
them. In many cases, cheek by jowl deposition of Late Wood-
land materials left the original association difficult to de-
termine. The total ceramic sample was 3883 sherds, 2249
(57.9%) shell-tempered and 1634 (42.1%) grit-tempered. Of
this total, only a small number bore any decoration at all,
confirming an impression formed in the field that the Porte
des Morts Oneota is an overwhelmingly plain complex (see
Table 2). Only 142 body sherds from approximately 21
vessels had any surface decoration; these represented 3.6%
of the sherd total, 2.0% (78) of them being shell-tempered
and 1.6% (64) grit-tempered. An additional 222 were rim-
sherds representing a minimum of 86 vessels. For the most
part, shell or grit was sufficiently plentiful in the sherds
as to leave no doubt which was tempered with which. In a
small number of cases, however, only microscopic exam-
ination revealed the presence of shell casts; and it is in this
area where the potter seems to have been either in the
process of changing from one tempering material to another
or unmindful of what was used that some difficulties a-
rose. Unless otherwise stated, the presence of shell or shell
casts in the sherds was the determinant for identifying a
sherd as shell-tempered, whether or not some grit was

196 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

Table 2. Decorated

body sherds.

sherds

shell

ves.

grit
% sherds % ves. %

Lake Winnebago

29.5

20.0 ....

unclass. trailed

10.2

6.6 6 9.4 3 50.0

Perrot Punc.

8
2

10.2
2.3'

13.3 ...

6.6 . . .

broad trailed

fine line incised 8

10.2

6.6 ...

punctated

5.1

3**

20.0 7 10.9 2***** 33.3

misc. incised
and trailed

27.2

6.6 7 10.9

incised and
punctated

2.3

13.3 ....

Barney Incised

1****

1.3

6.6

unclass. zigzag . . . $4 68.8 1 16.7

78 99.3 15 99.6 64~ 100.0 ~ 100.0

* this is a minimum figure; maximum figure is 8, ail represented

by 1 small sherd each.

** this Includes the one body sherd belonging to a probable Lake
Winnebago Trailed vessel defined from Its rims and 3 trailed
sherds included in the above count as one LI vessel.
*** most of these are small scraps; therefore no vessel counts have
been made. The single exception consists of 4 sherds from a
child's pot.
*<MH total representation is 6 sherds, 5 of which are rims and included

in Table 3 as plain rims.

***** many of these sherds are very small t and some offer only one big
dimple as a decoration; maximum count is 6 vessels.

additionally noted.

There are no observable differences in surface color
between the grit and shell -tempered sherds; gray to buff
and gray-buff are the dominant colors with all three color
variations sometimes being present on the same sherd. The

The Oneota Component 197

grit-tempered sherds are on the average harder than the
shell-tempered ones, but this may very well be because of
considerable leaching and rotting away of the surface as a
result of lying in the ground. There is a slight difference
between the two groups as regards sherd thickness: a
sample of 302 shell-tempered sherds was' found to have a
range in thickness of 2 to 8 mm. with a mean of 4.68,
while a sample of 266 grit-tempered sherds had a range
from 2.5 to 8.0 with a mean of 5.09 mm. No coil frac-
tures were observed in any of the sherds, and the most
common feature left from the manufacturing process are
wipe marks along the interiors parallel to the rim.
In contrast to the material at the nearby Mero site,
few of the Porte des Morts sherds were split or sloughed
off to any great extent. The annoying bedding of a layer of
rootlets through the sherd cores and their subsequent split-
ting as the roots dried was a common feature of the Mero
Oneota sherds, but at Porte des Morts it was not as fre-
quent;. Almost universally, however, the rimsherds exhibited
some evidence of burned organic material on the interior
or exterior (or both). This was never very great in actual
amount (only 1 gram was recovered in scraping all the
Oneota rims), but its presence was a consistent feature.
One other general characteristic of the pottery as a whole
should be noted: not a single fragment of a handle was re-
covered nor are there any suitable places on sherds where
handles might have been attached. The lack of handles or
of possible handle attachments fits well with the relative
scarcity of such appendages at the Mero site.

SHELL-TEMPERED POTTERY

Shell-tempered pottery from the Porte des Morts site
is seldom in as pristine a condition as, for example, that
from Lasley's Point. The sherds are heavily leached and as
a consequence are distinguished by pitted and irregular
surfaces. As a rule, however, the pottery did not contain
enough shell in the first place to leave the sherds porous
and crumbly after its loss.

Trailed sherds comprise the largest category of dec-
orated body sherds. Of a total of 31 such sherds, 8 are
small with straight line trailing (2-4 mm. across), some

198 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

of it becoming quite shallow and faint at the ends of lines.
In all cases, the trailing is of vertical lines in rows, prob-
ably always ending at the vessel shoulder. Although only
small and occasional sections of shoulder are preserved,
it seems likely that these vessels conform to the common
Oneota shape of a constricted necked jar with an everted
rim. Most of these 8 sherds fall within the range of
Lake Winnebago Trailed and probably could be assigned to
that type. However, in view of the strikingly narrow trailing
and the noncommital size of the sherds themselves, they
have been listed in Table 2 as "unclassified trailed. "
Seventeen of the remaining trailed sherds belong to a single
vessel of the type Lake Winnebago Trailed (PL 2, top);
this vessel is additionally represented by 2 rimsherds (in-
cluded separately in the rim table as Lake Winnebago
Trailed rims). It was a large globular jar with a con-
stricted neck and a rim angle so acute as to approach the
horizontal (Fig. 2, F). Some sherds were badly sloughed off,
but what could be measured of the vessel wall was thicker
at the rim (5 mm.) than toward the shoulder (7 mm).
Although heavily shell-tempered, these particular sherds
were generally less decayed than many other shell tem-

plate 2-Trailed and incised shell-tempered sherds: top, LAKE WINNEBAGO
TRAILED.

The Oneota Component 199

Figure 2 Representative rim profiles.

200 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

pered sherds on the same site; whether this is owing to
original excellence or luck in the ground is difficult to
judge. Decoration consists of trailed lines (5 mm. across)
arranged in a band of at least 5 deep (probably several more)
horizontal to the rim and underlain by an array of vertical
trailed lines encircling the vessel. The lip, is, very broadly
and shallowly scalloped as if smoothed discontinuously by
a thumb or paddle edge; subsequent to this scalloping, the
same* tool that was used in decorating the vessel body was
laid transversely across the lip, avoiding the original ridges
made during scalloping.

Three other trailed body sherds belong to a Lake Win-
nebago Trailed vessel represented additionally by 10 rim -
sherds and one other body sherd. There is unfortunately no
fit of the rimsherds with the body sherds; but judging from
the close physical correspondence between the two groups,
there is little doubt that they do in fact belong together. The
flat lip is shallowly scalloped with some clay extruded on
the exterior. The vessel originally was quite large, tan in
color, and had a flaring rim of about 20 mm. in height. At the
point of juncture of the rim with the body, there is a row of
large (5-6 mm. in diameter) round to oval punctations en-
circling the constricted neck at intervals of 6-10 mm; the
single non-trailed body sherd belonging to this vessel bears
one of these punctations. Little can be said of the trailed
design on this vessel other than that the trailing was of rows
of straight lines. In general, this vessel represents the best
that the Porte des Morts potters produced; hard and thin
(6 mm. on the body) for its evident size, it has an air of
quality conspicuous by its absence elsewhere on the site.

The third Lake Winnebago Trailed vessel is represented
by three small sherds (PI. 2, bottom row, left) with punc-
tations in addition to trailed lines. Decoration consists of
groups of 4 vertical trailed lines ( 3 mm. wide) alternating
with a single column of punctations. The area preserved is
at the shoulder and provides only a small segment of what
might have been a much more complex design.

Two sherds (PI. 1, top right) of a very broadly trailed
vessel were also recovered from the site (in the sod of
Excavation Unit F). Since these rimsherds were broken off
very close to the lip, it is impossible to determine whether
the original vessel was a bowl or a constricted necked jar;
the angle of the broad finger-trailing to the rim argues that

The Oneota Component

201

Plate 1 -Shell-tempered decorated sherds: left, section of a PERROT PUNC-
TATE vessel; top right, exterior and interior of broadly trailed rimsherds.

the vessel walls were perfectly straight but offers no cer-
tainty either way. Decoration consists of broad (10 mm.)
trailed lines running down the exterior at a slight angle to
the lip. The lip itself was rounded,, and on the interior there
is a row of deep, narrow punctations parallel to the lip,
each separate punctation set at an angle to it; these punc-
tations were made with a slender, pointed tool with a blunt
tip (an awl?). The presence of this kind of interior lip treat-
ment places the vessel outside the range of the type
Koshkonong Bold as defined by Hall (1962: 72-75) unless
very broad trailing can be construed a sufficient criterion
for that type.

Eight sherds of Perrot Punctate. (Hall 1962: 175-177),
representing two different vessels, were recovered from
widely separated sections of the site (PI, left). The first,
represented by 7 body sherds (Excavation Unit I), was ap-
parently a large constricted necked jar. The design consists
of nested chevrons composed of groups of three parallel

202 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

incised lines, each group of lines with a fringe of elongated
punctations appended to the bottom member (Fig. 3, top). The
chevrons themselves are intersected by still other groups of
parallel lines in a total design not readily interpolated from
the remaining sections. A single rimsherd from the same
unit and level may belong to this vessel: it has a plain ex-
terior surface, a flattened lip, and long punctations on the
interior perpendicular to the lip. Nine plain body sherds with
repair holes drilled neatly into them from the exterior also

Figure 3-Pottery decorations: top, PERROT PUNCTATED; bottom, grit-
tempered vessel with zigzag design.

The Oneota Component

203

belong to this vessel. The other vessel of Perrot Punc-
tate is represented by a single sherd from test pit 30. Part
of a shoulder, it clearly shows a fringe of long punctations
appended to nested parallel lines.

The only other trailed shell-tempered pottery from the
site is a vessel which most nearly resembles Ramey Incised
(PI. 3, top). While not a typical example, it would not stand
out in a large sample of Ramey Incised sherds from such
a site as Aztalan. Sherds from Aztalan in the collections
from Lawrence University, for example, include a rim that
except for its design might be from the same vessel. The 6
badly leached and porous sherds represent a jar (or per-
haps a bowl; no cazuela or other shoulders have been pre-
served, however) that was originally very heavily shell-
tempered. Before weathering distorted it, the surface was
very smooth; even now in spite of the irregularities re-
sulting from leaching, the surface? is very smooth to the
touch. Surface color is black with a lighter interior color
(fading to gray and tan in places); the exterior surface was

Plate 3 Shell-tempered rimsherds.

204 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

carefully smoothed before firing and reflects light. The
sherds are thickest just under the lip (7 mm.), thinning to
3 mm. farther down on the body. Decoration consists of a
broad trailed line (6 mm.) in what was at least a partly
curvilinear pattern; on the surviving sherds, a curved line
arches to the right and then abruptly terminates in a short,
straight line perpendicular to the rim. There is a cameo
effect on the interior. The rim itself is short (12 mm.) and
set on the body at an angle only slightly less than 90 de-
grees (Fig. 2, G). The lip is rounded, plain, and the rim is
noticeably smoother on the interior surface than on the
exterior. The sherds from this vessel were recovered
from Excavation Unit A, where Oneota sherds from all
levels comprised only 2.8% (a total of 36) of the sherd
count. In stratum C of this unit, where all the sherds des-
cribed above were found, only 4 other shell-tempered sherds
occurred; the major pottery types were North Bay. Thus
this trailed vessel is isolated both geographically, in the
sense that where it was found is not one of the major
areas of Oneota deposition, and culturally, in the sense that
its most immediate and plentiful neighbors in the ground
were North Bay sherds.

There are 30 body sherds with parts of incised designs
on them (PI. 2). Eight of these (Excavation Unit H) are from
a single vessel with extremely fine line incising as dec-
oration (PI. 2, second row center); the incised lines are so
fine as to resemble scratches made by a pin and in places
are so light as to be almost non-existent. On all the sherds
the design is very petite and consists of a single incised
line from which are suspended short parallel lines and
short pairs of lines so set with regard to each other as to
form triangles. Evidently groups of parallel lines and-
triangles were alternated around the vessel shoulder. On
several sherds, there are indications that more than one of
these decorative lines may have been present, perhaps in
widely spaced parallel units. Except for the facts that this
is a technically well-made vessel and is apparently full-
sized as well, it might have been a child's pot, so unsophis-
ticated and scratchy is the decoration. The remaining
shell-tempered incised sherds are all small scraps; and
with one exception were decorated with what seems to have
been units of parallel lines (ca. 1 mm. wide). Judging from
the curvature of some of the sherds, these lines were ar-

The Oneota Component 205

ranged vertically on the vessel. The single exception is a
small, crudely modeled child's pot (represented by four
sherds) with incised decoration put on at random; short
incised lines and gashes intersect and diverge in no rec-
ognizable pattern.

Two sherds of incised and punctated ware from this site
provide some variation in the decorative modes. The first
of these, a small section of vessel shoulder, is hard, thin,
and impressive on this site because of its evident quality.
The original vessel was decorated, at least in part, by ver-
tical parallel lines ending at the shoulder; between every
other pair of parallel lines at their point of termination is a
small (4 mm. in diameter) annular punctation, evidently
encircling the vessel as a shoulder decoration (PI. 2, bot-
tom row, left). The other sherd is part of a rimsherd with
the flaing rim itself broken away. The decoration consists
of a neat trailed line (3 mm. across) bordered by perfectly
round punctations (PI. 2, center row, right). The surviving
decoration looks like the apex of a triangle surmounted by
punctations. The rim of this vessel was originally set at an
obtuse angle to the body, but none of it survives.

Shell-tempered sherds bearing punctations are rare at
Porte des Morts, and those with punctations as the sole dis-
cernible decoration are rarer still (PI. 1). All four of the
surviving punctated body sherds are from the upper 4 inches
of Excavation Unit I. One of these was a section of a con-
stricted neck originally encircled by a row of enormous
(12 mm. across) finger-tip impressions. Two other sherds
were less spectacular: one with a single isolated row of small
horizontal punctations and the o;ther with a similar row of
slightly larger ones.

Two sherds with repair holes were recovered from Ex-
cavation Unit F, both from the same vessel. Except for the
Perrot Punctate vessel described earlier, this is the only
instance of vessel repair noted at the site; no other drilled
holes, whether for repair or for suspension, were found.
The only other post-firing modification found was a roughly
rectangular body sherd with a six-pointed asterisk crudely
engraved on the exterior surface; its significance, if any,
is unknown.

There are 109 shell-tempered rimsherds from Porte des
Morts, and they represent a minimum of 42 vessels (see
Table 3). The categories used to describe these sherds are

206 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

Table 3. Varieties of lip treatment; all rims, including those
attached to decorated body sherds are included in this count.

shell grit

sherds
lip form

plain

scalloped

exterior lip

inner lip rim

notched

notched, cord -marked -1

sinuous

other

Lake Winnebago Tr.

total, number of rimsherds: 225

total number of vessels estimated from rims: 86

* this count includes 5 sherds of the possible Ramey

Incised vessel.
** a. minimum count, 26 at maximum.

for the most part the same as those employed in Mason 1966
for the Mero site. In some cases, the categories did not fit
the Porte des Morts material as well in spite of the close
proximity of the two sites. In particular, distinctions between
"scalloped" and "notched" forms of lip treatment as main-
tained for the Mero site were less successful here since they
seemed disconcertingly to fade into each other, sometimes on
the same vessel. In general, "notched" is limited here to
instances where actual removal of the clay occurred with any
questionable examples being retained in the scalloped cate-
gory. The distinctions between degrees of scalloping (shal-
low, moderate, deep) have not been employed because the

sherds
no.

ves.
% no. %

sherds
no. %

ves.

no. %

14*

12.8

16.7

11.3

18.2

20.1

28.6

51.7

17**

38.6

13.8

14.3

6.0

11.4

3.7

7-1

11.9

14.3

12.1

20.4

d -1

2 ."4

11.9

4.3

11.4

13.6

14.7

11.0

4.7

110

99.9

42~

100.1

116

100.1

100. <

The Oneota Component

207

total sample size is so small and so uniform compared to the
material from the Mero site.

Judging from the small number of such rimsherds, few
of the vessels had plain lips (PI. 4, bottom row, center and
right). Only 14 plain rimsherds were receovered (including
5 from the possible Ramey Incised vessel described above)
representing at most 7 vessels. One of these was a small
bowl (or deep cup) with a crudely flattened lip; three others
had small flaring rims (9-12 mm. high) and constricted
necks. There is no uniformity in the treatment of the lips:
they range from slightly flattened to rounded to thickly
rounded. One of the flaring rims has traces of what might
be brushing (or nearly obliterated cord-marking) on the ex-
terior surface. The distribution of these sherds on the site
is interesting; with the exception of the bowl or cup (rep-
resented by 4 sherds), the 5 Ramey Inci sed rims, and a single
rim from an isolated test pit, all of them come from Excava-
tion Unit I, the locus of the greatest Oneota activity. The bowl

Plate 4-Shell-tempered rimsherds: top, scalloped; center, exterior lip deco-
ration; bottom, interior lip decoration and plain lips.

208 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

was one of the few Oneota vessels (or sherds, for that mat-
ter) recovered from Excavation Unit A, where the possible
Ramey Incised sherds were found. In the case of the bowl,
however, the sherds were from the upper midden and humus
rather than the more deeply buried deposits.

Thirteen sherds, representing at most 5 vessels, were
classified as having sinuous lips (PI. 3, bottom right). This
category is composed of lips that have been manipulated,
probably between the fingers in some instances, so as to
produce a wavy effect when viewed from above. In most
cases, this form of treatment has thinned the lips ap-
preciably, sometimes almost to a point; in others the thin-
ning has been effectively masked by sufficient secondary
smoothing of the rim as to produce a thicker Hat top with
heavy clay extrusions on both interior and exterior surfaces.
For most of these sherds, little can be said of vessel size
or shape; one is evidently from a broadly flaring rimmed
vessel (rim height 31 mm.) with a constricted neck, and one
other may be from a shallow bowl. They are as a whole the
least well-made of the rimsherds; massive sloughing char-
acterizes over half of them, and they are conspicuously
pitted, leached, and soft.

A total of 22 rimsherds, representing 12 vessels, were
classified as scalloped (PI. 4, top). In many instances,
scalloping was done with a narrow implement laid trans-
versely across the lip; the effect is a regular, closely
spaced scalloped decoration, not the broad scalloping assoc-
iated with the Lake Winnebago Traced rims described above.
Most of these rims are additionally low (between 17 and 23
mm. high), thick, and in only three cases was enough of the
rim preserved to confirm the presence of the familiar
Oneota flaring rimmed vessel shape.

A minimum of 6 vessels are represented by 13 notched
rims (PI. 3, center). Some of these may be from constricted
necked jars, but the very small size of the sherds makes ves-
sel shape difficult to determine. Notching was done by means
of a narrow tool that removed a section of the lip either ob-
liquely across it (2 sherds, 2 vessels) or transversely (9
sherds, 4 vessels). In some cases, certainly, the notching
grades into the narrower forms of scalloping, but as was
indicated previously, an attempt was made to limit the use of
the term "notched" to instances where clay had clearly been
removed in the decorating process. One additional notched

The Oneota Component 209

rim (listed separately in Table 3) is from a- deep bowl with
a cord-marked exterior (PI. 3, left). The lip was notched
obliquely with a narrow tool, and cord-marking covers the
exterior from immediately under the lip as far down as the
vessel wall is preserved, with some indication that it was
smoothed over nearer the base. The cord-marking is neatly
done from left down to right; the cord iteself was a very
loose two-ply with a clockwise twist. This vessel was heavily
shell-tempered with leaching and deep pitting observable on
both surfaces.

A sample of 15 rimsherds (6 vessels) was decorated
at the juncture of the lip with the exterior rim (PI. 4,
center row). This form of decoration involves notching or
impressing just beneath the lip and not impinging either on
it or across it. With but two exceptions, all of the sherds
were decorated with long narrow notches .set at right angles
to the rim or obliquely to it. In one case, the narrow notches
were replaced by deep punctation-like impressions made
from above, pushing the clay at the lip downwards; in another
case, the same effect may have been produced by a fingertip.
Two of the six vessels from this group were flaring rimmed
jars, one with a broad, gently flaring rim of 37 mm. and
the other a more typical 90 degree flare of only 15 mm. In
all cases, the lips were smoothed flat or nearly so before
decoration was attempted.

Only 4 sherds (representing 3 vessels) were decorated
on the lip interior (PI. 4, bottom left). Two of these have
been previously described in conjunction with broad surface
trailing. The remaining two sherds (and vessels) are alike
in the respect that their decoration consists of a bifurcated
impression (oddly resembling deer tracks) just inside the
rim. In one case the vessel was a flaring rimmed bowl with a
slightly greater than 90 degree flare and a rim height of
16 mm. The other vessel may or may not have had a smiliar
flaring rim. The lips of both of these vessels were neatly
squared with the decoration close to the inner edge in a
single row; in one instance, the impressions were quite small
(4 mm.) and in the other almost double that size.

GRIT-TEMPERED POTTERY
Excavation Unit H produced 44 sherds from a large flaring

210 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

Plate 5 Grit-tempered zigzag trailed sherds.

rimmed vessel, which in all respects is the most flam-
boyantly decorated of all the Oneota pottery recovered from
the site (PI. 5). There are 6 grit-tempered Oneota rim-
sherds (representing 5 different vessels) from the same
level (level 2) of the same unit, but none of these fits any of
the body sherds. The rimsherd nearest in ware character-
istics to them is a notched flaring rim of 30 mm. in height,
which, judging from some of the body sherds, was attached to
the vessel almost at a right angle. The body sherds rep-
resent an area from right under the rim to well down on the
shoulder, providing a good section of the body decoration.
These sherds incidentally provide an excellent example of
the uselessness of surface color as a guide to vessel sep-
aration or as a clue to reconstruction. The sherds as they
came out of the ground looked deceptively as if they rep-
resented several vessels since their surface and core colors
varied so extensively. A number of them were light buff
from one surface through the core to the other; still others
were red-buff or smokey-gray or even black all the way

The Oneota Component 211

through. In the laboratory it was discovered that all these
variously colored sherds were in fact from a single vessel,
the lines of color differentiation being the lines of original
breakage. Exposure to different conditions in the ground
produced the striking differences in color, even within such
a strictly delimited area as a single excavation unit. All
of these sherds are heavily tempered with grit particles,
many of which appear on the surfaces (especially on the in-
terior); this crushed grit is locally available hornblend or
gabbro, probably derived from granitic rocks. Some of the
particles are surprisingly large, reaching up to 4 or 5 mm.
in diameter. As a whole, this is a well-made vessel, hard
and with carefully smoothed surfaces; like so many other
Oneota vessels, it is thickest just under the lip (7 mm.)
thinning down (to 5 mm.) on the vessel body.

The same care cannot be observed in the execution of
the decoration, which, although complicated, is very sloppily
applied (Fig. 3, bottom). The area above the shoulder is dec-
orated by a combination of trailing (2-3 mm. across) and
punctating. A pair of parallel trailed lines form deep joined
chevrons all around the vessel, these are between 11 and 12
cm. from the apex of one chevron to its base where it joins
the next. The points of the bases are about 10 c.m. apart
while the tips at the top are about the same distance apart.
The figures are not geometrically precise or even artistically
well balanced. The underside of 'the set of trailed lines is
bordered by hemispheric punctations quite close to and in
some instances impinging carelessly on ,the lower line it-
self. The chevrons are filled with zigzag lines parallel to
the rim, and in one section where a good part of the design
still survives, a minimum of six such lines is present. The
topmost zigzag line is above the chevron peaks and com-
pletely encircles the vessel. All of the zigzags were drawn
with so little foresight that frequently their final zigs or-
zags had to be unnaturally stretched or compressed or con-
torted so as to fill a given space. As a result, the total design
lacks something in both symmetry and balance.

Three other vessels of trailed grit-tempered Oneota
pottery are among 6 trailed sherds recovered from the
site (PL 6, top right). One of the sherds is heavily grit-
tempered with a coarse texture and rough particles of quartz
protruding on the surface. The designs on almost all these
sherds appear to be groups of parallel straight lines, neither

212 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

1 3cm.

Plate '6 Grit-tempered decorated body sherds and rimsherds.

well done nor precisely executed. One of the vessels was a
globular jar, probably with a constricted neck, and trailing
from neck to shoulder. The one exception to the presence of
straight trailed lines is a single sherd (Excavation Unit
F) representing a section from the neck of a constricted
necked jar (J>1. 6, center, top). The section preserved is
just under the rim, 13 mm. thick at its thickest point, and
bearing a line of hemispherical punctations on the exterior
under the rim, very reminiscent of the similar line some-
times found on vessels of Lake Winnebago Trailed. The
trailed design on this sherd, however, is curvilinear, prob-
ably part of a guilloche or meander.

There are six small body sherds with incised lines on
them, but there is little that can be said of them other than to
record their presence. They are all very thin, tempered
with fine grit, and bear one or two short sections of narrow
incised lines. No comments can be made about design or
probable vessel shapes.

One of the punctated vessels found on the site has as its

The Oneota Component 213

sole surviving decorative mode a line of elongated punc-
tations around the shoulder at its widest point (PI. 6,
top row, left). As far as the 3 surviving sherds reveal,
there is nothing more on the vessel shoulder. A second small
gobular vessel has at least two rows (and probably several
more) of closely spaced "deer track" punctations above
the shoulder, each of which has left a cameo impression on
the interior. In this instance, the vessel was a remarkably
fine one- -thin, hard, and with a polished surface (PI.
6, center left). The remaining punctated sherds are Dis-
tinguished by the presence of large (10 mm. in diameter)
dimple-like impressions on the surface. In one instance,
these form a distinct row across the sherd, resembling
fingertip impressions. On the interior of this particular
sherd, there are internal impressions bearing overlapping
fingerprints as if the fingers of one hand had supported the
vessel interior as each exterior impression was made.

The 116 grit-tempered rimsherds from this site rep-
resent a somewhat larger sample than do the corresponding
shell-tempered rims (Table 3). The rims, however, are
spearable into fewer descriptive categories; and because one
of these categories, scalloped rims, accounts for over 50%
of the total number, the rim sample itself seems more
homogeneous.

There are almost exactly the same number of grit-tem-
pered plain rims as there are shell-tempered ones, and they
represent nearly the same number of vessels (PI. 6, bot-
tom). The 13 sherds are from 6 flaring rimmed vessels, 1
straight- sided vessel, and on^ whose shape is undetermined.
Four of the flaring rimmed ves'sels are distinguished only
by their temper and one of their dimensions from their shell-
tempered counterparts: the rims are higher, ranging between
21 and 33 mm. The single exception is a tiny flaring rim (4
mm. high) set at an angle of 90 degrees to the body and with
a neatly rounded lip. One of the vessels with a high rim is
very reminiscent of a small "funerary" pot from the Walker
Hooper sit<^ (McKern 1945 :P1. 66, no. 2). The lip is slightly
flattened d beneath it is a band of three deeply incised
lines encircling the vessel. The shoulder is more sharply
angled than the Walker-Hooper specimen, however, and the
body was considerably less bulbous (Fig. 2, H). The two
sherds representing this vessel were found isolated in test
pit 26, where their only companions in the ground were 2

214 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

^r*r.

Plate 7 Grit-tempered rimsherds: top, scalloped; center, notched, bottom, ex-
terior lip decoration and sinuous lip.

North Bay body sherds and a single shell-tempered Oneota
body sherd.

At least 17 and possibly as many as 26 vessels are rep-
resented by 60 scalloped rimsherds (PI. 7, top). In all cases
where vessel shape can be inferred, it is a constricted
necked jar; the rims flare from the bodies at a 90 degree
or greater angle, occasionally approaching 120 degrees (Fig.
2, C). Two of the vessels had very low rims (6 and 11 mm.)
and bulbous bodies beneath them without any discernible
constricted neck; both of these vessels were from the bottom
of Excavation Unit I and therefore represent some of the
earliest Oneota pottery on the site (Fig. 2, A and B). Be-
fore scalloping, the lips of the 60 rimsherds were rounded
and remain so between the scalloping; about one quarter
were additionally slightly smoothed and flattened, rather cas-
ually and perhaps with a thumb. Scalloping was accomplished
by impressing a round tool across the lip; the resulting
profile varies according to the diameter of the tool, the

The Oneota Component 215

closeness of the impressions, and the care taken by the
potter. For the most part, scalloping is spaced closely to-
gether, resembling in this respect the corresponding shell-
tempered sherds. There are three scalloped sherds, rep-
resenting two different vessels, with crude crosshatched in-
cising on the interiors of the flaring rims (Fig. 2, I). In
places this incising is partially obscured by burned food
material, but it is clearly not brushing or wiping. These
two vessels are the only instances of such decoration at
Porte des Morts, incising usually being confined to. more
traditional areas. One of the other scalloped vessels, an ex-
ceptionally heavily grit-tempered one, had a row of punc-
tations around the neck at the juncture of the rim and body
very like the punctations sometimes found on vessels of
Lake Winnebago Trailed. Most of the six sherds from this
vessel were from the bottom of Excavation Unit I and belong
to the earlier end of the period of Oneota occupation at the
site.

There are 7 sherds from 5 different vessels with deco-
ration at the juncture of the lip with the exterior rim (PI. 7,
bottom left). In four cases, these vessels were flaring rim-
med jars, but the shape of the remaining two is unkown.
Decoration was either by short vertical gashes (3 in-
stances) or by fingernail punctations (2 vessels). In all
cases, the rim was flattened before decoration was at-
tempted.

Fourteen rims, representing 9 vessels, have been clas-
sified as ' 'notched/' In mast eases, the sherds (PI. 7,
center) are so small that vessel shape is impossible to de-
termine. Some of the rims were certainly flaring, others
probably so. In all cases a narrow notching across the lips
(never oblique) is the mode of decoration; sometimes this
was done with so sharp a tool as to resemble slicing rather
than notching as such. In one instance, the rim had been
folded over before the lip was notched; in another, a line
of ragged punctations occurred on the rim exterior right
under the lip.

Five grit-tempered rimsherds from as many vessels
had sinuous lips (PI. 7, bottom right). Four of the sherds
came from flaring rimmed jars, some with very high rims
(30 mm.). In all cases, the manipulation of the lip appears
to have been done in the same manner as on the shell-
tempered rims.

216 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

STRATIGRAPHIC RELATIONSHIPS

There are only a few places on this site where strati-
graphic evidence is such as to provide information on
chronological change during the Oneota occupation. One
of these is Excavation Unit A, where the major component--
a heavy North Bay occupation in Level E--was the most
conspicuous feature. Above this level, in the humus and
underlying tan sand, there were only a few Oneota sherds
(11 plain shell-tempered body sherds plus 5 rimsherds rep-
resenting at most a plain cup and a scalloped rimmed jar),
but in Level C, in association with 102 North Bay and 7
Late Woodland cordmarked sherds, were the 6 sherds of the
possible Ramey Incised (PI. 3) vessel discussed above. If
this is Ramey Incised or a local derivative thereof, then
this is its northernmost manifestation in Wisconsin and the
earliest Mississippian pottery on the site. An intriguing
possibility is that the means that brought it here might be
part of the same trading complex or diffusion network that
put Ramey Incised vessels on Bois Blanc Island (Me
Pherron 1967:116-118). Below Level C, Oneota sherds were
almost totally absent (1 single shell-tempered body sherd
was found in Level E, otherwise exclusively North Bay in
content). Also within Level C, there were two rim sections
of a badly eroded shell-tempered vessel; both sherds now
exist in the form of sherd cores with only tiny fragments of
either of the original surfaces surviving. To begin with,
the vessel appears to have had a plain lip, but no actual lip
surfaces remain. The almost vertical rim is set at an angle
of slightly greater than 90 degrees to the body and measures
over 25 mm, in height (Fig. 2, D). The almost vertical place-
ment of thisjrim is unusual at Porte des Morts, where rims,
if they are anything, are more inclined to strive for a hori-
zontal position than to stand upright. A vertical rim profile
is certainly compatible with the identification of its near
neighbor as Ramey Incised; similar rim profiles are known
from some Grand River Plain vessels elsewhere in Wis-
consin (McKern 1945: PL 70, no. 12).

Excavation Unit I lacks natural stratification and was ex-
cavated in 4 inch levels down to a depth of 16 inches. There
is some indication of typological changes from top to bottom
as well as the previously cited temper shift from shell to
grit. Most of the sherds identified as Lake Winnebago Tr-

The Oneota Component 217

ailed (10 sherds, 2 vessels) occurred in the first 4 inches
with only 2 below 8 inches. Similarly, the single vessel of
Perrot Punctate from this unit is from the humus with only
1 sherd below 4 inches. Although the sample is regretably
small, it should be noted that while these two types were
most frequent in the humus, it is between 4 and 8 inches
that Oneota materials were absolutely most plentiful. Near
the bottom of Excavation Unit I were rimsherds from two
grit-tempered vessels with very low flaring rims. One of
these (represented by 5 rimsherds) has a tiny right angled
scalloped rim (in places under 5 mm. in height), which is so
low as to be almost not there. Beneath the rim, the vessel
body appears to have been bulbous and without much neck
constriction. This vessel and a companion with a slightly
higher rim (up to 11 mm.) were in a small pit appended to
the bottom of the midden.

In broader terms, the collection from Excavation Unit I
provides evidence of at least two different occupations:
one connected with Lake Winnebago Focus pottery and the
second with a grit-tempered complex dominated by Car-
cajou Plain. Both Lake Winnebago Trailed and Perrot Punc-
tate occur in the humus in company with an overwhelm-
ingly shell -tempered population of plain body sherds. Much
of the other pottery from this same unitjudging from the
absence of plain rims, the infrequency of body decoration,
and the presence of a substantial number of grit-tempered
sherds- -can be classified as Carcajou Plain. The 2 grit-
tempered vessels with very low flaring rims from the bot-
tom of this unit may additionally qualify as early Carcajou
Plain (Hall 1962:62). There is, however, little evidence to
argue for the presence of its companion type, Carcajou
Curvilinear, in view of the infrequency of curvilinear tr -
ailed lines. Since the type was defined to include an oc-
casional straight trailed decoration, some of the unidentified
straight trailed sherds- -particularly those with grit tem-
pering- -might be so identified.

Elsewhere on the site, few other examples of Lake Win-
nebago Trailed occur anywhere other than in the humus or
immediately below it. One trait, cited by Hall as a frequent
feature of Lake Winnebago Trailed pottery (Hall 1962:
172), the occurrence of a row of punctations encircling the
rim, may have an earlier antecedent here in the presence
of such rows on two grit-tempered vessels, one with curvi-

218 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

linear trailing from Excavation Unit F in the 4-8 inch zone
(PI. 6) and the other from the bottom of Excavation Unit I.

The single cord-marked rimsherd (PI. 3; Fig. 2, E) be-
longing to the Oneota occupation was found in an other-
wise undistinguished collection of plain Oneota body sherds
in Excavation Unit F'. It was beneath the humus zone ir 'ds
unit, and its only associates in the ground were all grit-
tempered. The Mero Complex Oneota contained 60 cord-
marked body sherds (Mason 1966:162), but none of these is
sufficiently diagnostic in its locus to provide any kind of
chronological information. Hall records a single cord-
marked Oneota sherd from a pit at Carcajou Point in as-
sociation with objects of historic date (1962:78), but again
there is no clear chronological assignment. Cord-marking
occurs late in a number of Oneota-like complexes in ihe
east, and there is no real evidence to countervene this im-
pression from the Porte des Morts site.

The total situation in Excavation Unit H is not very
helpful when attempting to decide where to place the 44
sherds of the grit-tempered, zigzag trailed Oneota vessel
(PI. 5) recovered from the upper midden. It is not com-
patible with Lake Winnebago Focus pottery, but it might
very well belong to the Grand River complex at Porte des
Morts. To use McKern's felicitous phrase, it may be a
"locally peculiar" specimen (McKern 1945:149). It might
also represent an entirely different occupation. The upper
midden in which it was found was a clearly defined stratum
below the humus; both grit and shell-tempered Oneota sherds
were plentiful, but decorated sherds were rare. The only
other decorated pottery associated with the zigzag trailed
vessel were 8 sherds from a shell-tempered pot with very
fine-line incising (PI. 2, center).

CONCLUSIONS

As a whole the ceramic complex at Porte des Morts
is distinguished by several characteristics. First among
these is the extraordinarily high incidence of grit-temper-
ing, which ranges from 42.1% of the total sherd count to 51.1%
of vessels estimated from rims. Nowhere else withir ^he foci
defined as Oneota does the percentage of grit-tempering
reach such a proportion unless it is at the Mero site, where

The Oneota Component 219

the percentage is 40% of gross sherd totals and 38% of all
vessels estimated from rims. This great emphasis on grit-
tempering may be mainly a phenomenon of the Lake Michigan
shoreline. Grit-tempered Oneota components have been
found in Wisconsin at least as far south as Kewaunee and
Sheboygan Counties (Hall 1962: 63-64; McKern 1945: PI.
54) and as far north as Rock Island. Oneota material in
the Neville Public Museum, Green Bay, includes a number
of very small surface collections from several sites in
Kewaunee County and some from immediately south of the
Sturgeon Bay ship canal in Door County. At one of these
locales- -that cited by Hall from Kewaunee County- -there is a
proportionately large sample of grit-tempered sherds. Most
of these 22 plain, grit-tempered body sherds are indisting-
uishable from the Mero site material, even to the presence
of tiny gold flecks in the paste on both surfaces. The single
grit-tempered rim has an obliquely notched lip and an almost
vertical profile over a very bulbous body. Another site
nearer Algoma again indicates the presence of grit-temp-
ered sherds, but in the museum collection, at any rate,
there are more shell than grit-tempered sherds present.
In the area near Sturgeon Bay but actually on the Lake Mich-
igan shore, a number of small surface collections in the
museum again attest to the presence of grit-tempered
Oenota, but here it seems to be associated with large quan-
tities of shell-tempered sherds. On Rock Island, preliminary
indications from recent Lawrence University excavations
are that there is at least one grit-tempered Oneota compo-
nent present, one stratigraphically separate from shell-tem-
pered material. Some grit-tempered Oneota sherds have even
been found on Bois Blanc Island in Lake Huron (McPherron
1967: 118-120).

Along the Green Bay shore, the major excavated Oneota
site at Point Sauble (Freeman 1956) has produced no grit-
tempered Oneota sherds at all, an odd contrast between the
east and west sides of the Door Peninsula. There is, however,
a multi-component site north of Point Sauble in what is now
Dyckesville that has produced a few grit-tempered sherds,
in a small surface collection in the Neville Public Museum,
there are 7 grit-tempered body sherds and 1 plain rim.
Associated shell-tempered material includes incised and
trailed body sherds as well as a typical Lake Winnebago
Trailed rim. Inland from Lake Michigan, there is as yet no

220 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

clear pattern for the distribution of the grit-tempered Oneota
variant. Preliminary examinations of Oneota sherds from a
series of sites along the central Fox River (Lawrence Uni-
versity collections) has produced very few grit-tempered
sherds at all. A site recently excavated by Lawrence Uni-
versity on an old channel of the Little Eau Plaine River has
produced a few grit-tempered sherds as far west as Stevens
Point, but the percentage may turn out to be quite small. As
far as present evidence indicates, the farther from the Lake
Michigan shore an Oneota site is, the less likely it is to have
a really large percentage of grit-tempered pottery on it.
Certainly, if the northeastern Wisconsin area is to contribute
significantly to the Oneota mainstream, the spatial para-
meters of the heavily grit-tempered Oneota will have to be
more closely determined.

Another important attribute of the Porte des Morts Oneota
is the total absence of handles of any kind. Whether a larger
sample of pottery might have included at least one or two is
a probability in view of the presence of three such appendages
at the Mero site with its 50% greater sherd total. A low fre-
quency of handles is, of course, a characteristic of the Grand
River Focus, but the almost complete absence of them in the
Door Peninsula sites is remarkable, even for Grand River.

Porte des Morts Oneota pottery is overwhelmingly plain,
again a trait linking the site within the Grand River orbit. Of
an estimated 86 vessels, only 24% (21 vessels) bore any dec-
oration at all (a total that would be reduced even further were
the Lake Winnebago Focus vessels eliminated from consid-
eration). As small a percentage as this represents, however,
it is actually double the percentage of decorated vessels at
the Mero site.

There is some difference betv/een the grit and shell-
tempered vessels with regard to variety of body decoration
present. Fewer grit-tempered vessels were decorated (6 of
21) and when body decoration is present on them, it is of a
more limited range of variation than is present on shell-tem-
pered vessels. Most of this sense of limited range is an arti-
fact of including the obvious Lake Winnebago Focus material
in any comparison; its presence weights the shell-tempered
component in an almost flamboyantly decorated direction.
Many more shell-tempered sherds are incised, and this rep-
resents one of the real differences between the shell and
grit-tempered decorative devices. The presence of more in-

The Oneota Component 221

cising is one of the differences between Porte des Morts
and the Mero site (36 sherds as opposed to 8 at the Mero
site). Another is the total absence of embossing at Porte
des Morts, a feature recorded in 10 instances at the Mero
site; embossing is a popular trait at some Grand River
sites, and its absence here may be a function of the smaller
sample size.

Plain lipped vessels are infrequent at Porte des Morts,
a fact of some importance in assessing its relationship
with the Grand River Focus. Of all rimsherds, only 12%
(representing 15 vessels) had plain lips, a percentage less
than half that at the Mero site (35%). The significance of
this low incidence lies in Hall's Grand River ceramic series,
where the important categories Grand River Plain and
Grand River Trailed are both defined as lacking lip modi-
fication. The low incidence of plain lips therefore places
much of the Porte des Morts Oneota in the alternate Car-
cajou series and specifically includes it in Carcajou Plain.
When breaking plain-lipped vessels down into their temper
categories, there is little difference between grit and shell:
9% and 8% of all vessels respectively (as determined from
rims). By way of contrast, there is a much higher incidence
of shell-tempered plain lipped vessels at the Mero site than
there is of grit-tempered ones (20% as opposed to 6% for
grit).

There is not much difference in the percentages of scal-
loped lips between the Mero site and Porte des Morts, even
allowing for the much larger sherd sample from the former
site. On both sites, the percentage of scalloped lips reaches
approximately 28% of shell-tempered vessels and 39% of
grit-tempered ones. At the Mero site, many of the remaining
shell-tempered vessels in the total vessel count have plain
lips, but at Porte des Morts many of the remaining vessels
fall into several minor lip categories, thus giving the impres-
sion of greater lip variation there.

One other trait of the Porte des Morts Oneota that de-
serves mention is the presence of incising on the broad
flat interiors of flaring rims (PI. 7, top right). There are
two vessels recorded for this trait, both with scalloped lips
and both grit-tempered. The incising on the rim interiors
is very crude crosshatching in both instances, some of it
so carelessly applied that it had to be repeated with some of
the later lines obscuring and crossing the original ones. In-

222 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

cising or trailing on rim interiors is not a common Oneota
trait, but it occurs at the Correctionville site (Henning 1961:
12, 13) on shell-tempered pottery and in much more formal
designs.

STONE TOOLS

All chipped stone tools at Porte des Morts, unless other-
wise noted, are made of a gray chert, sometimes white and
sometimes a mottled dark gray. Occasionally it is banded
or fossil bearing and very often it contains crystalline
inclusions whose presence must have rendered much of it
structurally unsound for tool making. Its source may have
been the tablular chert deposits found in the cliffs in places
along the shore; some of the artifacts still retain the long
flat planes of the original surface. Glacial gravels in this
area might have been an additional source of raw matt lal,
but no pebble rinds were observed on any of the surviving
artifacts nor was debris from smashing open glacial cobbles
recovered.

Twenty-five small triangular projectile points very
probably belong to the Oneota component. Their distribution
on the site, except for a few isolated specimens, corres-
ponds to the area of greatest Oneota concentration (Exca-
vation Units, I, H, F', E), but it should be noted that Late
Woodland materials occurred in these areas, too. Except
for two specimens which diverge sharply in two or more
attributes each, the triangular projectile points are quite
uniform within the limits of a not too precise flint tech-
nology. They are small and light. In length they exhibit
a mean of 20 mm. within a range of 20-33 mm. In breadth
and in thickness, the ranges f variation are 12-20 mm. and
3-6 mm. respectively. These dimensions are best reflected
in weight, a measurement providing striking contrast with
the stemmed and notched pc nts from the North Bay de-
posits (Mason 1967) and which *nust signify different modes
of hafting and possibly method of propulsion. The triangular
points range in weight from 0.6 to 3 grams with an average
of 1.5 grams. Actually, the arithmetic mean does not appear
to be a truly accurate appro" 4rv ^tion of the freq ? icy dis-
tributions within fractions of ^ ams since the distribution
curve is definitely skewed. For example, 7 points weigh

The Oneota Component 223

between 0.6 and 1 gram, 5 points between 1.1 and 1.5
grams, and the remaining 8 and 1.6 and 3.0 grams. Only
competed specimens were weighed, of course, but it is
evident from inspection of broken triangular points that
these determinations are not biased by this necessary
selection.

All of the triangular points have straight to some-
what convex lateral edges- -these attributes are fre-
quently combined on a single specimen. Two points have
very convex basal edges, 3 are slightly convex, 14 are
straight, and 3 concave. Only 4 points are really carefully
made; the majority are mainly flakes trimmed to triangular
form. Eight are classifiable as bifaces and 12 are es-
sentially uniface artifacts.

In addition to the triangular points, there is one small
triangular stemmed point (PI. 8) from the upper levels of
Excavation Unit B (Ex). It is unlikely Oneota specimen,
but it was found associated with both grit and shell-tem-
pered pottery as well as one single Late Woodland cord-
marked sherd. It is of a tan-gray chert, bifacially flaked
and lenticular in cross-section, 29 mm. long, 18 mm.
broad, 7 mm. thick, and weighs 2.6 grams.

Of a total of 7 scrapers that occurred in the deposits,
6 may belong to the Oneota occupation. Two are certainly
Oneota, the other 4 very probably so. The two scrapers with
unquestioned Oneota affiliation are quite small; one (from
the humus of Excavation Unit H) is a thumb-nail scraper
with a steep working face (roughly 25 mm. across), and
the other is a crude rectangular tool that resembles a
homemade gunflint that was never put to use (20 by 25
mm.). Two other scrapers are minimally worked flakes (P
1. 8) whose undersides are the unretouched bulbs of per-
cussion and whose only modification consists of the tiny
beveled scraper edge. Both are from Excavation Unit I in
contexts almost certainly Oneota in character. The scarcity
of scrapers in the Porte des Morts Oneota deposits con-
trasts with some other Oneota sites in Wisconsin, but it is
becoming increasingly clear that presence or absence of
end scrapers of itself is not as diagnostic a feature as was
once hoped (McKern 1945:133-134). It is more likely an in-
dicator of different emphases in the activities carried out
at each stie.

A single stone awl came from level 2 in Excavation Unit

224 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

cm.

Plate 8 Stone and copper tools: top, projectile points; center, stone awl; bot-
tom, "butter knife."

H. It has a general triangular shape with a blunted tip and
an overall use patina. It was held in the fingers, and a series
of long cuts are visible on one face where an attempt was
apparently made to cut it down for a more comfortable fit
(PI. 8, center; for a similar specimen see McKern 1945:
P 1. 40, 14).

A number of stone tools of unknown function occurred
in the deposits. One of these was a small assymmetrical
flake knife (?); it has few use marks along the blade, but
there is battering at the juncture of the haft section with
the blade, indicating a definite back and forth movement. Its
dimensions are 42 by 14 mm. Several small broken ovate
tools, bilaterally worked and not very well made, may have
been knives or preforms.

There are two unmodified, oval granitic cobbles that
exhibit signs of use, probably as hammerstones on chert.
The more intensively used of these (75 by 85 by 35 mm)
has seen service on both faces and along the edges; use
marks in the stone are deep, jagged, and striated. Appar-

The Oneota Component 225

ently, it was used to work material much harder than it is.
The second hammerstone (100 by 67 by 47 mm) had been
used less than the first with marks from use confined almost
entirely to the faces.

In Excavation Unit B (Ex), there was a concentration of
stone tools in the first two levels, which in this area in-
cludes the recent gray sand humus zone and the immediately
underlying light tan sand down to 9 inches below the surface.
Below 9 inches is a dark brown culturally sterile layer that
effectively seals off the North Bay middens beneath it. In
these upper two levels, at the juncture of the first two levels,
14 whole or partial stone tools were recovered from an area
less than five feet square in association with a large amount
of rough chert waste. These tools include large biface pre-
forms, 1 triangular point, the small stemmed point men-
tioned above, crude knives, and pieces of all of these. In
the same two levels, Oneota sherds were vitually absent.
This area contrasts with such loci as Excavation Unit I,
which contained 9 of the 25 triangular points but only 7
fragments similar to those in the B Extension. It also con-
trasts with sites elsewhere in Door County where masses
of crude unfinished tools and chert waste have been assumed
to be earlier in time. Here there is no question but that the
chert concentration is post North Bay and either Oneota or
Late Woodland in period.

BONE TOOLS

There were comparatively few bone tools from the en-
tire Porte des Morts site although bone preservation was
exceptionally good there. Only four of the tools could be
associated with the Oneota occupation. All of these were
from Excavation Unit H, and without exception they are
fragmentary and eroded. Three of the fragments are from
split bone awls; the fourth is a small section of worked
antler tine. The bone awls (2 from the humus, 1 from
level 2) conform to the usual Oneota variety: they are
made from split deer bone with an abrupt taper near the tip.
One of the surviving sections is a blunted base, the two
others tips only.

226 WISCONSIN ARCHEOLOGIST Vol. 51, No. 4

COPPER ARTIFACT

The only other artifact from the Oneota occupation is a
small copper ''butter knife" (PI. 8). It came from the humus
of Excavation Unit H where most of the pottery was shell-
tempered Oneota (3 Late Woodland cord-marked sherds
were also present). The knife is crudely hammered near
the tang, but in spite of age and corrosion, it is still sur-
prisingly sharp along its finely ground glade. Similar speci-
mens are associated with Late Woodland occupations at the
Juntunen Site (McPherron 1967).

SUMMARY

The archaeological site at Porte des Morts was oc-
cupied at least twice in its long history by peoples who may
be classified as part of the Oneota tradition. As |vas the
case at the Mero site, the bulk of the material fits best into
the Grand River Focus. The plainness of the pottery, the
absence of handles, and the compatibility of the pottery
types with categories formally set up by Hall for Grand
River are the principal criteria for so identifying it. In
the high incidence of grit tempering and the varieties of
decorated vessels, it is more nearly like the Oneota
complex at the nearby Mero site than any of the type sites
for the Grand River Focus. Like the Mero site, Porte des
Morts is remarkable for the number of Oneota traits that
are entirely lacking- -pipes, catlinite ornaments, bone tubes,
ground stone celts, sherd discs, etc. The second major
Oneota deposit is represented by a ceramic complex in-
cluding such types as Lake Winnebago Trailed and Perrot
Punctate and is on stratigraphic grounds the later of the
two occupations. Whether or not there were other distinct
occupations depends upon the interpretation given minor
ceramic groups present in the total collection. It is
probable that Porte des Morts was the site of many brief
occupations over an extended period and served as a con-
venient stopover for people moving out into the islands
north of the Door Peninsula.

The Oneota Component 227

Works cited:

Freeman, Joan E.

1956 An Analysis of the Point Sauble and Beaumier
Farm Sites. Unpublished Master's thesis, Uni
versity of Wisconsin.

Hall, Robert L.

1962 The Archaeology of Carcajou Point. Madison, the
University of Wisconsin Press.

Henning, Dale R.

1961 "Oneota Ceramics in low &." Journal of the Iowa
Archaeological Society, Vol. XI, no. 2, -pp. 1-47.

Mason, Ronald J.

1966 Two Stratified Sites on the Door Peninsula of
Wisconsin. Museum of Anthropology, The Uni-
versity of Michigan, Anthropological Papers no.
26. Ann Arbor.

1967 "The North Bay Component at the Porte des
Morts Site, Door County, Wisconsin/' The Wis-
consin Archaeologist, Vol. 48, no. 4 n.s. pp.
267-345.

McKern, Will C.

1945 Preliminary Report on the Upper Mississippi
Phase in Wisconsin. Bulletin of the Public Museum
'of the City of Milwaukee, Vol. 16, no. 3.

McPherron, Alan

1967 The Juntunen Site and the Late Woodland Pre-
history of The Upper Great Lakes. Museum of
Anthropology, The University of Michigan, An-
thropological Papers no. 30. Ann Arbor.

228 WISCONSIN ARCHEOLOGIST

Vol.51, No. 4

INTERESTING WISCONSIN SPECIMENS

SANDIA-LIKE POINTS FROM WISCONSIN AND IOWA
From the James Weida collection, Milwaukee. Provenience (left to right): Mil-
waukee or Waukesha Co.; Washington Co.; Milwaukee Co.; Wisconsin; Van
Bur en Co., Iowa.

BOOKS RECEIVED

THE DAVENPORT CONSPIRACY by Marshall McKusick, University of Iowa,
Iowa City, 1970. Price: $5.00, Paper, $3.00.

GODS WITH BRONZE SWORDS by Costa de Loverdo. Doubleday & Co., New
York, 1970. Price: $6.95.

Statement of Ownership 229 X

STATEMENT OF OWNERSHIP, MANAGEMENT AND CIR-
CULATION (Act of October 23, 1962; Section 4369. Title
39 United States Code)

1. Date of filing: October 8, 1970

2. Title of Publication: The Wisconsin Archeologist

3. Frequency of issue: Quarterly

4. Location of known office of publication: Leader Printing
Company, Lake Mills, Wisconsin, .53551

5. Location of the headquarters or general business offices
of the publishers: Milwaukee Public Museum Mil-
waukee, Wisconsin

6. Names and addresses of publisher, editor and managing
editor: The Wisconsin Archeological Society, Milwaukee
Public Museum; Robert Ritzenthaler, Milwaukee Public
Museum

7. Owner: The Wisconsin Archeological Society, Milwaukee
Public Museum

8. Known bond holders, mortagees, and other security
holders owning or holding 1 percent or more of total
amount of bonds, mortgages or other securities: None

9. For completion by nonprofit organizations authorized
to mail at special rates: Have not changed during pre-
ceding 12 months

10. Extent and nature of circulation: Membership

ACTUAL NUMBER OF

N0 - COPIES COPIES OF S1NGLE BSUE

EACH BSUE DURING PUBLISHED NEAREST

PRECEDING 12 MONTHS TO FILING DATE

A. Total No. Copies Printed 700 700

B. Paid Circulation

Mail Subscriptions 495 519

C. Total Paid Circulation 495 519

D. Free Distribution None None

E. Total Distribution 495 519

F. Office Use, Left-Over, Unaccounted,

Spoiled after Printing 205 181

G. Total 700 700

I certify that the statements made by me above are
correct and complete. Robert Ritzenthaler, Editor.

230 WISCONSIN ARCHEOLOGIST

Vol. 51, No. 4

NOTES

' '.- -'

COMMITTEE ASSIGNMENTS
WISCONSIN ARCHEOLOGICAL SOCIETY

MEMBERSHIP COMMITTEE: Phillip H. Wiegand, Chairman.
Herman Zander, J. K. Whaley.

PUBLIC* :"Y: Tom Jackland.

FRAUDULENT ARTIFACTS: J. K. Whaley, Chairman. Martin
Greenwald, Robert Hruska, Neil Ostberg, Dr. Robert Ritz-
enthaler, Paul Scholz, Phillip H. Wiegand.

SURVEYING AND CODIFICATION: Paul Koeppler, Chairman.
Darryl Beland, Elmer Daalmann, Wayne Hazlett, Ernest
Schug.

PRESERVATION OF SITES: Robert Hruska, Chairman. Dr.
Richard Peske, Dr. Robert Ritzenthaler. (Chairman has op-
tion to make further appointments.)

EDITORIAL: Dr. Robert Ritzenthaler, Chairman. Dr. David A.
Baerreis, Dr. Joan Freeman, Dr. Lee Parsons, Dr. Melvin
Fowler.

PROGRAM: Dr. Diehard Peske, Chairman. Paul Turney, Mrs.
P. H. Wiegand, Dr. Robert Ritzenthaler.

LAPHAM RESEARCH MEDAL. Dr. Robert Ritzenthaler,
Chairman, Dr. David A. Baerreis, Phillip *:. Wiegand.

THE CHARLES E. BROWN CHAPTER
Madison

(Meets second Tuesday, 7:45 P. M., State Historical Society,
October thru May)

President: James Stoltman
Secretary: Peter Storck
Treasurer: Lathel Duf field

THE FOX VALLEY CHAPTER

Oshkosh

(Meets second Monday, 7:30 P. M., Oshkosh Public Museum,
September thru May)

President: Elmer Daalmann

Vice President: Richard Mason

Secretary: Robert R. Jones

Treasurer: Nancy L. Hoist

THE WISCONSIN
ARCHEOLO6IST

HILGEN SPRING PARK MOUND GROUP 1

by Howard Van Langen and Thomas F. Kehoe

A GROOVED AXE TYPEOLOGY
by William A. Smith

CONSTITUTION AND BY-LAWS

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee, Wisconsin
Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. M., Milwaukee Public
Museum, September thru May

OFFICERS
PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Ostberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Greenwald.

TREASURER

Wayne J. Hazlett
3768 S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct, Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr./Robert E. Ritzenthaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL

MEMBERSHIP FEES

The Wisconsin Archeologist is distributed to members

as part of their dues.

Annual Members, $5.50

THE WISCONSIN ARCHEOLOGIST

New Series

MILWAUKEE, WISCONSIN - MARCH, 1971
Published Quarterly by The Wisconsin Archeological Society

HILGEN SPRING PARK MOUND GROUP
Howard Van Langen and Thomas F. Kehoe*

The following is a report of the excavation of a mound
group located in the SE 1/4 of the NW 1/4 of Section 35,
T 10 N, R 21 E, in the City of Cedarburg, Ozaukee County,
Wisconsin. The excavation was undertaken in the summer
of 1968 when it was learned that the mounds were to be
leveled in the process of expanding a housing project.

After receiving the kind permission _of the owner, Mr.
James Callan, to salvage the site the work was begun
immediately since the landscaping was scheduled to start
in two weeks. As it turned out, various delays slowed down
the building project and it was finally abandoned completely
after having leveled only the smallest mound of the group.

The mounds were a group of three conicals located on
the sloping crest of a wooded hill located 300 feet above
Cedar Creek. The woods consisted of a good stand of beech
with some birch and hickory. As indicated by the group
name, there were several good springs in the area.

The mounds were oriented in a generally north-south
direction. The largest mound, number one, measured forty
feet in diameter and varied in height from six feet when
measured from the natural slope level, to nine feet where
it extended down into one of several depressions in the area
formed by the removal of material used for mound fill.

Mound number two was oriented twenty-four feet south of
mound number one and measured thirty-five feet in diameter
and four feet in height.

Mound number three was oriented seventy feet south of
mound number two and measured thirty feet in diameter
and three and one-half feet in height above the natural
surface level. A line drawn through the centers of mounds

*The senior author conducted the excavations and is responsible for this
section of the report while the junior author made the analysis of the artifact
material.

2 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

number one and three, finds mound number two oriented
somewhat to the west.

Each of the mounds had been disturbed by relic hunters,
mound number three was extensively disturbed, and mounds
number one and two were pitted at the centers reaching down
to some of the burials. However, the considerable size of
mounds number one and two left much undisturbed area to
be excavated. Because mound number three was extensively
damaged and lay in a position closest to where the building
project was to start, it was not excavated. Excavation was
started on mound number two first, in the hope that it could
be excavated, while the least damaged mound, number one,
lying farthest away from where the building project was to
start, could somehow be spared long enough to permit a
careful excavation.

In light of what was expected at the start of the dig, this
was probably a good plan. However, had the ensuing delays
been anticipated, ample time could have been found for
excavating mound number three first. This would have been
worth the effort as in slicing through the mound the
bulldozer rolled out quite a bit of human skeletal material
including several skulls that were salvaged along with one
large pottery rim sherd. (PI. 1, 1)

Work on mound number two was started by laying out a
grid system and clearing brush and trees from the mound.
Excavation was started by taking a four foot wide cut through
the center of the mound and continued by taking alternating
three foot cuts from each side of the original cut across the
entire diameter of the mound. This method was also followed
in excavating mound number one in order to excavate the
usually more productive centers of both mounds first with the
intention of excavating the rest of the mound completely if
time allowed.

Natural soil stratification in the area outside of the
mounds consists of an average seven inch thick layer of dark
humus soil at the surface. This covered a yellow sub soil
layer that varies from very thin to six inches in thickness.
This is underlain by a layer of red soil averaging twelve
inches in thickness and in turn is underlain by a thin layer
of gravel, and then a deep layer of clear yellow sand. The
very abrupt color changes between these layers of soil made
them easy to recognize and interprete.

In both mounds number one and two, the Indian builders

Hilgen Spring Park Mound Group

Plate 1. Projectile Points: a-c, Knives: d-g, Gorget: h, Rim Sherds: i-1,
Body Sherds: m-v.

removed the top soil layer, before construction of the
mounds was started. This was very evident not only by the
absence of the humus soil, but also by the absence of the

4 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

many stones of various sizes that are found in the top soil
layer in the area around the mounds.

In mound number two, the yellow first sub soil layer
was used as the floor of {he mound. In mound number one,
the first sub soil layer was also removed and the second
red sub soil layer was used as the mound floor. No special
attempt was made by the builders of the mound to keep the
floor level. Instead, the contour of the sub soil color layer
chosen for the floor was followed and not broken except
for burial and fire pits and several small breaks that
possibly represent test pits dug by the builders in their
effort to follow the chosen color contour.

Each of the mounds was used for burial purposes. Skeletal
material salvaged from mound number three represents at

300 TO CBLDBR CREEK

V .

O K>' 20' 30*

Figure 1. Hilgen Spring Park Mound Group

Hilgen Spring Park Mound Group

U 6

2. V BUNDLE BURIALS IN PIT

SMALL BREAK CONTAINING BALL JOINT & SIISHD
S o DOG BUNDLE BURIAL

6 STONE FIRE HEARTH
7. EXTENT OF DARK MIDDEN LATER

A l-o STONE CONSTRUCTIONS
Figure 2. Hilgen Spring Park Mound 1. Plan View

least five adults, quite probably buried as bundle burials.

In both mounds number one and two, the main burials were
placed in pits at the mound centers. In mound number one, the
pit was taken through the red floor of the mound and into the
gravel and sand layers. In mound number two, the pit was
taken through the yellow floor of the mound and into but not
through the red layer. Each of the burial pits was roughly
circular in outline, ran to a depth of about eleven inches and
was from five to six feet in diameter.

In the pit in mound number two, the undisturbed lower
leg and foot bones of an adult, primary burial in a supine
position, was encountered. The rest of the skeleton had been
reached by the relic hunters in their vertical pitting at the

6 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

center of the mound. Apparently they were not interested in
skeletal material as they did not take the time to remove the
fill to get at the leg and foot bones. In sifting through the dug-
over material of relic hunters, many of the bones belonging to
this skeleton were found. This burial was oriented in a
general east-west direction, head to the west, and most
probably represents the only burial in the pit of mound
number two.

Just beyond the outside edge of the burial pit, a one-
foot-in-diameter break in the mound floor was found.
(Fig. 4, #2). In digging into this break, one large piece of
human skull bone and a human sternum were found lying
on the sand and gravel layer. Possibly these bones represent
secondary bundle burial material somehow previously left
unburied and then given proper burial in the mound primarily
built for the single in-the-flesh burial.

feet

1. BUHULIPIT.CONTAINING LOWER LIMBS OF AN EXTENDED BURIAL

2. SMALL BREAK IH FLOOR CONTAINING HUMAN BONE FRAGMENTS

3. EXTENDED INFANT BURIAL
U, INTRUSIVE BUNDLE BURIAL
5. FIRE HEARTH

6 FIRE PIT

? WHITE ASH FIRE REMAINS

8. EXTENT OF DARK MIDDEN LATER

Figure 3. Hilgen Spring Park Mound 2. Plan View

Hilgen Spring Park Mound Group

COMMON SAND & SOIL FILL

IUMDS SOIL & STONE SURFACE

LATER REMOVED } HUMUS SOIL COVER

YELLOW SUBSOIL FLOCR-
RED SUBSOIL

U 'BURIAL PIT CONTAINING LOWER LIMBS OF EXTENDED BURIAL

2. BBEAI IN FLOOR CONTAINING IUMAN BONE FRAGMENTS

3 EXTENDED INFANT BURIAL

it. INTRUSIIE BUNDLE BURIAL

5. FIRE HEARTH

$. FIRE PIT

7. WHITE ASH FIRE REMAINS

8. EXTENT OF DARK MIDDEN LATER

feet

Figure 4. Hilgen Spring Park Mound 2. West Profile

The remains of a child that died at or shortly after birth
was also found in mound number two. (Fig. 4, #3). The
skeleton was oriented in a generally east-west direction
with its head towards the east. The child was buried in a
supine in the flesh position and located to the northwest of
the burial pit and nine inches above the mound floor. Burial
in this position, somewhat removed from the burial pit,
may reflect a lack of status for this burial. However, grave
goods to be described later were found with this burial
indicating some status and it is more likely that the child
died while the mound was being constructed and was included
in the remote position.

Another burial found in mound number two was inserted
sometime after the mound was completed, as an intrusive
bundle burial. (Fig. 4, #4). The burial was placed one foot
below the top of the mound. A rim sherd (PI. 1, k) of pottery
found with this burial is foreign to the type of pottery found
in the undisturbed area of the mound and represents a group
of people other than the builders of the mound.

The burial pit in mound number one was also reached
by relic hunters and some of the burials were disturbed.

8 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

At least three bundle burials were represented in the
undisturbed part of the burial pit in mound number one.
In sifting through the dug over material of the relic hunters,
bones that probably represent another bundle burial were
found. The condition of the bones of the bundle burial were
poor. Also found in sifting through the dug-over material
were bones in a very good state of preservation including
the skull of a round-headed individual (in contrast to all
other skulls found being long-headed). It is very frustrating
to speculate on material worked over by relic hunters. But
this well preserved round-headed skull probably represents
an intrusive burial inserted quite some time after completion
of mound number one. As such, this speculation, while
adding nothing to an attempt to understand these mounds, will
at least not distort the picture obtained from the undisturbed
parts of the mound.

A very small break in the mound floor at the edge of the
burial pit contained one ball joint, one small pottery sherd
and a small piece of charcoal. This again, possibly
represents bundle burial material found by the mound
builders after the burial pit was filled, and inserted with
the pot-sherd, for proper burial.

In both mounds number one and two, the burials, burial
pits and other features, and a considerable area at the
centers of the mounds were covered with a fire blackened
soil layer. It measured as much as 18 inches in thickness
at the center and diminished in thickness towards the
outside. It contained much midden material including many
flint flakes and cores, snail and clam shell, antler and deer
bone and teeth, fragments of human mandible, charred
bone, worked bone, red ochre, potsherds, bits of charcoal,
fire cracked stone, broken implements and polished stone.

The material was packed very compactly around the
burials and there is no doubt that it was intended as a
special stratification covering for the burials, separating
them from the common sand and soil fill used to complete
the mounds. There can be some speculation as to where this
material came from. The midden material found in this layer
is what one would expect to find on the surface of a camp or
village site. The very uniform blackness of the layer
indicates it came from an area subjected to a great amount
of fire action. Such an area could very well be a part of a
camp or village set aside for temporary scaffold burials or

Hilgen Spring Park Mound Group

temporary burials in the earth until such proper time as the
remains were placed in a permanent mound burial. Such a
temporary burial area would no doubt be subject to many
fires and would serve as a source of material suitable for
covering burials in the mound. The proper location chosen
for the mound would not necessarily be close to the habitation
site and could possibly represent accumulated deaths over a
period of several years from different camp sites.

There was a marked absence of any large pieces of
charcoal that would normally be present in such fire strata.
Small tools and containers used to dig am) transport this
material would enable the builders to remove the charcoal
and any other material that they felt should not go into this
burial layer.

It would also follow that material found in this layer if
not specially included would at least not have been rejected
as undesirable and probably represents cultural material
associated with the builders of the mounds:

This black midden filled layer was the only special
stratification in mound number two. Above this layer mound
number two was completed largely with clear yellow sand.
However, mixed materials representing all the top and sub
soil layers was used for mound fill to some extent.

Mound number one above the black midden layer was
made up more completely of the clear yellow sand. This
may have been the special intention of the builders of mound
number one, or it may reflect the fact that the pits formed
by getting fill for mounds number two and three was present
when mound number one was built. Then a good supply of
yellow sand would be available without the need of going
through the top and sub soil layers. In mound number one,
in addition to the dark midden layer, a special golden color
sand layer was used in association with stone constructions.
These special sand stratifications were about three inches
in thickness and were from five to six feet in diameter. They
occurred on top of or inside the dark midden layer.

The five stone constructions found in mound number one
were of several sizes and locations as follows. Construction
number one was made up of mixed types of fieldstones and
limestone of various sizes. A stone much larger than any of
the rest was placed in a point or base position in the
construction. It was a multi-colored beauty that very likely
was specially chosen by the builders of the construction.

10 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

This construction was laid on the midden layer five inches
above the mound floor, on the north east edge of the burial
pit, and measured five and one half feet in length, two feet
in width and one and one half feet in height. The stones
were carefully placed to form a compact well shaped form.
The entire construction was covered with the dark midden
layer.

Construction number two was made up of various types
of field stones and limestone, and as in. construction
number one, an extra large base stone beautifully rough
and bicolored was used. This construction was placed
across the burial pit from construction number one about
five feet back from the opposite edge of the burial pit. A
roughly circular three-inch layer of a golden color sand
was found directly below this construction. A six-inch
thick layer of the dark midden, was below this. Construction
number two was of more loose and not as precise in out
line as construction number one. It appeared that some of
the top stones from this construction may have been toppled
accidently, possibly in the process of covering them with
fill. This construction measured five feet in length, three
feet in greatest width and fifteen inches in height, and was
also covered with the dark midden material.

Construction number three was made up of nine stones
laid in a circular outline two feet in diameter. The stones
were about the size of a football, and consisted of four
limestones and five field stones, one a very bright red and
one of a beautiful green color. It is quite probable that these
stones were not just chosen^ at random. This construction
was placed about eight feet from the outside edge of the
mound to the northeast of the burial pit and six inches above
the mound floor. The construction was placed on a three
inch layer of golden sand and this in turn underlain with a
three inch dark midden layer. The construction was also
covered with the dark midden layer.

Construction number four was made up of a single
layer of various sizes of field stone, sandstone and limestone
five feet long and one and one half feet wide! This
construction was located one foot above the mound floor
and about six feet to the east of the burial pit. The
construction was laid on a five inch layer of dark midden
material. This was underlain by a two inch layer of golden
sand, and this underlain by a five inch layer of dark

Hilgen Spring Park Mound Group 11

midden material. This construction was covered by the
yellow sand fill material.

Construction number five was made up of various sizes
of fieldstones and limestone. An especially large base stone
of rough limestone was present. This construction was about
four feet long, from two to three feet wide, one foot in
height and roughly oval in outline. This construction was
located to the northwest of the burial pit and extended to
within eight feet of the outside of the mound. The construction
was built on and covered with yellow sand j fill and placed
about two feet above the mound floor. The top stones of
this construction extended to about five inches below the
top of the mound.

There can be no doubt that these constructions were an
important part of the ritual connected with the building of
these burial mounds. It appears that all the constructions
were built after the burials in the pit had been placed and
covered with the dark midden material. The various positions
of the stone constructions in relation to the burial pit, in
distance and height, indicates that the ritual did not stop
with the covering of the burials but continued throughout
the building and completion of the mounds.

No one special function can be attributed to all of the
stone constructions. No one of the constructions was
directly associated with fire. The floor of the mound directly
below construction number two was blackened from a fire.
Directly to the east of construction number four, a fire
hearth was encountered lying on the mound floor. The hearth
was outlined with limestones and measured three feet wide
and four feet long. A good amount of charcoal was the only
thing found on this hearth.*

Skull fragments were found between the stones of
construction number two. A human tooth, human finger
bone, one small fragment of bone, several flint flakes and
three snail shells were found among the stones of
construction number five. No material was found directly
among the stones of the other stone constructions.

Construction number two was covered with dark midden

"Charcoal from this hearth was submitted to the radiocarbon laboratory at
University of Wisconsin and dated 460 B.C. (2410155, WIS-354). It is
isted in error as from Mound 2 in Bender, Bryson, and Baerreis 1970, p.337.
The date is considered too early for an Effigy Mound Complex site.

12 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

material and we cannot overlook the possibility that the
skull fragments could have just been included in this
material. However, construction number five was covered
with the yellow sand fill that was largely sterile, pointing
out that the material on construction number five, and
quite possibly that on construction number two, was purposely
deposited by the builders.

A concentration of one hundred and ninety flint flakes
was found in a one foot cubic area in the midden layer in
close association with construction number one. No such
concentration appeared anywhere else in the mounds.
The mound floor near construction number four had an
area of discoloration from some humus material.

Unbroken, well made or crude artifacts, and broken
artifacts, were found in very close and relatively close
association, on a common level with, and above and below
the stone constructions. It is possible that some if not all of
these materials were placed in the mound as grave offerings,
or could just be material included in fill dirt from the borrow
pit. It may seem dubious to suggest that broken artifacts
would be included as grave goods; however, it is suggested
as well as the possibility that the artifacts were purposely
broken before being placed in the mound. Two artifacts,
one broken and one complete (PI. 1, c) were founcl in the
usually sterile sand above construction number four; it
seems very unlikely that these were not purposely placed.
Pottery sherds were found above and below in close
association to the stone constructions and other features of
the mounds primarily in the dark midden layer, but
occasionally in the yellow sand fill, and in areas remote
from any feature. No whole pots were found, usually single
pieces and sometimes several pieces from one pot were
found together. A single rim sherd (PI. l,k) was found with
the bundle burial in mound number two, a single piece of
pottery appeared deep in the ash of a crematory fire, a
single piece with the ball joint buried at the edge of the
burial pit. This would seem to suggest that a single sherd of
pottery may have been significant to the Indian of this time
and that their inclusions in the mound cannot be written off
as accidental and may represent pottery associated with
the builders of the mounds.

Another feature found in mound number one was a dog
burial (Plate 2). The dog was not buried in the flesh but as

Hilgen Spring Park Mound Group

Plate 2. Dog Maxilla: a, Dog Madible: b-c.

a bundle burial. The burial was located in the yellow sand
fill two and one half feet below the surface of the mound
and twelve feet from the outside of the mound. One of the
few potsherds found in the sand fill appeared near the dog
burial. The reason for the inclusion of this dog burial is to
be pondered. Had it been buried in the flesh, its inclusion
as a sacrifice would seem feasible. Buried as a bundle
burial, it would not seem to be valid as a sacrifice. More
probable is that for some reason the dog was attributed
enough status to be included in the remote location in the
mound.

Mound number two did not contain any stone constructions
of the type found in mound number one.

A hearth of fire blackened and ^cracked stones was found
on the floor of mound number two five feet from the outside
edge of the mound. The hearth was made of a single layer
of stones, and was roughly oval in outline measuring two
feet by one and one half feet. This hearth had a good
quantity of charcoal but no other material was found on it.
The hearth was placed at the edge of a four-foot-diameter
fire pit that ran to a depth of one foot below the mound

14 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

floor. The pit was filled with blackened soil and charcoal
from successive fires and contained scattered fire cracked
stone and one small potsherd. The yellow sand fill around
the outside diameter of this pit was blackened from its
fire indicating the fire was burning as fill was brought up
to it. It is possible that this fire area was used through the
period of the building of the mound and covered in a final
completion of the mound.

Across the mound from this hearth and pit and nine
feet from the outside edge of the mound another fire remains
was found. This fire had a precisely circular diameter of
eighteen inches and left a very white ash layer five inches in
thickness. No trace of charcoal was present. The subsoil for
a few inches below the ash was turned a very bright red
color indicating an intense heat from this fire. The ash
contained one potsherd, (PI. 1, p) one small piece of charred
bone, and a few flint flakes. The remains of this fire was
covered with the black midden layer and made a very striking
color contrast between the very precise layers of black, white
and red. It is quite probable that this was a crematory or
sacrificial fire. The piece of charred bone found in the ash
indicates an animal material of some kind was burned but
the consumption of the material was so complete that no
identification could be made.

CULTURAL MATERIAL*

Stone Arti facts -

There was a total of 19 stone artifacts found while
digging the burial mound. Three complete and two broken
projectile points were found in various parts of the mound.
One distinct point type recognized in Wisconsin was present-
A Fox River Valley Stemmed Point.

The complete Fox River Valley Stemmed Point, (PI. l,c)
made of white chert, has a length of 45 mm., width 27.0
mm., and thickness of 8.0 mm. The stem length is 10 mm.
and base width 13.0 mm. It has a trianguloid blade, sloping
shoulders, contracting stem, and irregular fractured base.
There is no evidence of grinding. Flake scars are deep and

*The entire artifact and skeletal material as well as the excavation records
have been donated to the Milwaukee Public Museum.

Hilgen Spring Park Mound Group 15

irregular suggesting a hard hammer percussion method of
manufacture. The basal stem of the same type point was
found in the midden layer of mound 1. It is made of white
chert, and the stem is 15 mm. long and 18 mm. wide, and
thickness is 8 mm. The base is straight to slightly convex,
and is not altered showing the original cortex. A large
flake scar runs parallel with the stem. This type of point
-occurs in Archaic to early Middle Woodland complexes.

A complete Woodland type point (PI. 1, a) 41 mm. long,
18 mm. wide and 4 mm. thick, and made of grey chalcedoney,
was found in the northwest corner of construction 4, of
Mound 1. It is a side-notched point with a flaring convex
base. The broad shallow notches are 6 mm. wide and 4 mm.
deep. The distance between the notches is narrow (7 mm.),
and the base (8 mm.) is narrower than the proximal end
of the blade (13 mm.). There is no grinding on the convex
base, rounded basal edges, and shoulders of the artifact. The
blade is somewhat asymetrical. This point is somewhat like
a Monona Stemmed Point (Baerreis 1953, Fig 1, a) but
closer to a projectile point found at the Naomikong Point
Site (Janzen 1968, PI. XVII, d-m).

One complete unclassified side-notched point (PI. 1, b)
is made of pink chert with a length of 37 mm. width 28 mm.,
and thickness of 6 mm. The base is 20 mm. wide, straight,
with rounded basal edges. The specimen has large round,
deep, side-notches, 5 mm. deep and wide. The distance
between the notches is 11 mm.

An incomplete point base with the break across the
proximal end of the blade is made of tan quartzite. The
base is slightly convex and 17 mm. wide. There is no grinding
on the rounded basal corners. Broad shallow corner-notches
are present giving the base the appearance of an expanding
stem. The distance between the notches is 15 mm.

One end scraper-plane or gouge was found in the midden
layer of Mound 2. It was made from a thick flat ovoid chert
core. One end has been chipped to form a broad convex
edge. This forms a 50 degree angle from the ventral
unworked surface.

Twelve stone artifacts appear to be knife blades. One
(PI. 1, g) has a long and slender, leaf -shaped body with a
straight base, and is made of mottled gray chert. A large
fire spall is removed from one side. A similar but thicker
knife (PI. 1, e) with a straight base has the tip wrenched

16 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

off. Seven knives (PI. 1, f) are small ovate forms. Two are
broken across the blade and the others have the tip wrenched
off. The remaining two are complete. There are two broken
tips from knives, or very large, thick projectile points.
There is also a broken mid-section of a knife (PI. 1, d).

Six cores of gray chert show some evidence of use. Five
flakes show signs of usage. Other reject flake material is
present showing evidence of use. One block flake appears to
be heat treated or has been in a fire.

Metallic hematite or limonite occurs as flat oval
pebbles, dark red in color associated with the infant burial
in Mound 2, and were found just beyond the fingers of the
left hand.

An unfinished or broken three hole rectanguloid gorget
(PI. 1, h) of green chlorite shist was found in Mound 2 in
the center of the collector's spoil dirt material. It is 35
mm. wide and 4.5 mm. thick. It's incomplete greatest length
is 75 mm. Two holes, equally spaced from the sides, are
drilled completely through the gorget and a third is evident
in the break. One is 35 mm. from the end and the other an
additional 24 mm. from the first, and the third 15 mm.
further. The holes were drilled with a solid drill from each
side leaving a cone shaped hour glass-like hole. It is
similar to the Thiensville gorget (Ritzenthaler & Quimby,
1962, Fig. 123).

Ceramic Artifacts-
Seventy potsherds were found scattered in the three
mounds with the largest concentration of 35 sherds associated
with Construction 5 in Mound 1. Seven of the total were rim
sherds. These were described, except for one rim sherd
associated with an intrusive burial, as the Hilgen Spring
Ware. The method of construction is the patch method of
building up a vessel by adding lumps of clay. There is
lamination of the clay with longitudinal splits indicated.
Sherds are extremely friable and easily crumble especially
when wet.

Tempering is of crushed granitic rock. Fine silica grains
and mica occur probably as natural inclusions in the clay.
Temper size ranges from 0.5-2.5 mm. in size and shows
through the exterior and interior paste. The exterior is
moderately smooth and the interior moderately rough to
touch. Hardness ranges from 2.0-2.5, and color reddish

Hilgen Spring Park Mound Group 17

brown 5YR 5/3 to brown 10YR 5/3 on the exterior.

The surface finish, showing no luster, was paddled with
an instrument wrapped with cord 1-2 mm. wide with possible
braiding and an S-twist in the cord (PI. 1, n). One type
of Hilgen Springs pottery has been smoothed over by vertical
wiping with grass when the clay is partially dry
(PI. 1, p-u).

The vessels are completely undecorated with the
exception of two lips of the cord marked rim sherds.
Diagonal notches or flutes are made in damp clay using the
plain edge of a pointed stick 2.5 mm. apart (4 mm. w,ide and
1.5 mm. deep). These are inserted from the interior with
the right hand, slanting slightly toward the left and
terminating 1.5 mm. from the exterior edge of the lip.
The decorated rimsherds are 9 mm. thick. One sherd shows
possible evidence of incising in a "V" pattern over cord
roughening (PI. 1, v).

A minimum of four vessels are represented in the sample
but the sample of sherds is too small to reveal the exact
size and shapes of the vessels. They were probably large
conoidal or possibly flat based vessels with vertical rims.
One base flange is present. The lips were thick (12 mm.),
smooth and flat at right angles to the rim with a 45 degree
bevel on the interior. One variety had a rounded to angular
lip. Another variety (PI. 1, 1) had slight interior and
exterior extruding of the flattened lip plus the interior lip
bevel. Wall thickness ranged from 5.0 to 14 mm. and
averaged 7.0 mm.

There has been no previous type descriptions of this
pottery. It is believed to be of Early Woodland affiliation.*

The rim sherd (PI. 1, k) associated with the intrusive
burial is split longitudinally. Tempering is angular feldspar
and ranged from 2.0 mm. to 8.0 mm. in size and does not
show through the exterior paste. Hardness is 2.5, exterior
color reddish brown 5YR 5/3, and it does not crumble.
Surface texture is moderately smooth on a plain surface.
The lip is extruded 2 mm. on the exterior and decorated
by a scallop 2 mm. deep by thumb pressure on a two strand

*Dr. James B. Griffin of the University of Michigan examined the pottery and
recognized Early Woodland affiliations. Dr. David A. Baerreis of the University
of Wisconsin feels the radiocarbon date is in agreement with the pottery and
other artifacts.

18 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

Z-twist cord. The cord strand is 0.5 mm. wide. A 2 mm.
boss resulting from an interior punctate with a round
ended stick 7.0 mm. in diameter occurs 20 mm. below the
rim. Cultural relationship of this specimen is unknown.

INTERPRETATION

The three conical mounds excavated at the Hilgen Spring
Park present an intriguing situation. If the radiocarbon
date of 460 B.C. (WIS-354, 2410 * 55 b.p.) is valid, Mound
1 is the oldest constructed burial mound in the state of
Wisconsin and dates an Early Woodland mortuary complex.
The dog burial, extended human burial in a central pit,
ceramics believed to be of I?arly Woodland provenience,
and some of the point types give support to this. A. three
hole rectanguloid gorget, metallic hematite flat oval pebbles
associated with a burial, and the Fox River Valley Stemmed
Points suggest relationship with an earlier complex. The
stone constructions or so called "alters," suggest a later
relationship with Effigy Mound Culture. A later type rim
sherd associated with an intrusive burial supports the use
of the mound by later Indians for burial purposes.

The most feasible interpretation is that the radiocatbon
date is valid within the magnitude for plus and minus error
of deviation, and the contents of the mound except for the
intrusive burial material is in near approximation of time
and cultural relationship to this date. However, the mounds
could ^be interpreted as having a series of construction
stages undetected at the time of excavation with material
dating from Red Ochre, Early Woodland, Effigy Mound
Culture, and the unknown late intrusive culture. A third
interpretation could suggest that the radiocarbon date is too
old and the artifact bearing midden is derived fill antedating
the construction of the mound by many centuries and coming
from a site occupied by several cultures (except for the
intrusive mound material).

We hope that the Hilgen Spring Park Burial Mound group
may shed some new light upon but not add additional confusion
to Wisconsin's evasive but intriguing problem of burial
mound complexes.

Hilgen Spring Park Mound Group 19

Baerreis, David A.

1953 "Blackhawk Village Site (Da5) Dane County,
Wisconsin" JOURNAL OF IOWA ARCHEOLO-
GICAL SOCIETY Vol. 2, No. 4, pp. 5-20.

Bender, Margaret M., Reid A. Bryson, and David A. Baerreis

1970 "University of Wisconsin Radiocarbon Dates VII"

RADIOCARBON, Vol. 12, No. 1, pp. 335-345.

Janzen, Donald E.

1968 "The Naomikong Point Site and the Dimensions
of Laurel in the Lake Superior Region." MU-
SEUM OF ANTHROPOLOGY, UNIVERSITY OF
MICHIGAN ANTHROPOLOGICAL PAPER NO. 36

Ritzenthaler, Robert E. and George I. Quimby

1962 "The Red Ochre Culture of the Upper Great Lakes
and Adjacent Areas" CHICAGO NATURAL HIS-
TORY MUSEUM FIELDIANA: ANTHROPOLOGY
Vol. 36, No. 11.

20 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

A GROOVED AXE TYPEOLOGY
William A. Smith
INTRODUCTION

Little has been published concerning grooved axes.
This study is a compilation of the data acquired by the
author in personal communication, through viewing
collections, and by reviewing information available in the
limited amount of printed materials.

"The first incentive to the native in making a grooved
ax was to obtain an implement of practical service, which
could be securely fastened to a handle . . .Otherwise he
would not have used grooved axes at all . . ."(Moorehead,
287). This handle, acting as an extension of the forearm,
increases the force of the blow administered by the ax,
allowing the user to do more work in less time.

The function of the handle in increasing delivered
force is more easily understood by analogy with the modern
steel ax. More wood can be cut in less time if the ax is
attached to a handle than if the ax-head is used without
a handle. This same principle holds true for the hafted
ax and the hand-held chopper. By increasing the distance
from the pivot point to the ax (the striking radius),*both
the striking arc and the delivered force can be increased.

By hafting his chopping tool the user increases his
mechanical efficiency and decreases his expenditure of
muscular energy. However; in order to haft a chopper,
means must be found to secure the handle in such a way
as to minimize slippage of the tool in the handle, slipping
being the prime agent in destroying a haft. Securing is
successfully accomplished by notching or grooving the
chopper in such a way as to form a snug pocket in which
to rest the handle. A notch can be easily chipped or pecked
into a chopping tool. A groove may be formed by pecking
directly into the ax body, by flanging above and below and no
deeper than the body of the ax, or by raising a wide ridge
from the body in which to peck the groove. The occurrence
of chipped choppers notched on opposing sides is very
widespread and most collections contain at least one
specimen. It has not yet been determined if hafting was
parallel to or directly across a handle, the implements being
used as axes or hoes, respectively. If these artifacts were
used as axes, then they may represent a transition from

A Grooved Axe Typeology 21

the hand-held chopper to the hafted ax (Moorehead). This
statement, although providing a logical origin for the ax
tradition, is purely hypothetical.

The full-grooved ax, earliest in the ax tijadition as it
is now recognized, appears in association with pre-pottery
Archaic materials about 1000 to 700 B.C. The three-quarter
grooved ax, developed during the Early to Middle Woodland
period (300 B.C. to 500 A.D.), in many areas entirely
replaced the full-grooved ax, while in other areas the two
types were coeval. Beginning around 700 A.D. a gradual
change from the prevalence of the ax to that of the celt took
place; the grooved ax tradition was entirely displaced in
these areas by 1200 A.D. In areas marginal to the later
cultural development, those areas which still retained the
Woodland tradition, the grooved ax remained in manufacture
although in steadily declining numbers.

In the Southwest, the Mogollon, the Anasazi, and the
Hohokam culture areas, grooved axes appeared relatively
late from 500 to 900 A.D. The ax tradition continued until
the abandonment of the Pueblos and cliff dwellings around
1400 to 1600 A.D.

Just as environmental areas shift gradually from one to
another (as Woodlands and Plains), ax distribution shifts
gradually from common to scarce as one moves from an
area in which axes are useful tools to. an area where they are
not (ie, the Woodlands to the Plains), and from one type
area to another. Lines drawn on a map to show distribution,
while helpful in giving a general idea, tend to be highly
arbitrary and should be viewed in this respect.

Information as to exact distribution of axes is not
precise and for some areas nonexistent. Consequently,
boundaries shown on the map are " educated guesses" wit'h
gaps in those areas where information was not available.

The ax appears in various forms, many of which can be
classified into types; however, even those classes of axes
which have a similar appearance in flanges, vertical grooves,
poll forms, and blade forms also possess individual
characteristics of these parts, which set one specimen
apart from any other. Never are two axes found with
shapes exactly alike. Nor are axes similar in size. The
smallest, showing wear, perfect copies of much larger axes
and weighing only a few ounces, may have been used as
toys or in activities where a larger ax would have been

22 WISCONSIN ARCHEOLOGIST Vol: 52, No. 1

unsuitable. Large specimens, some weighing upward of
twenty pounds, have been reported for many locales (mostly
in Illinois). These, too, show polish on the bit and groove
which indicates use. Most axes, however, fall within a range
of from three to ten pounds; axes much larger and much
smaller are rare.

An ax weighing from five to seven pounds seems to be
the most practical, heavy enough to deliver a forceful blow
but not so heavy as to be cumbersome. If the bit is relatively
sharp, as it is in the majority of specimens, an ax of this
weight can cut wood easily when used in short, quick
strokes, swung from the elbows and not from the shoulders
(pers. comm. G. Metcalf). There is great disagreement,
however, on whether axes were the primary agent in felling
trees. Some authors are of the opinion that axes were used
either to girdle trees (C. Rau) or to bruise and crush
portions of the trunk so that fire could be used more
effectively in felling the tree. The ax was then employed
again to remove the burned and charred wood (A. Skinner,
G. Perkins). Others feel that axes were sufficiently sharp to
cut down trees without the aid of another agent such as
fire (W. Moorehead, C. Brown). Champlain told of Indians,
using stone axes to cut down trees fora stockade, who were
able "in two hours to make so strong a defense that five
hundred men could not break through without great loss."
Fire could not be used in this "lest the smoke reveal their
presence to their enemies" (F. Berlin). Most authors agree
that axes were used in both woodworking and warfare.
Moorehead cites the use of axes in soapstone quarries and
for bear hunting.

SELECTION OF MATERIALS

In the selection of raw materials for a grooved ax,
the prehistoric manufacturer had to apply his knowledge of
geology; he had to know at a glance the equivalent of the
chemical and physical properties of rocks and select the
proper "blank." He knew that certain rocks disintegrated
through hydration (rocks with mica or limestone) and
consequently avoided these.

The manufacturer knew that certain rocks did not lend

A Grooved Axe Typeology 23

themselves to pecking and grinding or to use as axes. Rocks
with coarse crystalline structures (feldspar-quartz
crystalline rocks and most porphyrys), rocks with well
developed bedding planes (gneiss, schist) and most
metamorphic rocks were avoided because they tend to
fracture when struck solidly. Greenstone, a metamorphic
rock, is an exception to this and is the second mosj: common
raw material for axes in the Woodlands. The most common
rocks used for axes are the fine-grained igneous extrusives
(basalt, trap, diabase, etc.). These were selected consistently
for their durability; igneous extrusives are hard enough to
hold a cutting edge (but not so hard as to make pecking and
grinding difficult), and do not fracture easily. This preference
for fine-grained rocks is also manifest in the Southwest in
combination with coarse-grained, very hard specimens.

The skill of the manufacturer and the time available to
him provided other bases for selection of a blank. If he were
skillful or had a greater amount of time to spend, he would
select a harder rock. If the form of the rock before working,
the blank, closely resembled the form of the finished ax,
the problem of spending much time in working down a large,
jagged piece of rock to a recognizable form was much
simplified or the manufacturer could ignore the odd shape of
the blank and produce an oddly shaped ax; the shape of the
blank often determined the shape of the finished ax.

Another time factor is the intended useful lifetime of the
ax. If the ax about to be made was to serve its owner for a
long time, the manufacturer would choose the best material
and apply all of his skills in producing the best ax he could.
However, if the ax was to serve for only one or two activities
and then be discarded, he could employ less suitable
materials, and use less time and skill in manufacture.

As stated before, aboriginal manufacturers showed a
preference for the fine-grained igneous extrusives as raw
materials for axes. Ninety percent or more of all axes
found are of these rocks. The greater part of the remaining
percentage are rocks with greenstone predominating. Axes
of inferior materials are few, two percent or less except
in locales such as the Driftless Area of Wisconsin, where
suitable materials were impossible to acquire. In areas such
as this the percentage is inverted with axes of inferior
materials (sandstone, limestone) predominating and
specimens of igneous material occupying a low percentage.

24 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

MANUFACTURE

After selecting a blank of the proper size, shape and
composition, the manufacturer chose a smaller cobble of
approximately equal or greater hardness, with one end
rather pointed and the other round and comfortable to hold.
The pointed end of the hammer-stone was struck against the
blank, which smashed away a small part of the surface of the
blank (and the hammer-stone) and produced a small pit. The
manufacturer continued to peck away the surface of the blanks
and worked out a groove and a blade. According to his own
cultural tradition and taste, he produced a variety of poll and
blade shapes, a full or a three-quarter groove, flutes,
flanges, barbs or vertical grooves, and did or did not conform
to a given pattern.

When the ax was pecked into the desired shape and size,
it could have been used in its rough pockmarked form or
could be ground and polished. Grinding entails the use of a
suitable stone for abrading the rough surface and produces
a smooth but grainy surface. Again the ax could have been
hafted at this point and put to use or it could be further
polished with sand until smooth and highly reflective.

Pecking, contrary to popular belief, does not require
infinite patience and time. At the U.S. Nation Museum, with
jasper hammerstones for shaping and quartzite for
smoothing a block of nephrite, the hardest rock known, was
worked into a very serviceable grooved ax and polished in
sixty-six hours. A block of granite was worked into an ax
in just two hours (Fowke).

H.L. Skavlem of Janesville, Wisconsin, claimed to be able
to make a crude ax in five to ten minutes, a good grooved ax
in forty-five minutes and a polished ax in an hour to an hour
and a half. No mention was made of materials used.
(C. Brown).

In any given collection the greater number of specimens
will be ground, having a smooth, grainy finish, while the rest
will be about equally divided between those with a rough,
pockmarked finish and those with a polished surface. Many
of the axes exhibiting the rough pockmarked surfaces also
are crude in general workmanship. This, to some, infers
greater age. This assumption is erroneous or at least
difficult to prove; these crude axes were probably
manufactured when emergency situations arose. The

A Grooved Axe Typeology 25

manufacturer, not having sufficient time to select the best
raw materials or to devote to the making of a fine specimen,
made an ax serviceable enough to last through the activity,
and then discarded it. Highly polished or very distinctive
axes, on the other hand, seem to have been cherished and
carried about. Axes of the Keokuk or the Michigan Barbed
types, whose manufacture is very localized, have been found
as much as 250 miles from the local area.

Well-made axes, after they had become dull or slightly
broken, were not discarded. Evidence of resharpening of
dull blades, found in many specimens, is shown by very
short thin blades and large bodies above. Occasionally a
damaged poll was repaired by pecking the rest of the poll
down to appear symmetrical but slightly lopsided.

RAFTING

With the completion of the manufacturing processes
the next step was to attach the ax to a handle. A branch
had to be found which lent itself to the method of hafting
used. The branch had to be green or at least pliable in
order for it to be bent around the groove and form a snug
pocket for the ax to rest in. This pocket could not allow any
slippage; for if the ax slipped in the haft, it would soon
splinter the wood or cut the binding materials. The diameter
of the branch had to be great enough so that a good grip could
be taken, larger in diameter for large axes, smaller with
small axes. Bark and twigs were removed and the branch
scraped down to bare wood. The preceding steps are common
to all methods of hafting; from here on methods differ
greatly.

One hafting method called for a forked stick with forks
of equal diameter and equal length, long enough to pass
through the groove of the ax and allow enough beyond the
front for wrapping. The inside half of the fork was cut
down for flexibility, and the ax was set in between the forks,
one side of the groove resting in the base of the fork. A wet
thong secured the ax to the handle by tightly wrapping the
two ends, the base, and the sides of the fork in a figure-
eight method. When the thong dried it contracted, minimizing
slippage of the ax.

A variation on this method is reported by Skinner. The

26 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

Cree, living around Hudson's Bay, reported that axes were
hafted by splitting a stick of sufficient diameter half way
down, setting the ax in, and binding securely above and below
with deer hide. Both this method and the one described above
worked well for thin axes, but thick axes or axes with square
corners within the groove put too much strain on the handle
and caused splintering and breakage.

For thick axes a branch was suitably prepared by
thinning it in the center, making it flexible, (perhaps by
soaking in water), and bending it around the groove
surrounding the ax tightly on three sides. This handle was
then wrapped in back and in a figure-eight method around
the ax. The handle itself .was also wrapped with hides. Such
a split handle, when the ax is struck solidly, will open
slightly and then snap shut upon the skin of the hand. This

Three Handle Types:

Left: Split Branch; Center: Forked Branch; Right: Bent Branch.

A Grooved Axe Typeology 27

problem is solved by wrapping the handle securely with a
thong and covering it with a sheet of hide.

Since several branches of smaller diameter are more
flexible than one of larger diameter, they can be wrapped
around the ax and secured with rawhide. This method
removes the problems of flexibility encountered in a larger
single branch.

The Keokuk ax type presents its own particular hafting
problem. Grooved on two faces only and having a vertical
groove on each side, this square-cornered ax would split
a wrapped or forked handle. Keyes has proposed that the
use of two withes, one in each groove extending a little in
front and tying of the withes together and covering the
handle following the usual form would solve this problem.

AREAL DISTRIBUTION

The areal distribution of grooved axes includes the
whole Mississippi Valley, Southeast and South central
Canada, and the Southwest. Axes are extremely rare in
Florida, on the Great Plains, and in the Ozarks. Few are
found east from Florida along the Gulf Coast, north from
Texas to southern Iowa, or at the headwaters of the
Columbia, Missouri, and Colorado Rivers. Rare on the
Pacific Coast, axes found in this area are considered by
some authorities to be strays from much farther east
(Moorehead).

Since the grooved ax was primarily a woodworking tool,
one would expect axes to be found where wood was available
in prehistoric times. This is generally true although
''pockets" of axless areas appear within the Woodlands and
in other areas where both ample materials for ax production
and sufficient wood supplies occurred. An example of this
deficiency is found in the Ozarks, where the greater number
of village sites have no axes, and very few sites produce
only occasional specimens (Moorehead). At a village site
elsewhere within the distribution area of the grooved ax,
five to fifteen specimens are usually found with numbers
sometimes as high as thirty.

The continental United States has few areas which are
truly axless, that is, areas in which no ax specimens can be

28 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

found, though they are rare on the Plains or the Rocky
Mountain Plateau and tend to assume odd and varied shapes.
In the midst of the "axless" area occurs an area of relative
commonness and homogeneity of size and shape. This area
of commonness, the Southwest, is separated from other
areas of common occurrence, the Woodlands and Southeast,
by several hundred miles of "axless" area.

Occasional mention in historical reports of axes used
by Plains Indians supports the theory that these migratory
buffalo hunters acquired knowledge of the ax and its uses
from the sedentary peoples in the Mississippi River area.
The idea was probably carried westward by these nomadic
hunters and passed on to the sedentary peoples of the South-
west who readily accepted the ax concepts from the South
and East. Better materials were available here for the
manufacture of axes and a wider variety of uses occurred
for this tool than was the case on the plains. On the
Plains, besides there being a restricted number of uses
and materials, a hafted ax was an additional and bulky tool
to be carried in migration and consequently was found very
infrequently in the Plains tool kit. This theory of the
diffusion of the ax concept is more logical than assuming

A Grooved Axe Typeology

separate invention in two relatively close areas, the
Southwest and the Woodlands.

NOMENCLATURE.

FACE

BACK

"Hi ree- Quarter
Grooved

BACK

Fo I I Grooved

TOP

BACK

.pal

FACE
Mode

can be discerned)

5/OE

SIDE

30 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

RjllTypes

onjedpoll

3"*

ROUND (hem.sphr,coO

-full grooved only

TRUNCATED (lop*y

THAEE QUAHTCft TRUNOSTEO

CONVEX.

r i

A Grooved Axe Typeology

SIDES

HORIZONTAL X STtoN

SECT.ON PttOfl,

Bit

Profile or So^al X Section

RQUHOED

Bottom

32 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

Groove

NARROW V*ST

vertical x Secfon of
Michigan bartMd

Blade

STRAIGHT S4DES

TAPERM6 5TWM&HT SIDES

CURVING SIOCS

ONE SIOC STRAISHr, ONE CuXv

A Grooved Axe Typeology

Blade

horizontal
X sections

S7WU6HT SIDS

Awwr0 END ELLIPSE

CONVEX *05
STRAI6WT fi%CTS

VERr/OU. GROOVE

CONVEX FACES

STTRAIGHT TAPERING

vert/ca/ x

STRAIGHT CUKVINC,

Curving

GROOVED AX TYPES

A. Notched

Poll, Blade, Bit: Forms variable

Groove: Two opposing notches one in each side near

the top or towards the center of the ax.
Distribution: Woodlands Southeast and Southwest
Frequency: Common

34 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

B. Full -Grooved

Poll, Blade, Bit: Forms Variable
Groove: Runs horizontally around body of ax
Distribution: Woodlands, Southeast and Southwest
Frequency: Common

C. Three-Quarter Grooved

Poll, Blade, Bit: Forms variable

Groove: On two faces and the front only, leaving the

back flat or rounded

Distribution: Woodlands, Southeast and Southwest
- Frequency: Common

D. Ovate

These axes are ovate in profile

Poll: usually low and rounded

Blade: contracting

Bit: convex

Groove: Three-quarter or Full

Distribution: Woodlands and Southeast

Frequency: Somewhat uncommon

A Grooved Axe Typeology

E. Flanged

Poll Blade Bit: Forms variable

Groove: Three-quarter or full with a flange above and/ or

below the groove
Distribution: Woodlands
Frequency: Relatively uncommon

Several sub-types exist within the flanged category,
all known from the Woodland area.

1. Vertical Groove: This sub-type always three-
quarter grooved, possesses a shallow depression
running down the backs from the poll to near the bit.

Frequency: uncommon

2. Double Vertical Groove: Also always three-quarter
grooved, these have a shallow depression both on the
back, running from poll to bit, and the front, from just
below the lower flange to the bit.

Frequency: Rare

36 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

3. Four noded: This fully grooved sub-type is char-
acterized by the four ends of the upper and lower
flanges isolated on the back by the intersection of the
ha f ting groove and vertical groove.

Frequency: Rare

F. Wisconsin Long-Bladed

Poll: May be peaked, hemispherical, or low, flat and
angled parallel to the groove, may be fluted as illus-
trated here

Bit: Convex

Groove: Three-quarter, angled, usually flanged

Blade: Long; back: usually straight, front: straight-
curving,

Faces: straight curving, may be fluted
Horizontal cross -^"tion: oval with flat back

Distribution: Found characteristically in Wisconsin in
an area between Green Bay and the Illinois state
line and from Lake Michigan to approximately -100
miles inland

Frequency: Rare

G. Michigan Barbed

Poll: Peaked, sides concave, sides tapering
Horizontal cross -section: ellipsoidal

Groove: Full, wide, very shallow, with four projections,
two on each side, one above and one below the groove.
Some specimens semi-flanged

Blade: Short to long, sides and faces straight- curving
Horizontal cross section: ellipsoidal

Bit: Convex

Distribution: South-central Michigan, found only in
CJinton County and counties bordering thereon

Frequency: Relatively rare within South Central Michi-
gan, extremely rare outside of Michigan

A Grooved Axe Typeology

H. Keokuk

Poll: Medium to high, flat-topped and truncated to round

topped and round- sided; vertical grooves may be

present
Groove: Grooved on two faces only, flanged above and

below
Blade: Medium, faces tapered to slightly curved, sides

straight to tapered, rectangular in cross -section
Bit: Convex ;

Distribution: Limited to Southeastern Iowa, west- central

Illinois and a small portion of northeastern Missouri
Frequency: Relatively rare within distribution area,

extremely rare outside

I. Rectangular Western

Poll: Medium to high, flat- topped and truncated, rec-
tangular in cross section

Groove: Three-quarter, most often angled although it
may be horizontal

Blade: Medium, faces tapered to slightly curved, sides
straight to tapered rectangular in cross -section

Bit: Convex

Distribution: Exact distribution unknown. Specimens
appear to be concentrated within 100 to 150 mile
radius of St. Louis, Missouri

Frequency: Unknown

J. Southeastern

Poll: Medium height, flat-topped, faces and sides

tapering, ellipsoidal in cross -section
Groove: Full, wide, shallow (ie., no deeper than and

continuous with, the body outline) flanges above

and below wide and flat
Blade: Medium length, broad, faces straight curving,

sides straight to expanding, ellipsoidal in cross -

section

Bit: Convex, continuous with body outline
Distribution: Southeast culture area
Frequency: Exact frequency unknown but form not

known outside of southeast U.S.

38 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

K. Southwestern Long-Bladed

Poll: Medium height, round topped, sides and faces

straight- curving, ovate in cross section
Groove: Three-quarter, flanges above and below may

or may not be present
Blade: Long sides and faces tapered- curving, round

cornered rectangle in cross section
Bit: Convex
Distribution: Exact distribution unknown, limited to

Southwest culture area
Frequency: Unknown

L. Southwestern Hammer-ax

Poll: Low to medium, most specimens badly battered,

sides and faces straight oval to rectangular in cross -

section
Groove: Full wide, shallow, usually angled, slightly wider

on lower right side, back deeply notched
Blade: Straight front and back, straight- curving faces

round -cornered rectangle in cross section
Bit: Convex, battered and dull
Distribution: Exact distribution unknown, limited to

Southwest culture area (Arizona and New Mexico)
Frequency: Unknown

M. Pocket axes

Poll, Bit: Forms variable

Groove: Three-quarter or full

Blade: Form variable except for a shallow to deep

depression in face on one side. "Pockets" may be

round to elliptical with rounded to sharp margins
Distribution: Exact distribution unknown, appear to be

concentrated within 100 to 150 mile radius of St.

Louis, Missouri
Frequency: rare

A Grooved Axe Typeology

N. Hematite axes

Poll Blade Bit: axes small in size with contracting bit,

otherwise forms variable
Groove: Three-quarter or full
Axes of this type made of hematite

Distribution: In and around hematite outcrop in Missouri
Frequency: Common around hematite outcrop area,

very rare elsewhere

Double Bit

Poll Blade Bit: Two bits and, consequently no poll;

forms variable
Groove: Full

Distribution: Occur throughout woodlands
Frequency: very rare

P. Double Grooved

Poll, Blade, Bit: Forms variable
Groove: Two, one above the other, full
Distribution: Occur throughout woodlands
Frequency: Very rare

40 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 1

Q. Fluted

Poll Blade Bit: Forms variable

Groove: Three-quarter or full

Axes of this type possess flutes on the poll or blade.

Flutes are found in vertical rows, horizontal rows

or chevrons
Distribution: Restricted to approximately the same area

as the Effigy Mound area of Wisconsin and small

portions of Minnesota and Iowa bordering on S.W.

Wisconsin

(See "The Ringeisen Collection of Fluted Axes" in
THE WISCONSIN ARCHEOLOGIST, Vol. 36; No. 2;
June, 1955.)

BIBLIOGRAPHY

Berlin, Alfred F.

1903 Prehistoric Implements. REPORTS AND COL-
LECTIONS OF THE WYOMING HISTORICAL
SOCIETY, Vol. 26.

Brown, Charles E.

1918 Grooved Stone Axes. THE WISCONSIN ARCH-
EOLOGIST, Vol. 17, No. 1 (OS), Lake Mills.

Fowke, Gerard

1913 Prehistoric Objects. MISSOURI HISTORICAL
SOCIETY BULLETIN No. 1, St. Louis.

Keyes, Charles B.

1931 Grooved Axes of the Keokuk Type. THE WIS-
CONSIN ARCHEOLOGIST, Vol. 10, No. 4, Lake
Mills.

McKern, W.C. and Robert Ritzenthaler

1956 Some Varities of Grooved Axes in the Woodland
Area. THE WISCONSIN ARCHEOLOGIST, Vol.
37, No. 2, Lake Mills.

Moorehead, Warren K.

1910 THE STONE AGE IN NORTH AMERICA, Vol. 1,
New York.

A Grooved Axe Typeology 41

Perkins, George M.

1885 Prehistoric Implements. THE AMERICAN NAT-
URALIST, Vol. 15, Salem.

Pond, Alonso

1930 Primitive Methods of Working Stone-Based on
Experiments of Malvor L. Skalem. LOGAN
MUSEUM BULLETIN, Vol. II, No. 1, Beloit.

Rau, Charles

1876 Archeological Collections of the U.S. National
Museum. SMITHSONIAN CONTRIBUTIONS TO
KNOWLEDGE, Vol. 22, Washington.

Skinner, Alanson

1909 Indians of Manhattan Island and Vicinity. AMER-
ICAN MUSEUM GUIDE, No. 29, New York.

42 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

CONSTITUTION AND BY-LAWS

WISCONSIN ARCHEOLOGICAL SOCIETY
(REVISED AND APPROVED FEBRUARY 15, 1971)

CONSTITUTION

Preamble

A meeting was held the 28th day of February, A.D.
1903 for the purpose of forming an Archeological Society
Corporation under Charter 86 of the revised statutes of
Wisconsin. Henry A. Crosby, Charles E. Brown, Lee R.
Whitney and George H. West were present at this meeting,
and the incorporation papers were prepared over their
signatures in the presence of G.E. Henrick and Arthur
Wenz. The first meeting of the corporation was held
April 3, 1903.

Article I. Name and Location

Section 1

This organization shall be known as the Wisconsin
Archeological Society.

Section 2

Its principal office shall be maintained in the city of
Milwaukee, Wisconsin.

Article II. Object

Section 1

This corporation is organized exclusively for literary,
educational and scientific purposes within the meaning of
Section 501 (c) (3) of the Internal Revenue Code.

Section 2

It will not, as a substantial part of its activities, attempt
to influence legislation, or participate to any extent in a

Constitution and By-Laws 43

political campaign for or against any candidate for public
office.

Section 3

On the dissolution of the corporation, the entire net
assets remaining after the payment of any and all liabilities
and obligations of the corporation shall be distributed
exclusively for the purposes of the corporation in such
manner, or to such organization or organizations organized
and operated exclusively for literary, educational and
scientific purposes as shall at the time qualify as an exempt
organization or organizations under section 501 (c) (3)
of the Internal Revenue Code of 1954 (or the corresponding
provisions of any future United States Internal Revenue
Law.)

Section 4

This organization is incorporated without capitol stock
and is not organized for profit and no part of its net income
shall inure to the benefit of any member, trustee or other
individual.

Article III. Dues

Section 1

The annual dues for regular members is $5.00 if paid
prior to January 1. ($5.50 if paid after January 1). This
fee entitles a member to a full year's subscription to the
WISCONSIN ARCHEOLOGIST.

Section 2

Sustaining membership annual dues are $10.00 Endow-
ment membership annual dues are $500.

Article IV. Application and Acceptance to Membership

Section 1

Application for membership shall be made to the cor-
responding secretary.

Section 2

Each application shall be accompanied by dues for the

44 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

calendar year.

Section 3

Applicant shall be voted upon at a regular meeting.

Section 4

In the event an applicant is rejected the applicant will
be notified and the dues shall be returned.

Article V. Quorum

Section 1

A quorum for a regular or special meeting of the
society shall consist of no less than 15 members.

Section 2

The society may not elect officers or change the
constitution or by-laws unless a quorum is present.

Section 3

A quorum for an Executive Board meeting shall con-
sist of no less than 50% of the board members.

Section 4

The Board may not make a binding decision unless a
quorum of members is polled and unless the decision has a
majority vote.

Section 5

Proxies shall not be voted at any time.

Article VI. Meetings

Section 1

The society shall meet on the third Monday of each
month, except the months of July and August when no
meetings are held. If circumstances warrant, the president
or executive board may set another date. The annual meeting
shall be held in March.

Section 2

The president or executive board may call a special

Constitution and By-Laws 45

meeting of the society at any time.

Section 3

A meeting of the executive board may be called by any
member of the board.

BY-LAWS
Article I. Election and Installation of Officers

Section 1

A nominating committee consisting of three or more
members shall be appointed by the president at the January
meeting of the executive board.

Section 2

The nominating committee shall submit at least one
complete slate of consenting candidates at the February
meeting.

Section 3

Additional officer candidates may be nominated from the
floor at the February meeting of the society.

Section 4

The election shall take place at the annual meeting
in March.

Section 5

The voting shall be by secret ballot where the number
of candidates exceeds the number to be elected.
Article II. Duties of the Officers

Section 1

The president shall preside at all meetings of the society
and the executive board. In his absence the duty shall fall
upon the other officers in the order named in Article IV of
the By-Laws. The president has the power to appoint
committees and shall be an ex-officio member of all

46 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

committees.

Section 2

The duties of the vice-presidents or any one of them shall
be to perform the duties of the president in his absence.

Section 3

The duties of the recording secretary shall be to keep
the minutes of all meetings of the society and the executive
board.

Section 4

The corresponding secretary shall attend to all the
correspondence of the society.

Section 5

The treasurer shall collect and disburse all monies
keeping an accurate account of all such transactions and
monies of the society.

Section 6

The executive board is to conduct the major portion
of the society business and to make recommendations to the
society on matters it cannot pass upon. The executive board
cannot elect officers or change the Constitution and By- Laws.

Article III. Membership and Dues

Section 1

Any person may be elected to membership in this cor-
poration by a majority vote of the members present at any
regular or special meeting thereof.

Section 2

Any person may be elected an honorary or corresponding
Any person may be elected an honorary or corresponding,

life or term, member of the society with all the privileges

of not paying dues.

Section 3

Any member in default in payment of dues shall be ipso
facto suspended from all privileges of membership, and if,

Constitution and By-Laws 47

after notice, such default be not cured within a period of
ninety days, the membership of such member shall auto-
matically cease and terminate.

Section 4

Any member may be removed from membership by a
majority vote of the members present at any regular
meeting or at any special meeting of the members called
for the purpose, for conduct deemed prejudicial to this
corporation, provided, that such member shall have first
been served with written notice of the accusations against
him, and shall have been given an opportunity to produce
his witnesses, if any, and to be heard, at the meeting at
which such vote is taken.

Article IV. Officers

Section 1

The elective officers of the society shall be a president,
five vice-presidents, two directors, a recording secretary, a
corresponding secretary, a treasurer and a maximum of
twenty advisors.

Section 2

The entire slate of elective officers except advisors
shall constitute the executive board.

Article V. Term of Office

Section 1

The term of office shall be one year for all elective
officers.

Section 2

The term of office shall run from election day to
election day at the annual meeting.

Article VI. Amendments

Section 1

The Constitution may be amended at any regular or
special meeting of the society by a two thirds (2/3) ma-

48 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

jority vote, provided a quorum is present, and provided
further that the proposed amendment shall have been
introduced at the preceeding regular meeting of the society.

Section 2

The By-Laws may be amended by a two thirds (2/3)
majority vote at any regular or special meeting of the
society at which a quorum is present.

Article VII. Miscellaneous Laws and Regulations

Section 1

The Society is formed for the mutual benefit of its
members and the general public. Each member therefore,
shall exert a reasonable effort to further the welfare of the
group by obtaining new members, by attending meetings
and faithfully discharging any and all duties assigned to
him.

Section 2

No unauthorized person or persons may incur any
expenses in the name of the society or demand payment for
injury or property loss without the express permission of
the society.

Section 3

The president and secretaries have the standing per-
mission of the society to incur expenses to the extent of
$20.00 for postage, printing, supplies, etc.

Section 4

The Executive Board has standing permission to incur
expenses to the extent of the treasury.

Section 5

A meeting of the executive board need not be a gathering
as such but merely a meeting of the minds through any
commonly recognized corresponding means.

Section 6

The recording secretary shall' call a special meeting
of the society upon the written requests of 15 members and
after notification of the membership.

Constitution and By-Laws 49

ORDER OF BUSINESS

1. Call to order of Executive Board.

2. Reading of the minutes.

3. Reports of officers.

4. Proposals and election to membership.

5. Report of committees.

6. Communications.

7. Unfinished business.

8. New business.

9. Adjournment of Executive Board.

10. Reading of Board minutes to society.

11. Election and installation of officers or any society
business.

12. Program.

13. Adjournment.

50 WISCONSIN ARCHEOLOGIST Vol. 52, No. 1

BOOKS RECEIVED

THE ETRUSCAN SURVIVAL by Christopher Hampton.
Doubleday & Co., New York, 1970. Price $6.95.

THE ARCHAEOLOGY OF MICHIGAN by James E. Fitting.
Natural History Press, New York, 1970. Price $18.95.

STUDIES IN THE HISTORY OF CIVILIZATION by Julian
A. Joffe. Philosophical Library Inc., New York, 1970.
Price $8.95.

-NOTES-

COMMITTEE ASSIGNMENTS
WISCONSIN ARCHEOLOGICAL SOCIETY

MEMBERSHIP COMMITTEE: Phillip H. Wiegand, Chairman.
Herman Zander, J. K. Whaley.

PUBLICITY: Tom Jackland.

FRAUDULENT ARTIFACTS: J. K. Whaley, Chairman. Martin
Greenwald, Robert Hruska, Neil Ostberg, Dr. Robert Ritz-
enthaler, Paul Scholz, Phillip H. Wiegand.

SURVEYING AND CODIFICATION: Paul Koeppler, Chairman.
Darryl Beland, Elmer Daalmann, Wayne Hazlett, Ernest
Schug.

PRESERVATION OF SITES: Robert Hruska, Chairman. Dr.
Richard Peske, Dr. Robert Ritzenthaler. (Chairman has op-
tion to make further appointments.)

EDITORIAL: Dr. Robert Ritzenthaler, Chairman. Dr. David A.
Baerreis, Dr. Joan Freeman, Dr. Lee Parsons, Dr. Melvin
Fowler.

PROGRAM: Dr. Diehard Peske, Chairman. Paul Turney, Mrs.
P. H. Wiegand, Dr. Robert Ritzenthaler.

LAPHAM RESEARCH MEDAL. Dr. Robert Ritzenthaler,
Chairman, Dr. David A. Baerreis, Phillip H. Wiegand.

THE CHARLES E. BROWN CHAPTER

Madison

(Meets second Tuesday, 7:45 P. M., State Historical Society,
October thru May)

President: John R. Halsey
Treasurer: Frederick E. Lange
Secretary: Marjorie Staab

THE FOX VALLEY CHAPTER

Oshkosh

(Meets second Monday, 7:30 P. M., Oshkosh Public Museum,
September thru May)

President: Elmer Daalmann

Vice President: Richard Mason

Secretary: Robert R. Jones

Treasurer: Nancy L. Hoist

HE WISCONSIN

THE LITTLE WOLF SITE
by John R. Halsey

WINNEBAGO

TO THE FOX

by G. Richard Peske

TERWAY

A BURIED SOIL PROFILE AT

THE SITE OF AZTALAN, 47. JE 1,

WISCONSIN

by Manfred E. W. Jaehnig

THE BOOKSHELF

71
78

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee, Wisconsin
Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. M., Milwaukee Public
Museum, September thru May

OFFICERS

PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Ostberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Greenwald.

TREASURER

Wayne J. Hazlett
3768 S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct., Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr. Robert E. Ritzen thaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL

Dr. David Baerreis, Dr. Joan Freeman, Elmer Daalmann,
Robert Hruska, W. O. Noble, R. W. Peterman, Ernest Schug,
Frank Squire, E. K. Petrie, J. K. Whaley, Mrs, Phillip
Wiegand, Paul Koeppler, Tom Jackland, Gale Highsmith,
Robert VanderLeest, Dr. Melvin Fowler, Robert Salzer,
Leo Klecker.

MEMBERSHIP FEES

The Wisconsin Archeologist is distributed to members

as part of their dues.

Annual Members, $5.50

THE WISCONSIN ARCHEOLOGIST

New Series
MILWAUKEE, WISCONSIN - JUNE, 1971

Published Quarterly by The Wisconsin Archeological Society

THE LITTLE WOLF SITE

John R. Halsey
The State Historical Society of Wisconsin

In 1894, Francis J. Doran and his road crew disturbed
a burial of four individuals and their accompanying grave
goods while grading roads in Section 16, T-23-N, R-13-E,
Town of Little Wolf, Waupaca County. There were
approximately 500 artifacts discovered including a copper
axe and the items described below. This find was divided
among the crew members and Mr. Doran. Later, Mr.
Henry P. Hamilton of Two Rivers, Wisconsin, bought
Doran's portion of 34 artifacts for $7. The great number
of artifacts with this burial is unusual. If the estimate of
500 is correct, this would be the largest Red Ochre-like
assemblage known. The diversity of artifact classes and
the occurrence of at least two distinct projectile point
types, make it extremely regrettable that this find was not
kept intact at the time of discovery. While it cannot be
proved, I assume that Mr. Hamilton made attempts to
locate the rest of the artifacts and was unsuccessful. There
is no way of estimating how accurately the artifacts
described below reflect the composition of the original
deposit, since no indication of how or on what basis tine
division was made by its discoverers. The material from
this site and the rest of the Hamilton Collection is in the
collections of the State Historical Society of Wisconsin.

The most unusual specimen in the collection is a
stemmed spearlike object of bone, probably the rib of a
large mammal, e.g. moose or elk (Plate 2, J). In its
present condition it is 186.5 mm. long, 33 mm. wide and
4.5 mm. thick. The point is broken and blunted so that the
artifact was probably longer at one time by at least 5 or
possibly 10 mm. Both surfaces have badly flaked and
scaled although still visible on the bottom or ventral surface
are traces of cross-hatched incising which apparently

52 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 2

/ "*,

i ! I

The Little Wolf Site 53

covered the entire surface. No such incising is present
on the dorsal surface. The stem is quite neatly made and
expands to a convex base. In cross section this piece is
plano-convex.

Two large blades of differently colored Hixton silicified
sandstone (Porter, 1961) are present (Plate 2, H, I). The
largest one is lanceolate in shape and has a very slightly
convex base. It is 192.5 mm. long, 57.5 mm. wide and 10.5
mm. thick. Primary flaking produced medium to large
expanding flakes which meet near the center, but do not form
a ridge. Secondary retouch is confined to the edges. It does
not appear that this was used as a knife, since the edges
are still fairly keen.

The second blade is a bipointed lanceolate. The sides
are less regular in profile than the blade described above.
It is 188 mm. long, 40.5 mm. wide and 10 mm. thick. The
flaking technique is identical to that described above. The
edges are dulled, but whether through use, as a step in
manufacture or through more recent handling cannot be
accurately determined. In contrast to the first blade which
is gray in color, this blade is a light yellowish gray.

Two projectile points are present. One (Plate 1, A)
is a large side-notched point with relatively straight sides
that come to a point. It is 132.5 mm. in length. The other
measurements are in Table 1. The stone is a mottled pink
chert of uneven texture. Primary flaking scars are rather
small and along the edge tend to hinge out at a millimeter
or less up onto the blade. This has created a rather thick,
blunt edge on both sides of the point. The base is very
slightly concave. There is neither basal nor notch grinding.
The point is biconvex in cross section.

The second projectile point (Plate 1, B) is a well-made
expanding stem point of medium size. The stone type is
again a mottled pink chert of the same color as the first
point, however it is even-textured. The blade sides are
symmetrically convex and terminate in nearly right-angle
shoulders. The stem and convex base are heavily ground.
The complete measurements for both of these points and
the large bone point are in Table 1.

Eight small to medium-sized knives or blanks of Hixton
silicified sandstone make up part of the collection (Plate 1,
C-J). These are generally lanceolate in outline and contract
near the bottom to form either a convex or straight base.

54 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 2

-9f*W

The Little Wolf Site

Their metrical attributes are presented in Table 2. The
edges are not sharp. This may be due to use as knives or
as part of the manufacturing process.

The major portion of the collection consists of 21 chert
knives or blanks (Plate 1, K-X; Plate 2 f A-G). They exhibit
the same kind of dulling from use or preparation as noted
for the silicified sandstone specimens. Also like those
above, the shapes are generally lanceolate and contract
near the bottom to a straight or convex base. One specimen
is extremely aberrant in shape, being bipointed with
asymmetric edges (Plate 2, G). Viewed as a group, exclusive
of the aberrant specimen, they are remarkably uniform
in shape, manufacutring technique and amount of wear.
The size range present shows no evidence of a bimodal
distribution. Their metrical attributes are in Table 3. Both
the silicified sandstone and the chert blades bore red
ochre stains.

DISCUSSION

The artifacts described above belong to the divers and
elaborate Late Archaic burial tradition which covered the
whole northeastern part of the North American continent.
Having been unable to view or locate the original specimens
for similar chipped stone caches in Wisconsin, I can only
-iggest that the following finds may be related: the Lyons
Cache (Stout, 1906, p. 257); the Schumacher Cache (Brown,
1907, p. 55); Richland City Cache No. 2 (Brown, 1907,
pp. 55-56); the Fulton Township Cache (Brown, 1907, p.
56); the Tustin Cache (Brown, 1907, p. 56); the Berlin
Cache (Brown, 1907, pp. 57-58); the Honey Creek Cache
(Brown, 1907, pp. 58-59); the Meyer's Mill Cache (Brown,
1907, p. 60); the Reynolds Cache (Falge, 1915, p. 146);
the Lyon Cache (Brown, 1916, p. 93); and the Silliman Burial
(Brown, 1924, p. 73). These typically contain small (50-100
mm.) lanceolate or "leaf-shaped" blades, usually of finished
quality, as opposed to crude "roughed out" blanks.

Similarities may also be noted with a burial from Port
Washington (Quimby, 1957), classified by Ritzenthaler and
Quimby (1962, p. 258) as a component of the Red Ochre
Culture, which contained seven lanceolate quartzite blades
ranging from 59 to 94 mm. in length. It is the large blades,

56 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

however that offer the most intriguing and diagnostic
possibilities for correlation. The Riverside site in
Menominee, Michigan (Hruska, 1967) yielded large blades
of both the straight-based and bipointed varieties although
these are made of dark gray flint.

Recently, in Jackson County, Wisconsin, a cache of 74
quartzite blades was found in the bank of the Black River
(David Kriesel, personal communication). There are two
large straight-based blades (190 and 165 mm. long). The
rest range from 138 to 62 mm. in length. The smaller
blades have straight or convex bases.

Outside Wisconsin, in the Middlesex Phase of Ritchie's
New York sequence, there occur large and small blades of
the types described from the Little Wolf burials (Ritchie,
1969, p. 203).

It is to the south, in Illinois, where the most striking
similarities are found. Cole and Deuel (1937, p. 90, PI. IV)
include large straight-based blades as an integral part of
their definition of Red Ochre Culture.

Perino (1968, pp. 72-73) describes a "Red Ochre" point
type from specimens found in association with turkey-tails
in a small mound in St. Clair County, Illinois. The size range
and the forms presented by the specimens illustrated indicate
that nearly all the blades, both silicified sandstone and
chert, could be encompassed within Perino's description.

I believe it is likely that the "Red Ochre' ' point as
defined and illustrated by Cole and Deuel, Perino and Scully
(1951, p. 9) is a variant of or companion type to the large
"ceremonial" blades from such sites as Carey and Port
Washington in Wisconsin and Oak Grove and Dyer in Indiana
(Ritzenthaler and Quimby, 1962).

The large side-notched projectile point may be a large
representative of the Raddatz side-notched type (Wittry,
1959a, p. 44). The stemmed point most closely resembles
the "Table Rock" points described and illustrated by Perino
(1968, pp. 96-97) and "Bottleneck" points described by
Converse (1963, p. 111). It is also similar to the "side-
notched" point illustrated from the Thiensville site by
Ritzenthaler an Niehoff (1958).

The large bone spear is extremely unusual and I have
been unable to discover anything comparable to it in bone.
There are some chert and quartzite forms, however, that
quite closely approximate it.

The Little Wolf Site 57

CONCLUSIONS

With its estimated 500 artifacts, the Little Wolf burial
is the largest Late Archaic burial find known in Wisconsin
and probably one of the largest in the western Great Lakes.
Although the missing artifacts compromise any major
conclusions, the following seem justified on the basis of the
evidence available. The Little Wolf burial cannot be
accommodated in the Red Ochre Culture AS DEFINED by
Ritzenthaler and Quimby (1962) since it simply lacks too
many of the diagnostic "nuclear" traits (large white
blades, turkey tails, tubular marine shell beads). It is
obviously closely related though and in future syntheses
would probably be included with the more "typical" Red
Ochre sites. It may be contemporaneous, but is more
likely slightly later than most of the definitive Red Ochre
sites. The evidence for chronological placement rests mainly
on the dating of several of the comparative sites mentioned
above. Dates on comparable material from the Riverside
site range from 510 B.C. to A.D. 1. The Middlesex material
dates from 610 B.C. to 352 B.C. Most estimates for the age
of Red Ochre place it between 1000 and 500 B.C. Also in
regard to dating, the association of the rather different
types of projectile points is somewhat surprising since
both types have been interpreted by other writers as being
much earlier than 500 B.C. in the Archaic sequence.

The purpose of presenting this material is to help fill in
the picture of Late Archaic -Early Woodland burial practices
in Wisconsin. Examination of caches, when and wherever
possible, such as those cited as possible relatives of the
Little Wolf material may yield further surprising and
illuminating associations of previously unrelated artifact
types.

Table 1
METRICS OF LITTLE WQLF PROJECTILE POINTS

Cat. No.

Length

Width

Thickness

Minimum Stem
Diameter

Rase to
MSD

Plate

1919.2702

132.5

36.0

12.5

22. 5

i :; . o

I, A

1919.2703

62.5

25.5

6.5

11.0

1 . "l

1,B

1919.1474

186.5

33.0

4.5

14.5

13.0

II, J

58 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 2

Table 2

METRICS OF LITTLE WOLF SMALL S1LICIJ-

IKD SAMJSTOM; ML AUKS

Cat No.

Length

Width

Thickness Plate

1919. 1490

94.5

30. 5

6.5 i,r

1919. 1491

91.5

27.0

8.5 1,13

1919.1492

87.0

28.5

7.5 1,K

1919. 1493

83.5

31.0

6.5 1,F

1919. 1494

75.5

33.5

9.0 l.G

1919. 1497

74.0

26.0

8.0 l.H

1919.1496

57.5

25.2

7.0 1,1

1919. 1495

52.5

28. 5

6.0 1,J

616.

230.5

59.0

77.00

28.81

7.38

15.35

2.72

1.06

.1993

.0944

.1436

Table 3

METRICS OF THE

LITTLE WOLF CHERT BLADES

Cat. No.

Length

Width*

Thickness Plate

1919.1832

103.0

28.0

6.5 1,K

1919.1833

93.5

27.5

6.0 1,L

1919.1834

81.5

28.5

6.5 1,M

1919.1831

79.0

31.5

8.5 1,N

1919. 1835

77.0

30.5

7.5 1,O

1919.1836

75.0

28.5

7.0 1,P

1919. 1829

70.0

28.5

7.0 1,Q

1919. 1837

68.

27.5

6.0 1,R

1919.2706

68.

27.0

6.0 1,S

The Little Wolf Site

1919.

2704

27.

1919.

1844

28.

1919.

1846

. 5

26.

1919.

2705

28.

1919.

1839

28.

1919.

1S42

58.0

29.

1919.

1840

29.

1919.

1838

.tr

26.

1919.

1845

24.

1919.

1841

28.

1919.

1932

21.

1358.

554.

135.5

67.

27.

6.78

13.

.64

1998

0732

.0943

BIBLIOGRAPHY

Brown, Charles E.
1905

Wisconsin Caches. RECORDS OF THE PAST Vol 4

Part 3, pp. 83-95. Washington.

1907

The Implement Caches of the Wisconsin Indians. THE
WISCONSIN ARCHEOLOGIST, Vol. 6, No. 2, pp. 47-
70. Milwaukee

1916

Archaeological History of Milwaukee County. THE WIS-
CONSIN ARCHEOLOGIST, Vol. 15, No. 2, pp. 25-105.
Milwaukee.

1924

Indian Gravel Pit Burials in Wisconsin. THE WISCON-
SIN ARCHEOLOGIST, Vol. 3, No. 3, pp. 65-82. Milwaukee.

60 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

Cole, Fay-Cooper and Thorne Deuel
1937

REDISCOVERING ILLINOIS. The University of Chicago

Press. Chicago.

Converse, Robert N.
1963

Ohio Flint Types. OHIO ARCHAEOLOGIST, Vol. 13, No. 4,

pp. 79-121. Columbus.

Hruska, Robert

1967

The Riverside Site: A Late Archaic Manifestation in
Michigan. THE WISCONSIN ARCHEOLOGIST, Vol. 48,
No. 3, pp. 145-260. Lake Mills.

Perino, Gregory

1968

Guide to the Identification of Certain American Indian
Projectile Points. OKLAHOMA ANTHROPOLOGICAL
SOCIETY, SPECIAL BULLETIN NO. 3. Oklahoma City.

Porter, James

1961

Hixton Silicified Sandstone: A Unique Lithic Material
Used by Prehistoric Cultures. THE WISCONSIN ARCH-
EOLOGIST, Vol. 42, No. 2, pp. 78-85. Lake Mills.

Quimby, George I.

1957

An Old Copper Site ? at Port Washington. THE WISCON-
SIN ARCHEOLOGIST, Vol. 38, No. 1, pp. 1-5. Lake
Mills.

Ritchie, William A.
1969

THE ARCHAEOLOGY OF NEW YORK STATE. The

Natural History Press. Garden City.

Ritzenthaler, Rpbert and Arthur Niehoff

1958

A Red Ochre Burial in Ozaukee County. THE WISCON-
SIN ARCHEOLOGIST, Vol. 39, No. 2, pp. 115-120. Lake

The Little Wolf Site 61

Mills.

Ritzenthaler, Robert and George I. Quimby

1962

The Red Ochre Culture of the Upper Great Lakes and
Adjacent Areas. FIELDIANA: ANTHROPOLOGY, Vol. 36,
No. 11, pp. 243-275. Chicago.

Scully, Edward G.
1951

Some Central Mississippi Valley Projectile Points.

Museum of Anthropology, TT iversity of Michigan. Ann

Arbor. (Mimeographed paper.)

Stout, A.B.
1906

Summary of the Archaeology of Eastern Sauk County.

THE WISCONSIN ARCHEOLOGIST, Vol. 5, No. 2, pp.

230-288. Madison.

Wittry, Warren L.

1959a

The Raddatz Rockshelter. THE WISCONSIN ARCHEO-
LOGIST, Vol. 40, No. 2, pp. 33-69. Lake Mills.

1959b

Archeological Studies of Four Wisconsin Rockshelters.
THE WISCONSIN ARCHEOLOGIST, Vol. 40, No. 4, pp.
137-267. Lake Mills.

62 WISCONSIN ARCHEOLOGIST Vol. 52, *No. 2

WINNEBAGO CULTURAL ADAPTATION
TO THE FOX RIVER WATERWAY

G. Richard Peske
University of Wisconsin-Milwaukee

Within a culture area perspective, American ethnologists
have placed the Winnebago in the Eastern Woodlands Area
(Wissler 1938) or in the Wisconsin Wild Rice Sub-area
(Kroeber 1939). Thus, the Winnebago have been culturally
typed as akin to the Central Algonkians with a cultural
pattern including an exploitive emphasis on hunting, fishing,
and gathering of wild plants with associated seasonal
movement of habitation sites and a concomitant lack of
emphasis on agriculture and a sedentary existence. Such an
interpretation resulted from a study of ethnographical and
ethno-historical data which was accumulated after cultural
disruptions brought about by European contact and the
displacement of many tribes during the 18th and 19th
centuries.

The Winnebago culture as seen from the archaeological
data is quite different and is indicative of a more complex
Mississippian pattern although not as highly developed as
Middle Mississippian culture to the south and southeast.
Although comparisons of Winnebago culture with Middle
Mississippi and Late Woodland are informative and
meaningful, the scope of this paper is limited to Winnebago
culture and its relationship to its local environment.

The archaeological equivalent of historic Winnebago
culture is the Lake Winnebago Focus. During the 1930's
and 1940's, Midwestern archaeologists, and in particular
W.C. McKern (1945), defined the Lake Winnebago Focus
and historically associated this focus with the Winnebago
tribe by relating their respective geographical distributions
and associating historically-known Winnebago villages with
Lake Winnebago Focus archaeological sites.

Although by the early 19th century, the Winnebago were
scattered across most of south and central Wisconsin, all
the way to the Mississippi River, the 17th and 18th century

Paper presented at the 34th Annual Meeting of the Society for American
Archaelogy, Milwaukee, Wisconsin, May 3, 1969.

Winnebago Cultural Adaptation 63

accounts (see Lawson 1906) place their villages in east
central Wisconsin along the central portion of the Fox River
which includes Lakes Winnebago, Butte des Morts, Winne-
conne, and Poygan. It is this geographical region in which
we find the Lake Winnebago Focus. Through the centuries,
the Winnebago 's oral histories have been influenced by
anthropologists and historians. At the turn of this century,
effigy mounds were thought to be a product of prehistoric
Winnebagos and this idea was inculcated into the modern
Winnebago folklore. Also the Winnebago origin myths of the
Red Banks homeland (an area near Green Bay) were
reinforced by the historian's view of Jean Nicolet's 1634
discovery voyage to Wisconsin in which it wa^s thought that
the Winnebagos he met lived at Red Banks. However, the
historical data do not adequately support this hypothesis. On
the contrary, the Green Bay area was inhabited by
Menominees. The connotation of the term "Red Banks" may
refer to the Valders till which is prevalent in the Lake
Winnebago region as well as in parts of Green Bay.

The archaeological Lake Winnebago Focus sites are found
within the Wolf River-Fox River drainage basin which flows
into Green Bay and abuts the Wisconsin River at Portage,
Wisconsin. This river system is primarily within two major
geographical provinces of Wisconsin: 1) the Central Plain
Province and 2) the Eastern Ridges and Lowlands Province.
(Martin 1965). The Central Plain, except where it was
glaciated, is smooth but relatively infertile as a province
for corn agriculture. The Eastern Ridges and Lowlands
province is smooth, low, fertile, the most densely populated
area of the state today, and is the richest agricultural
province of the state. It is noteworthy, that the Lake
Winnebago Focus is limited in its distribution to the more
fertile Eastern Ridges and Lowlands Province.

What are some of the environmental attributes of this
province? The eastern and western borders are delineated
by two north-south oriented limestone cuestas. The western
limestone ridge is the Black River and Magnesian Cuesta
and the eastern ridge is the Niagaran Cuesta. Between these
two topographical features is a lowland called the Rock
River-Lake Winnebago-Green Bay Lowland which was the
main channel of the Green Bay lobe of the Valders glacial
advance. The Valders has left a heavy till bed and glacial
lake deposits over much of the Lake Winnebago Focus region."

64 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

The Fox River, which at one time in history was called
the Riviere des Puans or Winne-bago River, drains
most of the northern portion of the Eastern Ridges and
Lowlands Province. The Upper Fox, which starts near
Portage, Wisconsin and flows into Lake Winnebago, has a
gentle slope of less than 1/2 foot per mile which is the
same grade as the Wolf River from Shawano Lake to its
confluence with the Fox. In contrast to the slow meandering
portions of these rivers, the Lower Fox which traverses
from Lake Winnebago to Green Bay and the Upper Wolf
above Shawano both are rapid rivers with steep grades.
The lower Fox River drops 5 feet per mile and the Upper
Wolf descends at a rate of 10 feet per mile. The Lake
Winnebago Focus sites are crncentrated along the banks
and lakes of the slower portions of these two rivers.
The culturally selected riverine-lacustrine environment
is one in which the water is slow moving, shallow, and silty.
Within these waterways and Lake Winnebago Focus refuse
deposits, there are large quantities of fish, mollusca, and
waterfowl. The most heavily exploited fish species are
sturgeon, catfish, walleye, bowfin and fresh water drum.
Even today, Lake Winnebago, the Upper Fox, and the lower
Wolf are highly productive fishing waters as is indicated by
a year around fishing season, heavy spring spawning runs,
and sturgeon spearing seasons. Ducks and geese cover the
water during the migratory seasons as this waterway is along
one of the major migratory flyways between the South and
Canada. Many early explorers noted the abundance of
waterfowl in this area. Henry Schoolcraft navigated the
lower Fox River in late August of 1825, and commented:
"Why should I relate to you our dull progress through fields
of rice - through intricate channels, and amidst myriads of
ducks and water fowl" (Schoolcraft 1851:223). Schoolcraft
(ibid:222) goes on to state that the lower Fox River "is
serpentine, almost without a parallel; it winds about like
a string that doubles and redoubles, and its channel is
choked with fields of wild rice." Father Marquette's
description of his travels down the Fox River in 1673 is
similar to that of Schoolcraft ii that he says: "The way
is so cut up'lsy marshes and little lakes that it is easy
to go astray, especially as the river is so covered with
wild oats (i.e., wild rice) that one can hardly discover the
channel" (Thwaites 1882:156-7).

Winnebago Cultural Adaptation 65

Both the preceding quotations stress the heavy growth
of wild rice along the lower Fox River. This feature of the
environment stands out in historical descriptions. It was
not until this century that the extensive wild rice beds
were eradicated. Flood control, damming, and the con-
comitant rise and fall of water levels has led to the demise
of wild rice but during the Winnebago occupation of the
area, wild rice flourished and attracted fish and waterfowl.
In addition to its importance in the ecological system of
native flora and fauna, wild rice was probably an important
food source for the prehistoric Winnebago. Unfortunately,
the archaeological work that I have conducted has not
yielded grains of wild rice; however, indirect evidence in
the form of postulated wild rice thrashing pits has come from
Lake Winnebago Focus sites (Jenks 1901; Lawson 1902;.
Thus, the types of rivers associated with prehistoric Winne-
bago settlements are slow, shallow, meandering rivers which
are silty and frequently broaden out into shallow lakes and
within this habitat, the associated flora and fauna were
abundant and comprised a concentrated food source for
man.

In terms of the growing season, the Lake Winnebago
Focus sites are within the range of 140 to 160 consecutive
frost-free days which has been used (Yarnell 1964:126-137)
as a minimum number of frost-free days for productive
corn agriculture. Archaeological and topographical data
reflect a heavy exploitation of corn agriculture by prehistoric
Indians in south and central Wisconsin (Peske 1966; Fox
i959). Within this region numerous prehistoric agricultural
fields have been reported and the greatest concentration of
garden beds is in the Lake Winnebago Focus area. In an
earlier paper, these ridge and furrow systems have been
associated with the Winnebagos (Peske 1966). Extensive
fields are and were encountered around Winnebago villages'.
In association with the villages are numerous indicators of
corn agriculture which are corn kernels, corn cobs, corn
pollen, scapula hoes, and the cleared agricultural fields
which occur in the forested river flood plains and lowlands
which rise above the river flats.

The land vegetation of Wisconsin and the Lake Winne-
bago Focus region has been reconstructed by John T. Curtis
(1959). The floristic provinces and various plant communities
which he maps geographically are based on contemporary

66 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

observable plant distributions and reconstructions of local
floras from land survey notes of the middle 19th century
When Wisconsin was surveyed for township and section
plats, the government surveyors described the types of local
vegetation and frequently used trees, which were identified
to species, as markers. Thus, Curtis compiled the copious
data of plant types and distributions from the land survey
records and was able to reconstruct floristic zones for a time
period prior to extensive alteration of vegetation brought
about by encroachments of civilization with its emphasis on
lumbering, farming, and later, urban development and
industrialization.

The Lake Winnebago Focus sites fall within a floristic
tension zone which is comprised of a great variety of plants
and associated animals. This tension zone, which is 30 to
50 miles wide, marks the border of major floristic provinces
(Curtis 1959), major biotic provinces (Dice 1943) and major
air masses (Bryson and Wendland 1967). At least 182 plant
species, which represent different major floristic provinces,
reach their range limits in the tension zone. This means
that there is a vast variety of life within a narrow spatial
area. In terms of ecological communities, the Winnebagos
were in easy reach of prairies, oak openings, southern
hardwoods, northern hardwoods, and conifer stands in
addition to the riverine-lacustrine environment. They had
the advantage of being able to exploit a wide range of
habitats. The refuse remains consist of animal species
which are indicative of numerous environments; however,
faunal identifications show that nearly seventy percent of
the species at one site (Lasley's Point) are either aquatic
animals or animals closely^ associated with this habitat
(Cleland 1966: 85-87). At Lasley's Point 176 birds of 181
individuals were waterfowl.

Deer and elk total 78% of the pounds of usable meat of
mammals and they represent 54 of the 141 individual
mammals. The tension zone was the most heavily populated
deer area of the state prior to the establishment of second-
growth timber and brush in northern Wisconsin and the
white population pressure in the Eastern Ridges and Lowlands
Province. Besides the forest and forest-edge dwelling
mammals, bison are present on Lake Winnebago Focus
sites. The reconstruction of the bison range in Wisconsin
(Schorger 1937) places bison within central Wisconsin so

Winnebago Cultural Adaptation 67

that there would have been easy access to bison; however,
the excavated bison bones are all worked scapulas which
could have been received through trade with western
Wisconsin Orr Focus tribes like the loway and Oto. A good
percentage of the ceramic sample from Lasley's Point are
Orr Focus sherds. Late Woodland sherds are notably absent
from the excavated sample. Thus, I postulate stronger social
ties with Orr Focus social units and Chiwere Siouan
speakers than with nearby Algonkian tribes.

The prehistoric Winnebago settlement pattern is one of
sedentary villages along waterways and having an adjacent
cemetery and surrounding agricultural fields. They are
located in regard to immediate access to water for
drinking, fishing, wild rice gathering, shellfish collecting,
hunting, and travel. The fields extend a mile or more back
from the villages and are interspersed with rock heaps
derived from Yield preparation. When the fields are far
from the village, the stone heaps primarily contain cultivation
by-products like broken hoes and large rock. When the stone
heaps are closer to the habitation sites, they also include
village refuse debris which has been carried to the dumping
area. The cemeteries contain numerous single, extended or
semi-flexed interments which frequently have ceramic
vessels and shell spoons as burial goods.

Some of the historic and/or prehistoric village sites are
Doty Island, Karow, Furman, Eulrich, McCauley, Pipe, and
Lasley's Point. All of these sites are on Lake Winnebago,
the upper Fox River, or the lower Wolf River. Many of
these sites are located at strategic travel junctions. Doty
Island is situated at the mouth of the Fox River where it
empties out of Lake Winnebago, McCauley is at the mouth
of the Fox River where it empties into Lake Winnebago at
Oshkosh, and Lasley's Point is at the junction of the Fox
and Wolf rivers. The water travel routes linked the
Winnebago to the northern Pre cambrian Shield via the Wolf
River, to the Great Lakes via the lower Fox River and
Green Bay, and to the Mississippi River via the upper Fox
River and Wisconsin River.

The abundance of copper refuse reflects northern
economic ties; lake trout bones indicate interaction with
tribes along Lake Michigan; and catlinite, Orr Focus
sherds, and bison bone are indicative of interaction to the
west. Further evidence for the importance of the Green Bay-

68 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

Fox-Wisconsin-Mississippi water route for travel is the
history of French, English, and American exploration in
the Upper Great Lakes where this waterway was the focal
line of travel.

The anthropologists' treatment of the Winnebago is to
place them in a marginal culture area and/or on the northern
margins of Mississippian expansion. Perhaps we should look
at their ecological setting as being well centered for the
exploitation of a wide variety of environments and a
diversity of Indian tribes. Their cultural setting is analagous
to that of the Hurons in that they were a central cog for
distribution of goods between tribes over a vast area. No
wonder Jean Nicolet planned a voyage to set up trading
relations with the Winnebago rather than the Menominees
or other Algonkian Indians in Wisconsin. He had heard of
the dominant tribe of Wisconsin: the Winnebago.

BIBLIOGRAPHY

Bryson, Reid A. and Wayne M. Wendland

1967

Tentative Climatic Patterns for some Late Glacial
and Post-Glacial Episodes in Central North America.
In Mayer-Oakes, William J. ed., LIFE, LAND AND
WATER, pp. 271-298. Univ. of Manitoba Press. Win-
nipeg.

Cleland, Charles E.

1966

The Prehistoric Animal Ecology and Ethnozoology of the
Upper Great Lakes Region, ANTHROPOLOGICAL PA-
PERS, MUSEUM OF ANTHROPOLOGY, UNIVERSITY
OF MICHIGAN, No. 29. Ann Arbor.

Curtis, John T.

1959

THE VEGETATION OF WISCONSIN. University of Wis-
consin Press. Madison.

Dice, Lee R.

1943

THE BIOTIC PROVINCES OF NORTH AMERICA. Uni-
versity of Michigan Press. Ann Arbor.

Winnebago Cultural Adaptation 69

Fox, George R.

1959

The Prehistoric Garden Beds of Wisconsin and Michigan
and the Fox Indians. THE WISCONSIN ARCHEOLOGIST,
N.S., Vol. 40, No. 1, pp. 1-18.

Jenks, Albert E.

1901

The Wild Rice Gatherers of the Upper Lakes. BUREAU
OF AMERICAN ETHNOLOGY, 19th ANNUAL REPORT.

Kroeber, Alfred

1939

Cultural and Natural Areas of Native North America.
UNIV. OF CALIFORNIA PUBL. IN AMERICAN ARCH-
AEOLOGY AND ETHNOLOGY, Vol. 38.

Lawson, Publius V.
1902

Clam Eaters and Their Shell Heaps in Winnebago County.

THE WISCONSIN ARCHEOLOGIST, O.S., Vol. 2, No. 1,

pp. 6-8.

1906

The Habitat of the Winnebago, 1632-1832. PROC. OF THE
STATE HISTORICAL SOCIETY OF WISCONSIN, pp. 144-
166.

Martin, Lawrence
1965

THE PHYSICAL GEOGRAPHY OF WISCONSIN. Univ. of

Wisconsin Press. Madison.

McKern, W.C.

1945

Preliminary Report on the Upper Mississippi Phase in
Wisconsin. BULL. OF THE PUBLIC MUSEUM OF THE
CITY OF MILWAUKEE, Vol. 16, No. 3, pp. 109-285.

Peske, G. Richard

1966

Oneota Settlement Patterns and Agricultural Patterns in
Winnebago County. THE WISCONSIN ARCHEOLOGIST,

70 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

N.S., Vol. 47, No. 4, pp. 188-195.

Schoolcraft, Henry Rowe
1851

NARRATIVE JOURNAL OF TRAVELS.

Schorger, A.W.
1937

The Range of the Bison in Wisconsin. TRANS. OF THE

WISC. ACAD. OF SCIENCE, ARTS, AND LETTERS,

Vol. XXX, pp. 117-130.
Thwaites, Reuben
1882

HISTORIC WATERWAYS.

Wissler, Clark
1938

THE AMERICAN INDIAN, 3rd ed. Oxford Univ. Press.

Yarnell, Richard Asa

1964

Aboriginal Relationships Between Culture and Plant Life
in the Upper Great Lakes Region. ANTHROP. PAPERS,
MUSEUM OF ANTHROPOLOGY, UNIV. OF MICHIGAN,
No. 23. Ann Arbor.

A Buried Soil Profile 71

A BURIED SOIL PROFILE AT THE SITE OF AZTALAN,
47, JE 1, WISCONSIN

Manfred E.W. Jaehnig
Central Washington State College

During the summer of 1967 a field crew under tne
direction of Dr. Joan E. Freeman and Mr. Jay Brandon of
the State Historical Society of Wisconsin, and funded by the
Wisconsin Conservation Commission, Department of Natural
Resources, State of Wisconsin, undertook to excavate several
confined areas of the site of Aztalan (47 JE 1), Town of Az-
talan, Jefferson County, Wisconsin. One of the areas
excavated was a very low platform mound located in the
northeastern one third of the palisaded area. The mound,
referred to as Mound #3, is so low that it cannot be distin-
guished very readily from the natural topography. It was,
however, located on the early maps drawn by Judge Hyer
and Dr. Lapham (see Barrett, 1933: Plates 4, 5, and 6), and
had been relocated by Dr. Freeman and Mr. Brandon during
the 1964 excavations at the site.

DESCRIPTION AND CLASSIFICATION

Buried under the above discussed Mound #3 was a soil
which, since it had lain virtually unchanged since the
construction of the mound, could serve as an indicator of
microhabitational conditions of the area under the mound
before the latter was built. The soil was exposed in the
north profile wall of a trench dug in an east- west direction
during the 1967 excavations. The following description is
based on the Soil Description Form filled out at the site
by Mr. Joseph A. Steingraeber, soil scientist of the
U.S.D.A. Soil Conservation Service. Only the buried soil is
of interest here, and the weakly formed soil of the mound
fill will not be described.

The buried soil is classified as a Kibbie silt loam.
Present vegetation on top of the mound is a hay crop of red
clover and quack grass. The parent material consists of
eolian silt over either glacial outwash or river deposited
banded silts and fine sands. The physiography consists of
a river terrace or a lacustrine basin between the Crawfish
River and a glacial moraine. Relief is gently sloping at 3

72 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

percent, with an eastern aspect. The Kibble silt loam is
described as a somewhat poorly drained soil, and the ground-
water table stood at 9 feet on July 27, 1967. Root disturbance
consists of elm tree roots to a depth of 54 inches. There
has been no erosion, rather, deposition has occurred at this
part of the site. The permeability of the soil profile is
moderate in the upper 34 inches (from the present surface),
and moderately slow from 34 to 54 inches.

A technical description of the buried soil follows (refer
also to Figure 1):
Alb 24-29"

Very dark grey brown (10YR3/2M) silt loam; medium
moderate subangular blocky; friable; pH 6.3; abrupt,
wavy boundary.
A2b 29-34"

Grey brown (10YR5/2M) silt loam; medium weak platy;
friable; pH 6.3; clear, wavy boundary.
A3b 34-36"

Grey brown (10YR5/2M) loam; medium weak subangular
blocky; friable; pH 6.0; clear, wavy boundary.
B21tb 36-45"

Brown (10YR4/3M) silty clay loam; medium strong
subangular blocky; firm; pH 6.0; gradual, wavy boundary.
B22tb 45-54"

Olive brown (2.5Y4/4M) silty clay loam with many very
dark grey brown (10YR3/2M) clay skins and organic
coatings; medium strong subangular blocky; firm; pH
6.5; gradual, wavy boundary.
B23tb 54-60"

Dark yellowish brown (10YR4/4/M) silty clay loam with
very dark brown clay skins and organic coatings and
yellowish brown (10YR5/6M) common fine distinct mot-
tles; medium moderate subangular blocky; firm; pH
6.5; gradual, wavy boundary.
IIC1 60-72"

Yellowish brown (10YR5/4M) fine sand; single grain;
loose; effervescent; abrupt, smooth boundary.
IIC2 72-?"

Brown (10YR5/3M) laminated silts and fine sands;
single grain; effervescent.

In order to classify the buried soil profile it is desirable
to use two systems of soil classification: the 1949 system

A Buried Soil Profile 73

and the 7th Approximation. Although the former is supposedly
superceded by the 7th Approximation, the latter has not yet
been universally accepted.

Classifying the buried soil according to the 1949 system
would put it into the Order of the Zonal soils, and the Sub-
order of the light-colored soils of the timber region. In
the Great Soil Group it is assigned to the Gray-Brown-
Podzolic soils, and it is the Kibbie silt loam Type of soil.

In the 7th Approximation system of classification the
buried soil belongs to the Order of the Alfisols for the fol-
lowing reasons: it has an argillic horizon; it is usually
moist between 7 and 20 inches, it has a base saturation of
more than 35 percent at 72 inches below the soil surface;
it does not have a spodic or oxic horizon overlying an argillic
horizon; nor has it pJinthite forming a continuous phase
within 12 inches of the soil horizon; and it does not have
surface and upper sub-horizons meeting the requirements
for a mollic epipedon.

The soil belongs to the Sub-order of the Udalfs because
it has a mean summer soil temperature of 59 degrees
Fahrenheit or more at 20 inches below the surface; it is
usually moist between 7 and 20 inches and lacks a calcic
horizon; it lacks the characteristics of an Aqualf, and it
lacks an ablic horizon.

The buried soil has to be classified in the Great Group
of the Hapludalfs because it fulfills all the requirements of a
normal Udalf. It belongs to the sub-group of the Aquollic
Hapludalfs because it has mottles in the B23tb horizon, is
slightly gleyed in the B22tb horizon, and the Alb horizon is
thicker than the typical horizon of a Hapludalf. The Family
designation is fine silty, mixed mesic; and the soil is a
member of the Kibbie Series. The full classification of the
buried soil under Mound #3 at the site of Aztalan is Aquollic
Hapludalf, fine silty, mixed, non-acid, mesic: Kibbie silt
loam.

INTERPRETATION

An investigation of the literature leads to the conclusion

that the buried soil at Aztalan developed under a vegetation

cover of a broadleaf deciduous forest of the mid-latitude
regions (Baldwin, 1927: 282; Hole, n.d.). Since the soil is
somewhat poorly drained it can be stated that the deciduous

forest was a mesic forest, and that sugar maples were the

74 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

predominant species of trees present (Curtis, 1959: 104).
Climatic conditions under which the Gray-Brown Podzolics,
or Alfisols, form are given by Baldwin (op. cit.) as a climate
with a mean annual temperature of 45-50 degrees, with a
summer temperature of about 70 degrees Fahrenheit and a
mean winter temperature of 25-30 degrees; with about
120 to 180 average number of frost-free days per year; an
annual precipitation ranging from 30 to 45 inches, rather
uniformly distributed throughout the year; and with a
relatively high humidity of 70-90 percent most of the time.

The buried soil, however, is not a typical Gray-Brown
Podzolic, or Alfisol, because the Al horizon is 5 inches
thick rather than the 2 inches defined by Baldwin (1927:297)
and described by Hole and others. For an interpretation
of the thick Al horizon another soil, which also occurs in
southern Wisconsin and which has a dark thick Al horizon,
is of importance. This is the Udoll, or Brunizem in the
1949 classification, which develops under a prairie grass
vegetation and has an Al horizon 7 inches or more in
thickness.

The thickness of the buried soil lies between that of an
Alfisol and that of an Udoll, and therefore is interpreted
to be an intergrade between the two soils. According to
studies done by White, Everett and Riecken (1955:504-506),
the Aztalan soil is a typical intergrade soil, more closely
related to the Alfisols than the Udolls. Because of its well
developed A2 horizon and strongly formed structural units
of the B horizons indications are that the buried soil was
formed under a deciduous forest and that a prairie was
invading the area now covered by Mound #3 just prior to the
construction of the mound.

Arnold and Riecken (1964:353) report that gray ped
coatings are found in the lower horizons of Brunizem
(Udoll) soils in areas where the prairie was invading the
deciduous forests but that these coatings were absent in
areas where forests were invading prairies. The buried
soil at Aztalan exhibited very dark brown (10YR3/2M)
clay skins and organic coatings on the ped faces in the B22tb
and B23tb horizons, and an inspection of Arnold and
Riecken's Figure 3 (1964:358) suggests that the horizon
sequence of the Aztalan soil, including the thickened Al
horizon and the grainy gray ped coatings, can be found at
the very edge of the deciduous forest at which the prairie

A Buried Soil Profile 75

is beginning to get a foothold.

The well-developed solum of the Gray-Brown Podzolic
soil at Aztalan reaches a depth of 45 inches. This develop-
ment would have taken approximately 4,500 years (Arnold
and Riecken, 1964:355). It would not, however, have been
necessary for a forest to have covered the area for the
total span of that time. Since it would have taken about 2,075
years to form the A1-A2 horizon sequence (Parsons,
Scholtes, and Riecken, 1962:494), forest vegetation covered
the area for at least that length of time. A thickening of
the Al horizon from 2 to 5 inches by an invading prairie,
moreover, would have taken approximately 90 years
(Simonsen, 1959:154-155).

CONCLUSIONS

The buried soil located under Mound #3 at the site of
Aztalan, Wisconsin, acted as a storehouse of data facilitating
the reconstruction of the microhabitat at that particular
area of the site. As the above discussion indicated, a
deciduous forest covered the area for at least 2,000 years
without interruption, and for about 4,500 years before the
construction of the mound, although the area might have
been invaded by a grassland between the earlier and the
later date. About one hundred years before the earthwork
was constructed a prairie type of vegetation gained a foot-
hold at the spot, but this recession of the forest was
probably due to man's activities rather than climatic
factors.

As was mentioned above, only conditions at the spot at
which the soil profile was located are reflected in the
paleosol, and the reconstructed microhabitat cannot be
projected over the whole site. There are, however, other
buried soils still present at Aztalan. These are located
under parts of the reconstructed truncated pyramidal
mounds, and although these soils were disregarded during
the original excavations of Mounds #1 and #2, it is hoped
that any future excavations at these mounds will facilitate
microhabitational reconstructions at the western margins of
the site, which would then indicate whether the forests
covered the complete area of the site, or whether Barrett's
original postulation of an open prairie (Barrett, 1933:37-
38) at the site before the aboriginal occupation is valid.

76 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 2

Bll

BI2
Alb
A2b
A3b

B2l+b

B22fb
B23+b
ICC I

HC2

SOIL DEVELOPED
IN MOUND FILL

PALEOSOL DEVELOPED
BEFORE MOUND
CONSTRUCTION

UNDEVELOPED
PARENT MATERIAL

FIGURE!. SOIL PROFILE LOCATED UNDER
MOUND 3 AT AZTALAN

BIBLIOGRAPHY

Arnold, R.W., and F.F. Riecken
1964

Grainy Gray Fed Coatings in Brunizem Soils. IOWA

ACAD. SCI. PROC., Vol. 71, pp. 353-358.

Baldwin, M.

1927

The Gray-Brown Podzolic Soils of the Eastern United
States. PROC. 1st INTERNAT. CONG. OF SOIL SCI,
Vol. 4, pp. 282-297.

Barrett, S.A.
1933

BULLETIN OF THE PUBLIC MUSEUM OF THE CITY

OF MILWAUKEE, Vol. 13.

A Buried Soil Profile 77

Curtis, John T.

1959

THE VEGETATION OF WISCONSIN. University of Wis-
consin Press. Madison.

Hole, Francis D.

n.d.

Soils of Deciduous and Mixed Forests of Mid-Latitude
Regions and Mediterranean Scrub Forest Regions. In
Manuscript.

Parsons, R.B., W.H. Scholtes, and F.F. Riecken

1962

Soils of Indian Mounds in Northeastern Iowa as Bench-
marks of Soil Genesis. SOIL SCI. SOC. AM. PROC.
Vol. 26, pp. 491-496.

Simonsen, Roy W.
1959

Outline of a Generalized Theory of Soil Genesis. SOIL

SCI. AM. PROC., Vol. 23, pp. 152-156.

White, C., M. Everett and F.F. Riecken
1955

Brunizem-Gray-Brown Podzolic Soil Biosequence. SOIL

SCI. AM. PROC., Vol. 19, pp. 504-506.

78 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

THE BOOKSHELF

EXPLORATIONS INTO CAHOKIA ARCHAEOLOGY, Melvin
L. Fowler, Ed. Illinois Archaeological Survey Bulletin No. 7.
Sixty-four figures, 10 tables, 175 pages. Urbana, 1969. $3.

Across the Mississippi River from St. Louis in southern
Illinois i a fertile section of the Mississippi floodplain
kaown as the American Bottoms is a group of over 100
prehistoric mounds collectively referred to as the Cahokia
site. This is without doubt the largest and surely one of the
most important aboriginal sites anywhere in the Americas
north of Mexico. Dominating the site, and forming the
nucleus of tiw Cahokia Mounds State Park, is Monk's Mound,
tbe largest wan-made structure in North America. So large
is this structurebasal area of about 15 acres and a height
above ie surrounding plain of some 100 feet- -that its
artificial character was disputed well into the Twentieth
Century. Naturally, such an impressive cluster of mounds
has attracted amateur and professional archaeologists alike.
Somewhat anomalously, and to the regret of everyone
seriously iaterested in the archaeology of the Eastern Wood-
lands, the volume of literature on the site has not been
commensurate with the volume of earth that has been
excavated. As a result, detailed knowledge of Cahokia
archaeology has been an esoteric commodity, largely
passed by unpublished papers and word of mouth among
those actively engaged in research in or near the site.
The report wider review is a commendable effort on the part
of a group of Cahokia-area researchers to make available
to a wider audience at least some of the results of their
current endeavors in advance of the final reports which are
presently caught up in the lengthy preparatory process for
publication.

The volume consists of eight papers by nine different
authors currently engaged in research on Cahokia data
along with a Preface by the editor and a separate, com-
prehensive bibliography of the archaeological literature
pertaining to the site. Five of the papers grew out of a
symposium on Cahokia held at the 1968 annual meeting of
the Society of American Archaeology in Sante Fe, while
two had been published earlier in rather obscure sources
and are reprinted here with some revision. The eighth

The Bookshelf 79

paper, by Cutler and Blake on corn from the site, was
neither part of the Sante Fe symposium nor has it appeared
in an earlier publication. All papers are billed as "but
preliminary reports," and thus we shall evaluate them on
this basis.

The opening paper, by the editor, sets the stage by
describing the site's limits and major features and then
summarizes the history of archaeological research at the
site. A brief concluding section of this paper presents
some remarks on Fowler's own recent research, which has
included the preparation of a detailed map of the entire site
and excavations into the spectacularly rich Mound 72.
The selection and excavation of this latter mound was a
direct outgrowth of the mapping project during which it
was surmised that the distribution and orientation of many
of the mounds followed a non-random pattern. Mound 72
was selected for excavation because of what appeared to be
its strategic location along the principal north-south and
east-west axes of the site's central mound group. An
especially interesting discovery in this mound was a post
mold one meter in diameter and 2.5 meters deep at pre-
cisely the point where the north-south axis of the site
intersected the mound. Although the true meaning of this
feature may never be known, its precise location relative
to various geometric alignments on the site indicates that
considerable care and planning must have gone into the laying
out of the site by the aborigines.

The second paper is a brief report by Nelson A. Reed
of his survey of the literature concerning Eastern temple
mound sites, Cahokia among them, in search of recurrent
patterns of intra-site mound distributions and alignments.
He examined the literature on 54 sites containing 131
mounds, but found no obvious, recurrent patterns. Although
detailed documentation is not provided, Reed offers a
number of tentative conclusions. Among them are the
observations that when plazas are present, principal mounds
tend to be situated on the west sides and that the orientation
of Mississippian mound groups tends to be dictated more
by the local terrain than by the cardinal directions or
celestial phenomena.

The next two papers, one by Warren Wittry, the other
by Harriet M. Smith, are somewhat revised versions of
works formerly published in the NEWS LETTER of the

80 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

Cranbrook Institute of Science and in the JOURNAL OF
THE ILLINOIS STATE ARCHAEOLOGICAL SOCIETY, res-
pectively. Wittry's brief paper describes one of the four
known circular post hole alignments at Cahokia uncovered
during the 1960, 1961, and 1963 seasons. The immensity
of these circlesabout 410 feet in diameter- -makes it
highly unlikely that they were roofed structures. While
Wittry admits we may never know for sure what their
function was, he hypothesizes from the alignments between
a central post (located within 5 feet of the circle's true
center) and various peripheral posts that these circles
served as solar observatories. The astronomical and
mathematical data required to critically evaluate Wittry's
hypothesis are largely omitted from the paper, but the
more we learn of this immensely complex site the less
unlikely such hypotheses so'uid. Smith's paper, by contrast,
is much lengthier and is mainly descriptive in character.
The 1941 excavations in the Murdock Mound (Mound 55
on Moorehead's map of 1928) are the subject of this
paper, and the focus is almost completely on the mound
itself and its internal features, i.e. stratigraphy and
architectural remains. The lowest levels of the mound
were composted of at least 5 stages of village refuse that
began with a Woodland occupation and ended with a fully
Mississippian component after apparently passing through a
developmental sequence. A six stage platform, with what
Smith feels were elite-class residences on top, follows
the village sequence and is in turn overlain by a temple
mound. An impressive sequence of superimposed houses
beginning with the earliest Woodland occupation are des-
cribed and illustrated in admirable detail and constitute
the major contribution of the report. Unfortunately, there
is virtually no description of artifacts recovered in the
excavations and none is illustrated. For whatever reason,
during the ceramic analysis, provenience data were not
recorded for individual sherds so that the great potential
this site seems to hold for the Mississippian origins
problem has so far gone unrealized.

The fifth paper, by James Anderson, reports on "A
Cahokia Palisade Sequence." In the early stages of their
current Cahokia research project, Fowler and his associates
carefully studied aerial photographs of the site, some of them
dating as far back as 1922, in order to gain a broader

The Bookshelf 81

perspective of the site. Several "white lines" observed in
these photographs could not be attributed to natural
phenomena and were thus suspected of being man-made,
probably stockades because of their straightness and
regularity. Test excavations were conducted in 1966, 1967,
and 1968 (Anderson supervised the first two seasons' work)
along over 120 meters of these presumed stockade lines
and demonstrated conclusively that, indeed, that is exactly
what they were. In places evidence for as many as three
different stockade lines, each with bastions, was found. The
earliest has been dated to Mississippian times ca. A. D. 1100.
Here is one of the few, and certainly one of the nicest,
examples of how aerial photography can further the interests
of archaeological research in the Eastern Woodlands.
The sixth paper, by Patricia J. O'Brien, is concerned
primarily with establishing synchronisms between the
ceramic sequence of the Powell Tract at Cahokia .and the
Caddoan and Lower Mississippi Valley regions. A six-
fold sequence of periods based on the Powell Tract ceramics,
principally on the presence, absence, or stylistic changes
of the type Ramey Incised, is the framework for discussion.
The position of presumed trade sherds within this frame-
work establishes the synchronisms. Since O'Brien's periods
do not conform to the widely accepted, though perhaps
over-simplified, sequence of Mississippian phases for
Cahokia (Pre-Old Village, Old Village, and Trappist), it
is unfortunate that documentation for her proposed sequence
is omitted. It seems to this reviewer, as Walter Taylor
justly emphasized over 20 years ago, that within-site
problems should be tackled before one embarks upon cross-
cultural comparisons. Thus, before inter-regional compari-
sons were attempted, I should think a reconciliation between
the traditional and the newly-proposed ceramic sequences
would be in order. This is especially so when one notes
that the proposed sequence in some ways drastically
contradicts the old and was defined on the basis of a meth-
odology involving the presence or absence of traits rather
than statistical fluctuations in their occurrence, a practice
many would regard as a regression. We are told, for
example, that the second youngest of O'Brien's six periods
(V) is characterized by, among other traits, a high incidence
of limestone tempered pottery and the occurrence of the
enigmatic "stumpware." Yet, both of these have been

82 WISCONSIN ARCHEOLOGIST Vol. 52, No. 2

traditionally considered characteristic of the earliest not the
later part of the Mississippian sequence at Cahokia. Who is
one to believe? When such fundamental beliefs are called
into question it seems to i me that early publication of con-
clusions without ample supportive data is more apt to
generate heat than light.

The contribution of Hugh Cutler and Leonard Blake to
this volume is found in the seventh paper. In it they describe,
mainly in two tables that comprise 9 of the paper's 15
pages, corn cobs excavated from various sites in the
American Bottoms, throughout the wider Midwest, and into
parts of the Southeast. Their basic conclusion is that,
"Cahokia corn, then, follows the general pattern for this
region" (p. 134), i.e. a gradual shift from small, many-
rowed ears to larger, harder, 8-rowed ears. Further, "There
is no evidence of sudden acceptance of new kinds of corn,"
a point of relevance to the age-old problem of local develop-
ment versus intrusion of the Mississippian tradition.

The final paper, by James W. Porter, reports briefly
on his excavations at the Mitchell site, a mound group
situated within the American Bottoms about 6 miles north
of what he refers to as "downtown" Cahokia. Porter's
main purpose, however, is not descriptive, but is to offer
the hypothesis that "a market system developed in and
around Cahokia by A.D. 1100." (p. 159). Such an exchange
system as contrasted with those based on reciprocity or on
redistribution is presumed by Porter to be the highest rung
of a 3-stage evolutionary ladder and comparable, in many
ways, to that of Mesoamerican high cultures. In developing
his hypothesis he relies heavily on wide-ranging ethno-
graphic analogies and on the work of economists and
economic anthropologists, especially Karl Polanyi. The
archaeological evidence Porter marshalls in support of his
hypothesis boils down mainly to the immense size and
complexity of Cahokia, from which it is reasoned a market
system of exchange must logically have existed. He also
feels the site served as a "port of trade" to which and from
which long-range trade caravans moved. These ideas all
seem plausible enough, but can hardly be considered
compelling in light of the existing evidence. What are needed
now are a series of carefully formulated hypotheses to test
these propositions against the Cahokia-area archaeological
data via problem -oriented field research.

The Bookshelf 83

Overall, this monograph must be considered a welcome
addition to the literature of Eastern archaeology, but, in
all honesty, that statement smacks a little of the sentiment
that something is better than nothing. The strengths of the
report lie in its illustrations and descriptions of the total site
and of various important constructional features on it - con-
sidering the maps, photographs, and diagrams, it now rates
as THE source to consult for an overall appreciation of this
immensely important site. However, it is definitely not the
place to go for an appreciation of the kinds of artifacts
found at the site, nor does it define or critically evaluate
the various assemblages that have been traditionally rec-
ognized at the site. To read intelligibly at least parts of
virtually every paper in the monograph, a prior familiarity
with the literature on the site is thus required. The report,
then would seem to be pitched largely toward those with a
serious interest in Illinois archaeology. However, the
glimpses it affords into the research now being conducted
at Cahokia will certainly serve to whet the appetite of an
even wider audience for the more comprehensive reports
that will eventually follow.

James B. Stoltman,
University of Wisconsin- -Madison

It is with deep regret that we announce the death of one
of our oldest and most faithful members, Mr. Paul Scholz,
who passed away on April 25, 1971, at the age of seventy
six.

Mr. Scholz had been a member of the Wisconsin Arch-
eological Society since 1932 and he was present at most
of our meetings.

He will be greatly missed and fondly remembered.

BOOKS RECEIVED

EASTER ISLAND by John Dos Passes. Doubleday & Co.,
N.Y. 1971. Price: $6.95.

THE DEADLY FEAST OF LIFE by Donald E. Carr.
Doubleday & Co., N.Y. Price: $7.95.

-NOTES-

COMMITTEE ASSIGNMENTS
WISCONSIN ARCHEOLOGICAL SOCIETY

MEMBERSHIP COMMITTEE: Phillip H. Wiegand, Chairman.
J. K. Whaley.

PUBLICITY: Tom Jackland, Chairman, Wayne Hazlett, Gale
Highsmith.

FRAUDULENT ARTIFACTS: J. K. Whaley, Chairman. Martin
Greenwald, Robert Hruska, Neil Ostberg, Dr. Robert Ritz-
enthaler, Paul Scholz, Phillip H. Wiegand.

SURVEYING AND CODIFICATION: Paul Koeppler, Chairman.
Darryl Beland, Elmer Daalmann Wayne Hazlett, Ernest
Schug.

PRESERVATION OF SITES: Robert Hruska, Chairman. Dr.
Richard Peske, Dr. Robert Ritzenthaler.

EDITORIAL: Dr. Robert Ritzenthaler, Chairman. Dr. David A.
Baerreis, Dr. Joan Freeman, Dr. Melvin Fowler, Dr. Richard
Peske.

PROGRAM: Thomas Kehoe, Chairman. Paul Turney, Dr. Robert
Ritzenthaler.

LAPHAM RESEARCH MEDAL. Dr. Robert Ritzenthaler,
Chairman, Dr. David A. Baerreis, Phillip H. Wiegand.

THE CHARLES E. BROWN CHAPTER
Madison

(Meets second Tuesday, 7:45 P. M., State Historical Society,
October thru May)

President: John R. Halsey
Treasurer: Frederick E. Lange
Secretary: Marjorie Staab

WIJCONSIN
OLOGIfT

THE BORNICK SITE: A GRAND RIVER

PHASE ONEOTlA SITE IN MARQUETTE

COUNTY

by Guy Gibbon

A CACHE FROM JACKSON COUNTY
by David Kriesel

THE BOOKSHELF

1971 ARCHEOLOGICAL SHOW AWARD
WINNERS

138

139
145

WISCONSIN ARCHEOLOGICAL SOCIETY

Milwaukee/ \Visconsin
Incorporated, 1903

For the purpose of advancing the study and preservation of
Wisconsin Indian Antiquities

Meets Third Monday of Month, 8 P. M., Milwaukee Public
Museum, September thru May

OFFICERS

PRESIDENT

G. Richard Peske

VICE - PRESIDENTS

Neil Ostberg, Paul Scholz, Allen Prill, Thomas Kehoe
Martin Greenwald.

TREASURER

Wayne J. Hazlett
3768 S. 89th St., Milwaukee, Wis. 53228

SECRETARY

Mrs. Edward Flaherty, Corresponding
N62-W15127 Tepee Ct., Menomonee Falls, Wis. 53051
Harry Brown, Recording

EDITOR

Dr. Robert E. Ritzen thaler

DIRECTORS

Paul Turney, Phillip Wiegand

ADVISORY COUNCIL

MEMBERSHIP FEES

The Wisconsin Archeologist is distributed to members

as part of their dues.

Annual Members, $5.50

THE WISCONSIN ARCHEOLOGIST

New Scries
MILWAUKEE, WISCONSIN - SEPTEMBER, 1971

Published Quarterly by The Wisconsin Archeological Society

THE BORNICK SITE: A GRAND RIVER PHASE
ONEOTA SITE IN MARQUETTE COUNTY

Guy Gibbon
University of Illinois-Urbana

The Bornick site was excavated by the writer and a
smattering of volunteers during a six day session in late
August, 1967. The excavation of this site was part of a
larger research strategy concentrating on the stockaded
Walker-Hooper type-site of the Grand River phase. A
single component Grand River phase 'control' site was
sought, because the Walker-Hooper site was multicomponent
and the association of some kinds of artifacts (e.g. wedges,
bipolar cores) with the phase might be questioned. The
Bornick site was selected as the 'control' site.

Although the site is a single component occupation with
more than half of the cultural debris in the plow zone, the
brevity of the excavation period dictated a concentration
on in situ items. A grid pattern of five foot squares was
established and six inch arbitrary levels were maintained
below the seven inch plow zone. All of the dirt was screened
and a large number of flotation samples were collected for
later processing in the laboratory.

The writer wishes to thank Mr. Emil Bornick for per-
mission to excavate the site, Mr. Raymond Flaherty for
locating the site, and Messrs. R. Rhey, F. Jaehnig, and D.
Shea and their wives for participating in the actual ex-
cavation.

SITE SETTING AND DESCRIPTION

The Bornick site (47MQ65) is a previously unexcavated
camp site approximately one acre in extent in the N.E. 1/4
of the S.W. 1/4 of the S.E. 1/4 of Section 1, Montello
Township, Marquette County, Wisconsin. This small site is
jn the north-eastern fringe of the Grand River Marsh
(Map 1), eight and one-half miles from the large, stockaded
Grand River phase Walker-Hooper site. It was discovered

86 WISCONSIN ARCHEOLOGIST

Vol. 52, No.

^^ ^

Grand Marsh

Map. 1. Location of the Bornick Site in Relation to Marshland in Monte
Township, Marquette County (after Cook 1851).

The Bornick Site

and reported to the writer by Mr. Raymond Flaherty of.
Montello and named after the present owner, Mr. Emil
Bornick.

Mr. Ina Cook, who surveyed Montello Township in
December, 1851, described Section 1 as gently rolling and
typified by second rate soils covered with white, bur, and
black oak (Cook 1851). The site itself is two -fifths of a

WHITE LAKE

E I. 797

Map 2. Location and Topographic Setting of the Bornick Site.

88 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

mile south of White Lake in a flat field disrupted in areas
by natural pits or "kettles" (Map 2). A spring in a shallow
ravine thirty yards east of the site may have served as a
source of water. The soil is a slightly acid, excessively
drained, brownisn loamy sand of the Oshtemo-Gotham
complex. All midden samples recorded a pH of 8.(K
The site area has been plowed for many years and no
surface features remain visible except for a thin scattering
of potsherds and food debris. A series of at least five conical
mounds are situated 300 feet from the site, but only one of
them has escaped the plow. A survey of this plowed field in
September, 1967, produced no cultural material.

HABITATION FEATURES

Habitation features consist of five scattered postmolds,
sheet midden, and five cache or refuse pits. A brown B

BORNICK SITE
47-MQ-65

1-3 tot-

Wisconsin

H| - BASIN SHAKO PIT
B - UNDER CUT PIT
- POST MOLD

M.N.

Figure 1. Excavated Area of the Bornick Site.

The Bornick Site 89

horizon obscured the definition of both postmolds and pits
until the lighter C horizon was reached. Postmold stains
were 0.5-0.6 feet in diameter and 0.5 to 1.0 feet in depth
after discovery one foot below the plow zone (Figure 1).
The sheet midden was represented by irregular splotches
of light gray clayish soil rich in refuse materials. A large
number of aquatic snails were present in the midden, sug-
gesting that a combination of decaying organic material and
clay brought from a near-by source was responsible for its
texture. A similar pattern was found at the Walker-Hooper
site (Gibbon 1969). Remnants of sheet midden were found in
all units.

Four basin-shaped pits (1, 2, 3, and 5 in Figure 1) and
one undercut pit (4) were clustered together in the north
end of the area excavated. Clay "caps" or zones were not
associated with any of the pits. The vertical profiles of the
undercut pit and one of the basin-shaped pits are illus-
trated in Figure 2. The tops of all features were truncated
by the plow, but the neck of the undercut pit seemed to rise
vertically without a pronounced flare. Dimensions of the
five pits are entered in Table 1.

Figure 2. Pit Profile Examples from the Bornick Site
(A Basin -shaped, Feature 5; B Undercut, Feature 4).

90 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3
Table 1. Dimensions of Pits from the Bo r nick Site

pit depth encountered diameters in feet depth from orifice

below plow zone orifice . neck body* to base in feet

1 0.4' 1.7-1.5 - 1.1

2 0.5' 3.0 - - 2.4

3 O.y 2.0 - - 1.4

4 0.2' 3O 3.1 4.8 4.0

5 0.6' 2.4 - - 1.4

ARTIFACTUAL ASSEMBLAGE

The chipped stone debitage represents a non-blade flak
industry with bipolar and non-bipolar flaking techniques.
A small number of artifacts of chipped stone (25), ground
stone (20), metal (9), bone (7), and shell (8) were shaped into
tools or ornaments. The frequency and percent of classes
of stone debitage are listed in Table 2. Potsherds frorr
globular shell tempered jars were the most common item
recovered from the Bornick site (6587 sherds). Other objects
of clay include five artifacts, eleven miniature vessels, and
a small amount of clay waste. No grit tempered or Woodland
sherds were found.

Table 2. Frequency and Percent of Classes of Chipped
Stone Debitage

category . frequency percent

decortication flakes 21 3.1

shatter 38 5.6

bipolar flakes 11 1.6

bipolar cores 6 0.9

non- bipolar cores 3 0.4

utilized flakes 44 6.5

other unused flakes 555 81.9

(chips) "(32)

total 6?8 100.0

total weight * 858.2 gms.

(* decortication flakes excluded)

A. Tools for Food Procurement

1. HUNTING TOOLS. Hunting tools consist of nineteen
stone and two antler projectile points. Stone points are the
most numerous formal category of chipped stone artifact
at the site and occur in two basic forms.

The Bornick Site

Plate 1. Unnotched Triangular Projectile Points (A), Notched Projectile Point
(B), and End Scrapers.

a) Unnotched Triangular Points (Plate 1). Fifteen
unnotched triangular projectile points were found in situ and
four in the plow zone. Provenience, maximum dimensions,
extent of marginal retouch, configuration of edges and
presence or absence of serrated sides are tabulated in Table
3. Ten points are chert (1, 2, 7-13, and 19 in Table 3) and
nine quartzite (3-6, and 14-18). Most of the broken points

isist of tips (7, 8, 10 and 19). Nine are isosceles and five
are skewed or scalene.

b) Notched Point (Plate 1).^A single corner-removed

92 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

expanded- stem point was discovered in the midden in level 1
of unit 7. Attributes of this points are: haft width-15 mm.,
basal width-18 mm., shoulder width-25 mm., total length-49
mm., haft length-13 mm., thickness-9 mm., and weight-
10 gms. A similar specimen was found in an apparent
Middle Woodland context at the Silver Creek II site (Hurley
1966:218).

Plate 2. Bone and Antler Artifacts from the Bornick Site.

Two antler projectile points have hollow cores and
smooth carved surfaces with ridges around their centers
(Plate 2, b). Their proveniences and lengths are: Feature
3, level 2, 22 mm.; unit 17, level 1, 35 mm.

The Bornick Site 93

Table 3.

, Attributes of Unnotched Triangular Projectile Points
dimensions in ran. retouch configuration

CO
CU

a - straight
b - convex

-P

H
CO

9
CU

H
CU

g
V

c
o

'O
H
CO

sides

base

H
L,

-P

errat(

-P

0.8

0.6

0.5

(11)

(18)

0.6

(20)

(17)

0.7

(20)

1.4

0.6

0.9

0.5

1.8

1.1

- -

= attribute not present; X = attribute present; - = absent data;
() = approximate value

2. FISHING IMPLEMENTS. Fishing implements consist
of six small shell lures (Plate 5, a) and one flat ground
stone net sinker (Plate 4, E). Two lures came from the
plow zone and four from undisturbed deposits: Feature 3,
level 3; unit 19, level 1 (three lures). The net sinker found
in the plow zone has two opposing notches on the sides and
maximum dimensions of 95 x 83 x 17 mm. There are no other
signs of intentional modification.

94 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

Plate 5. Shell Artifacts from the Bondek Site.

Plate 4.

A. Sandstone Abrader (Feature 4)

B. Iron Acretions (Feature 4; surface)

C. Grinding Stone (Feature 4)

D. Paint Stone (?) (unit 14, level 1)

E. Net Sinker (surface)

F. Grinding Stone (unit 4, level 1)

G. Paint Stone (?) (surface)

The Bornick Site 95

B. Fabricating and Processing Tools

1. STONE PROCESSING TOOLS AND DEBIT AGE. Low
quality chert of light to dark gray was the most common
stone material in the chipped stone assemblage. Quartzite
was the second most common stone, approximately fifteen
percent, while quartz and other stone occurred in amounts
less than one percent of the total chipped stone. Table 2
contains the frequency and percent of classes of chipped
stone debitage. The distribution and frequency of shatter,
decor tication flakes, and the weight of debitage by excava-
tion unit are listed in Table 32. Shatter consists of angular
and cubical chunks of chert lacking bulbs of percussion and
signs of deliberate modification. It is apparently a product
of the initial percussion used to shape raw material (Bin-
ford and Quimby 1963:286-287). Decortication flakes are
unworked flakes having unmodified cortex covering one face.
Only 0.15% of the stone debitage consists of individual
pieces weighing 10 gms. or more. The largest fragments
are two cortex chunks weighing 136 gms. (Feature 4) and
134 gms. (level 2 of unit 19). All other chunks weigh between
10 and 17 gms. Small secondary remnants appear in the
form of cortex on about eleven percent of the debitage.

a) HAMMERSTONES. Six irregular stones have signs
of battering and hammering on their surface. Table 4 con-
tains their provenience, weight, and maximum dimensions.

b) ANTLER FLAKER. A detached deer tine with a
cut base was found in Feature 3, level 3.

c) BIPOLAR CORES AND FLAKES (Plate 3, F). Six
cores are tentatively identified as representative of the
bipolar technique. Three are polyhedral in shape (Feature
3, level 2; unit 4, level 2; unit 6, level 1) and three are flat
(unit 11-, level 1; unit 14, level 1; unit 19, level 1). The
maximum length of these cores in order are 32 mm.,
24 mm., 18 mm., 23 mm., 25 mm., and 17 mm. Table 32
contains the distribution of eleven bipolar flakes. The
range in length of the bipolar flakes is 16 to 38 mm., with
a mean length of 25.7 mm.

d) NON-BIPOLAR CORES. Three small wasted cores
of gray-white chert were found at the site. The provenience
a attributes of these cores are: Feature 3, level 3
(2<i x 18 x 12 mm., 0.06 kgms.), unit 10, level 1 (24 x
22 x 18 mm., 0.05 kgms.), plow zone (28 x 22 x 18 mm.,

96 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

Plate 3. A-E: Side and End Retouched Flakes; F: Bipolar Flakes; G: Chipped
Stone Knives; H: Perforator; I: Notched Flake; J: Copper Beads.

The Bornick Site 97

Table 4. Provenience and Attributes of Hamper stones

n unit/ .. . weight maximum dimensions in mm.

feature in gms length width thickness

1 F4 - 242 67 54 52

2 F5 2 42? 77 75 59

3 2 1 54 42 37 30

4 10 1 41 41 33 26

5 13 1 202 66 55 40

6 14 1 275 67 62 39

0.09 kgms.). The size of some cortex chunks indicates that
relatively large cores may have been initially shaped.
e) UNUSED FLAKES AND CHIPS. All remaining
unused flakes with the exception of chips have been sep-
arated on the basis of material composition and placement
of striking platform and/or bulb of percussion if present.
These flakes range in weight from 0.5 to 17.0 gms., but with
the exception of only a few pieces weight less than 10 gms.
The distribution of these subcategories is recorded in Table
32.

(1.) unworked chert (total: 350)

(a) end struck total: 97

(b) side struck total: 27

(a) end struck total: 51

(b) side struck total: 32

(a) absent total: 2
(distribution: unit 17, level 1; unit 19, level 1)

The frequency and distribution of chips is probably a
function of the amount of flotation undertaken for any par-
< 'cular unit. The distribution of chips by material composition
and placement of striking platform and/or bulb of percus-
sion if present is tabulated in Table 32.

98 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

(1.) chert chips (total: 29)

(a) end struck total: 5 .

(b) side struck total: 2

(2.) quartzite chips (total: 3)

(a) end struck total: 1 (unit 13, level 1)

(b) side struck total: 1 (unit 18, level 1)

2. WOOD-BONE WORKING TOOLS: Wood-bone working
tools consist of two notched flakes, a beaver incisor chisel,
and a sandstone abrader.

a) NOTCHED FLAKES (Plate 3, I). One notched flake
came from level 1 of unit 22 and has maximum dimensions
of 19 x 8 x 3 mm. The notch is 2 mm. in depth and 5 mm. in
length. An end- struck flake found in the plow zone has a
notch on each side and maximum dimensions of 25 x 15 x 2
mm. Like the first flake the notches are unifacially re-
touched, although each notch is retouched from a different
face.

b) SPLIT BEAVER TOOTH INCISOR (Plate 2, c).
This incisor was found in level 1 of unit 10 and has con-
siderable wear on both ends.

c) SANDSTONE ABRADER (Plate 4, A). A single
abrader 77 x 43 x 28 mm. in maximum dimensions came
from Feature 4. This irregular stone has a single groove
3 mm. in width on each of its two parallel flat surfaces.

3. DRILLING-PERFORATING TOOLS: Three bone awls,
a bone punch, three flake drills, and a chipped stone per-
forator comprise this tool category.

a) DEER ULNA AWLS (Plate 2, a). The broken tip
of a right ulna came from Feature 2. A left ulna 96 mm.
in length came from level 1 of unit 2.

b) FLAT SPLINTER AWL (Plate 2, d right). A single
flat splinter awl from the plow zone is 53 mm. in length.

c) BONE PUNCH OR NEEDLE (Plate 2, d left). A
broken bone punch or needle was recovered from the plow
zone.

d) FLAKE DRILLS. Three flakes exhibit continuous
unifacial wear or secondary retouch along the margins of a

The Bornick Site 99

thin natural point. These flakes range in length from 15-21
mm. All lack striking platforms or bulbs of percussion.
Their distribution is: Feature 3, level 3; unit 1, level 1;
unit 17, level 1.

e) CHIPPED STONE PERFORATOR (Plate 3, H).
A single crude perforator on an irregular flake lacking a
striking platform or bulb of percussion was found on the
surface. The maximum dimensions of this implement are
27 x 17 x 7 mm. Length of drill point is 6 mm. and width
of point base is 9 mm.

4. MAT-FABRIC MAKING TOOLS: A single, broken,
shell tempered POTTERY DISC from level 2 of Feature
3 was too small to discern the presence of a central
perforation or its diameter.

5. POTTERY MANUFACTURING WASTE: Single lumps
of fired clay of various sizes and shapes were found in
level 2 of unit 19 and in level 1 of units 7-9, 13, 15, 16,
and 22. Two lumps came from the first level of units 10
and 18. In addition to the twelve pieces of ceramic waste
recovered in situ, four were picked from the plow zone.
All of these pieces were quite small and may be untem-
pered waste from the manufacture of clay artifacts, such
as pots.

6. GENERAL CUTTING AND SCRAPING TOOLS: General
cutting and scraping tools at the Bornick site consist of two
knife fragments, two scrapers, and unifacially and bi-
facially modified flakes.

a) BIFACIAL KNIVES (Plate 3, G). A small chert
knife base came from level 1 of unit 17 and has a maximum
basal width of 17 mm. A white quartzite knife base with a
maximum basal width of 24 mm. came from the plow zone.
Both knives have beveled working edges and carefully
flaked faces.

b) SCRAPERS (Plate 1, ,C). Two scrapers, an end
and a side scraper, were found in situ. The end scraper is
of chert, has maximum dimensions of 43 x 22 x 14 mm.,
and came from level 3 of Feature 3. Secondary retouching
extends 10 mm. down the right side; the working length
measures 4 mm. and the dorsal surface is highly ir-

^ular. There is no striking platform or bulb of per-
cussion. The side scraper is of yellow quartzite and came
from level 1 of unit 17. Length of the working edge is 19

100 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

mm. and maximum dimensions are 20 x 16 x 5 mm. There
is no lateral retouch on the flake, which is side-struck
c) SIDE AND END RETOUCHED FLAKES (Plate 3^
A-E). The remainder of the worked and utilized flakes
have been divided into a number of categories depending
on location and position of wear and striking platform and/or
bulb of percussion. Of these forty flakes twenty-three are
unifacially retouched and seventeen bifacially retouched.
The flakes range in length from 10-52 mm. A standardized
size or form seems to be lacking. The distribution of these
flakes are listed in Table 33.

(1.) side
unifacial
of flake

(a) end struck

(b) side struck

retouched (total: 24)

retouch along one edge parallel to long axis

total: 4
total: 1
total: 5

unifacial retouch along both edges parallel to long
axis of flake

(a) absent total: 1

bifacial retouch along both edges parallel to long
axis of flake

(a) absent total: 2

bifacial retouch along one eage parallel to long axis
of flake

(a) end struck total: 3

(b) side struck total: 1

(2.) end retouched (total: 13)
unifacial retouch along one end perpendicular to long
axis of flake

(a) side struck total: 1

O>) absent total: 8

unifacial retouch along both ends perpendicular to long
axis' of flake

(a) side struck total: 1

(b) absent total: 1

The Bornick Site 101

bifacial retouch along one end perpendicular to long
axis of flake

(a) absent total: 2

(3.) end and side retouched (total: 3)

unifacial retouch along one end and side of flake
(a) absent total: 1

bifacial retouch _along one end and side of flake

(a) absent total: 2

7. GENERAL FOOD PROCESSING IMPLEMENTS: Im-
plements in this category consist of seven grinding stones,
two shell spoons, a clay spoon, and a possible metate.
a) GRINDING STONES (Plate 4, C and F). Seven
ground stone artifacts with a single flat surface apparently
the result of abrasion are classified as grinding stones. The
provenience, weight, maximum dimensions, and number of
these artifacts are recorded in Table 5. All of these im-
plements are irregular in shape and size. A surface find
not included in the table may represent a fragment of a
large metate.

Table 5 Provenience and Maximum Dimensions of Grinding Stones

n unit/ . . weight maximum dimensions in cm.
feature e in gins, length width thickness

1 P3 3 -H 32 29 13

2 ?3 3 23 32 30 21

3 F4 - 28 26 24 16

4 ?5 2 140 56 45 42
54 1 561 99 8? 59

6 11 1 163 64 51 29

7 19 2 ?4 (55) 40 26

b) SHELL SPOONS (Plate 5, c). The tang and part
of the body of a shell spoon made from a right valve was
found in the plow zone. This spoon is of the Ha type in
J.W. Griffin's classification (1946:35). The lateral edge of
a second spoon came from level 3 of Feature 3. Notching
along the lateral edge suggests that it may be part of a type
"b spoon.

c) CLAY SPOON OR LADLE (Plate 8, G). A small,
broken, shell tempered implement may represent a spoon

102 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

or ladle. The implement came from level 1 of unit 10 and
was apparently molded into shape by hand. It is slightly
irregular in outline and has a tab on one end.

C. Recreational and Ornamental Artifacts

1. CLAY PIPES. Small fragments of three shell tem-
pered pipes were found in the site debris. An elbow pipe,
from level 3 of Feature 3, is peculiar in that both arms
are' welded together at the elbow. The diameter of the hole
in the stem is 7-8 mm., narrowing to 6 mm. at the juncture
of stem and bowl. Lip thickness is 5 mm. The bowl is too
fragmentary to note more than that it is plain and bells
outward. A swelling or crest beginning 5-6 mm. below the
flat bowl lip is the only visible decoration.

A second pipe consists of two fragments of a bowl from
level 3 of Feature 3 and Feature 4. Except for color, these
fragments could represent pieces of the first pipe. The third
pipe is represented by a sub-angular corner of a bowl found
in the plow zone.

2. PAINT STONES (Plate 4, D and G). A stone found in
level 1 of unit 14 has a reddish residue on one flat surface.
The maximum dimensions of the stone are 66 x 44 x 21
mm. The surface of a stone from the plow zone has been
worn down into a groove possibly through abrasion of a
pigment, for a white and gray residue adheres to the sur-
face. Maximum dimensions of this stone are 105 x 62
x 24 mm.

3. GALENA. Two galena cubes came from level 1 of
unit 14.

4. COPPER BEADS ANt) FLAKES (Plate 3, J). Two
tubular 'beads' and two waste flakes were found in un-
disturbed midden. A third waste flake was picked from the
plow zone.

Table 6. Provenience and Maximum Dimensions of Copper

u unit/ , .. maximum dimensions in mm. . .

feature xe ex length width thickness e

1 V 1 22 14.5 5 float copper

22 t4 6.0 0.5 flake

3 16 1 15 5.5 - bead

4 18 1 11 6.0 - bead

5 Sf 14 8.0 0.5 flake

D. Containers

A globular or ellipsoidal shell tempered pottery vessel

The Bornick Site 103

with a rounded base and shoulder area, constricted neck,
and out-flaring rim was the only type of large container
(with the exception of pits) found at the site. No complete
or nearly complete vessels were recovered. A count of large
rim sections suggests that there was a minimum of at least
seventy-one of these jars. Fragments of eleven miniature
shell tempered jars also came from the site. Ther ceramic
analysis of large jars is based on the following sample:

in situ 3888 12?

surface 2165 124

site total 6053 251

Most sherds have a moderate to abundant amount of
shell temper, few pieces of which exceed 3 mm. in length.
Every large sherd examined also contained sperical iron
oxide particles ranging in size from minute to 3 mm. in
diameter. The surfaces of all sherds are dull and smooth.
Nearly all vessels exhibit scraping or smoothing in the
form of numerous parallel striations. The consistency of
the sherds ranges from flaky to a more moderately compact
structure. The thickest areas of the average vessel are the
neck and base. Table 7 indicates the frequency in percent of
body and shoulder sherd thickness. Vessel height and
capacity could not be determined. The majority of vessels
had the same interior and exterior colors. Table 8 contains
a representative sample of vessel colors. Sherd cores are
generally grayer than surface colors.

Table 7. Body and Shoulder Sherd Thickness
maximum thickness in mm. percent

1.5 to 2.5 1.0

2.6 to 3-5 1 3-0
3.6 to 4.5 16.0
4.6 to 50 27.0
5.6 to 6.5 23.0
6.6 to 7-5 10
7.6 to 8.5 6 -0
8.6 to 9.5 1 -5
9.5 to 10.5 .2.0

10.6 to 11.5 '0*5

104 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

Table 8#* Percentages of Surface Colors on Vessels

yellowish- reddish-

locatlon brown brown brown gray black

exterior 25 19 1^ 5 17
interior 28 3^ ^ 6 18

The terminology used to describe vessel form consists
of lip, rim, neck, shoulder area, and body:

1. lip- juncture of the inner and outer surfaces of the
vessel

2. rim -zone between the lip and the neck

3. neck-point of minimum constriction where rim and
shoulder are joined

4. shoulder area- space between the neck and the point
of maximum body diameter

5. body-area between the neck and the base of the vessel.

1. ATTRIBUTES OF LARGE JARS. Individual attribute
classes and their elements are described below. Com-
binations of elements on vessels (rimsherds with a fairly
large portion of the shoulder attached) and unattached rims
and lips are recorded in Appendix A. Tables in the appendix
also contain the weight of pottery and number of plain sherds
by unit and level. A brief discussion of established types
follows the discussion of attribute classes.
CLASS I: LIP PROFILE

Lip profiles from large jars from the Bornick site
are of five different kinds (Figure 4, A). Table 9 below
contains the frequency and percentage of these elements.

The Bornick Site

105

o O oo
o0 o
o o o o

Figure 3. Class XVI Shoulder Design Elements

,n <n ,h

A. Class I Lip Profile Elements (interiors to left)

10 (J fll

B. Class III Lip Design Elements

C. Classes VIII
and X

O O OO

0000000

figure 4. Lip and Rim Elements for Classes I, III, VIII, and X

106 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

Table 9 Lip Profile Elements

lip profile

lips and riais
no. J6

vessels

no. %

site

no.

total

thickness (IKE.)
range mean

rounded

136

69.2

1*5

51.0

3-7

*.7

flattened

15,*

29.0

- 8

5.6

inward bevel

15.*

10.0

- 8

5.*

outward bevel

1.5

- 5

thickened

8.5

-11

6.0

total

270

100

100.0

283

100.0

CLASS II: LIP DECORATION TECHNIQUE

All fragments large enough to consider vessels have
plain lips. Techniques employed on unattached rim and lip
sherds are: 1) NARROW TOOL-IMPRESSIONS: shallow
elliptical impressions 4 to 7 mm. in width made by pres-
sing a narrow object downward at an angle to the still
malleable lip surface; 2) FINGERTIP PUNCTATE: shallow
ovoids 5 to 7 mm. in width produced by pushing the finger-
tip straight downward onto the still soft lip-a fingernail
imprint can usually be found in the punctations; 3) FINGER
OR BROAD TOOL-IMPRESSION: impressions usually be-
tween 10-14 mm. in width made by pushing the finger or a
broad tool down onto the lip.

The following table lists the frequency and percent of
these elements.

Table 10. Lip Decoration Techniques
unattached

rims

lips

vessels

site total

no.

_ 2L

no.

0.45

0.3

3.5

0.45

1.1

1.40

1.1

plain

96.5

201

ffil

100

226

97-5

total

100

212

100

283

100

CLASS III: LIP DESIGN ELEMENTS

Designs on decorated unattached rims and lips are of
three kinds: 1) consecutive finger or broad tool-impres-
sions 3 to 20 mm. apart; 2) consecutive fingertip punc-
tates 2 to 30 mm. apart; 3) consecutive transverse, narrow
tool-impressions 3 to 8 mm. apart. These elements are
illustrated in Figure 4, B. The provenience of sherds with

The Bornick Site 107

lip design is recorded in Table 30. Table 11 lists the per-
cent and frequency of Class III elements.

Table 11. Frequency and Percent of Lip Design Elements

unattached lips unattached rims vessels site total
deslgn no. % no. _% *o._* n

13 1.4 o o o o 3 j.i

21 0.5 2 3.4 o o 3 1.1

51 0.5 o o 001 0.3

97.6 56 96.6 I2.JOO 276 97.3

99.9 58 100.0 13 100 283 100.0

CLASS IV: RIM PROFILE

Rims of unattached rimsherds and vessels were as-
signed to seven basic profiles (Figure 5). Profiles 1, 2,
and 6 are characterized by a straight to gently curving rim
5 to 45 from a vertical line drawn through the neck and
are differentiated by arbitrary rim height limits (1 - greater
than 35 mm., 2 - between 20 and 35 mm., 6 - less than 20
mm.). A study of standard deviation and other measures of
dispersion may suggest less arbitrary breaks than these.
Curved (Figure 5, left in 1) and fairly straight rims (Figure
5, center and right in 1) in these three categories are
designated by a and b respectively. Table 12 records the
frequency and percent of basic profiles.

Table 12. Frequency and Percent of Basic Aim Profiles

profile. unattached -rims vessels site total

no. no. no.

1 10 17.8 3 23.1 13

2 22 39.3 4 30.6 26

3 4 7.1 3 23.1 7

4 1 1.8 00 1

5 2 3.6 1 7.7 3

6 16 28.6 2 15.4 18

7 1 1.8 0_ _j

total 56 100.0 13 100.1 69 99.9

108 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

Figure 5 . Class IV Rim Profiles

(1-3 profile 1; 4-6 profile 2; 7-9 profile 3; 10 profile 4;
11-12 profile 5; 1>16 profile 6; 17 profile 7)

CLASS V: RIM CROSS SECTION

Rim cross sections displayed a rim either of even
thickness (1), thinning to the lip (2), or expanding to the
lip (3).

The Bornick Site 109

Table 13- Frequency and Percent of Rim Cross Sections

cross section unattached rims vessels site total

no. % no. % no. %

1. even thickness 1? 3^7 2 15.4 19

2. thinning to lip 28 57.1 11 84.6 "39

3. expanding to lip 4 8.2 4

total 49 100..0 13 100.0 62 100.0

CLASS VI: RIM HEIGHT

The height of rims from the Bornick site ranges from
8 to 51 mm. with a mean of 24 mm. Table 14 lists the
frequency and percent of rim heights by arbitrary cate-
gories. Individual rim heights are tabulated in Tables 27
and 30. Measurements of dispersion for rim height and for
lip, rim, and neck thickness are listed in Table 31.

Table l4. Frequency and Percent of Rim Heights

height unattached rims vessels site total
in mm. no. no. no. %

1 to 9 1

10 to 19 20

20 to 29 22
30 to 39 9
40 to 49 3

50 to 59 1

total 56 100.0 13 100.0 69 100.0

CLASS VII AND VIII: INNER RIM DESIGN TECHNIQUE AND
ELEMENTS

Two unattached rims and two lip fragments are dec-
orated on the upper edge of the inner rim with continuous
(Class VIII, element 1) finger or broad tool-impressions
(Class VII, element 1). Impressions are 14 to 20 mm. in
maximum diameter and are ovoid in outline. The basic
decorative design is a single row bf impressions 5 to 15
mm. apart circling the upper edge of the inner rim (Fig-
ure 4, C). The provenience of these sherds is recorded in
Table 30.

CLASS IX AND X: LIP EXTERIOR DESIGN TECHNIQUE
ND ELEMENTS

The lip exterior of three rim sherds and four lip frag-
ments are decorated. The rim sherds and three of the lip

1.8

7.7

2.9

35.7

23-1

33-3

39.3

46.1

40.6

16.1

15.4

15-9

5-3

7.7

5.8

1.3

1 .3

110 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

fragments are decorated with continuous (Class X, element
1) finger or broad tool-impressions (Class IX, element
1). Decoration on the fourth lip fragment consists of a band
of continuous, vertical (Class X, element 2) elliptical tool-
impressions (Class IX, element 2) (Figure 4, C). Ellipti-
cal tool-impressions are 4 to 7 mm. in width. Other
: measurements conform to those given in Classes VII and
VIII. The provenience of sherds with lip exterior modi-
fication is listed in Table 30. . , {
.CLASS XI: ANGLE OF RIM TO BODY

The angle of the rim to the body on all vessels that could
be measured is outslanting (element 1), with an angular 1
divergence of 5 to 45 .
CLASS XII: SHOULDER-NECK CONFIGURATION

The juncture of the rim and neck on all sherds large,
.enough to consider representative of a vessel was curved:
(element 1). Necks on one unattached rimsherd (unit'
14, level 1) and two plain shoulder sherds (level 1 in units
5 and 18) are angular (element 2). One hundred and sixteen
other sherds have curved necks. The small percent of<
angular necks (2.3%) is comparable to the Walker-Hooper'
site sample (1.9%). Table 29 contains the frequency and.
provenience of sherds consisting only of curved neck
fragments.
CLASS XIII: SHOULDER DECORATION TECHNIQUE

Seyen decoration techniques were used on shoulder!
sherds found at the Bornick site. None of the shoulders!
of the thirteen vessels were decorated. Only 4% of all body-
sherds are decorated. The decoration techniques used are:
1) TRAILED TOOL^- a narrow blunt-ended tool pulled
across a still malleable shoulder surface; 2) TRAILED
TOOL AND BLUNT PUNCTATE - trailing combined with
punctates 2 to 4 mm. in diameter made by pushing a blunt
ended tool straight down into the soft clay; 3) REED
PUNCTATE - small round punctates usually *2 to 3 mm.
in diameter and" 1 to 5 mm. in depth made with a hollow
tube, presumably a reed or bone; 4) TOOL-IMPRESSED -
an impression formed by placing a blunt-ended object on
the soft vessel surface and pressing downward at an angle
to the surface; 5) REED TRAILED - a variant of 1 with
numerous strjations within the trailed line; 6) 'HALF-
MOON' PUNCTATE - punctates formed by a solid, blunt-
, ended, crescent 2 to 4 mm. in maximum width; 7) FINGER-

The Bornick Site

111

TIP PUNCTATE. Some of these techniques are illustrated
in Plates 6 and 7. The provenience of decorated bodysherds
and their associated attributes are recorded in Table 29.

Plate 6. Class XVI Shoulder Design Elements

A. element 1 E. element 5

B. element 2 F. element 6

C. element 3 G. element 7

D. element 4 H. element 8

112 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

Plate 7. Sherd Design Elements and Miscellaneous Ceramics.
Sherd Design Elements

A. design 6 : two or more curvilinear trailed lines

B. design 9 : two or more tool -impressions

C. design 9 : two or more punctations, blunt -ended

D. design 3 : two intersecting trailed lines

E. unique surface sherd with two chevrons

Miscellaneous Ceramics

F. pottery4ug (unit 18, level 1)

G. pottery ladle (?) (unit 10, level 1)

H-L. miniature vessel fragments (H. surface; I. unit 4, level 1;
J. unit 1, level 1; K. unit 9, level 1; L. surface)
M. ceramic waste

The Bornick Site 113

Table 15. Frequency and Percent of Shoulder Decoration Techniques

shoulder decoration .shoulder sherd fragments

1. trailed tool 21 4

2. trailed tool and blunt punctate 2
3 reed punctate 1

4. tool-impressed 4

5. reed trailed 29

6. 'half-moon' punctate 1

7. fingertip punctate 2

(plain sherds: in situ ... 3883)
: surface ..

total 253 100.1

CLASS XIV: WIDTH OF TRAILED LINES

The width of trailed lines decorating the exterior
surfaces of shoulder sherds in Class XIII was measured
and arbitrarily assigned to one of three categories.

Table 16. Frequency and Percent of Categories of Trailed Line Widths

width category shoulder sherd fragments

no.

1. fine .1 to 2.5 mm. 35

2. medium 2.6 to 4.5 ram. 126

3. wide 4.6 to 8.0 mm. 82

total 243

CLASS XV: SHOULDER DESIGN

Decorated bodysherds were separated into categories
on the basis of the kinds of designs displayed. Table 17
records the frequency and percent of these categories.

Table 17. Frequency and Percent of Shoulder Designs

shoulder design shoulder sherd fragments

no. %

1. rectilinear 220 8?. 6

2. curvilinear 17 6.8

3. both 1 and 2 4 1.6

4. tool-impressions or punctations 8 3.2

5. both 1 and 4 2 O.o

total 251 100.0

CLASS XI: SHOULDER DESIGN ELEMENTS

Decorated shoulder sherds large enough to display a

114 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

complete or nearly complete shoulder element exhibited
seven different elements. These elements are illustrated
in Figure 3 and in Plate 6 (A-G). Sherds too small to
exhibit an element were separated into nine sherd design
element categories. Table 18 lists the frequency and percent
of both shoulder and sherd design elements. The sherd
design elements are:

1. single straight line

2. two or more parallel straight trailed lines

3. two or more intersecting trailed lines in a chevron
pattern (Plate 7, D and E)

4. criss-crossing trailed lines

5. single curvilinear trailed line

6. two or more curvilinear trailed lines (Plate 7, A)

7. one straight and one curvilinear line

8. trailed line with bordering punctates or tool-im-
pressions

9. two or more punctates or tool-impressions (Plate 7,
B and C).

Table 18. Frequency and Percent of Shoulder Design Elements and Frag-
mented Sherd Design Elements

shoulder bodysherds fragmented bodysherds.

design sherd design

elements no. % elements no. %

1 15 55.6 1

2 3 11.1 2
3 4 14.8 3

4 1 3-7 4

5 2 7.4 5

6 1 3.7 6

7 1 3-7 7

8
9

27.... 1.00.0 222 99.9

CLASS XVII: LOCATION OF DECORATION

None of the thirteen sherds large enough to consider
representative of whole vessels are decorated. The re-
maining sherds are too small to determine combinations
of decoration on the lip, rim, and shoulder. However,
the scarcity of decorated lips, rims, and shoulders suggests
that vessels from the Bornick site are best typified by the
absence of decoration.

The Bornick Site 115

CLASS XVIII: APPENDAGES

No loop or strap handles were found at the Bornick
site. A fragment of what appears to be a shell tempered,
rectangular lug came from level 1 of unit 18 (Plate 7, F).
This appendage was made by folding-over and shaping a flat
strip of clay. Opposing holes in the upper shoulder area
were apparently the primary means of suspension. The
distribution of the seven sherds with holes drilled through
from the exterior is recorded in Table 30.

2. Large Jar Pottery Types

The thirteen large, globular, shell tempered jars that
are complete enough to treat as whole vessels can all be
assigned to the type GRAND RIVER PLAIN (Hall 1962:
68-70). Table 27 contains the provenience of the^e vessels
and their attribute combinations. The large number of
plain lips and bodysherds indicates that most of the vessels
in the sample were of this type. The predominanace of
rectilinear shoulder designs and plain lips also suggests,
that GRAND RIVER TRAILED vessels were the next most
frequent type. A few vessels combining lip notching and
curvilinear shoulder designs (Class XV, elements 2 and 3)
or lip notching and plain shoulders may have been of the
CARCAJOU CURVILINEAR and CARCAJOU PLAIN types.
The actual presence of Carcajou types cannot be verified,
because of the fragmented condition of the remaining
vessels.

3. MINIATURE VESSELS (Plate 6, H; Plate 7, H-L)
Sherds representing eleven shell tempered miniature

vessels are described in Table 28 in the appendix. Seven
are unattached rim sherds, one is a shoulder -herd, and
three are complete enough to consider vessels. Attributes
are those described for large jars, except for shoulder
design element 8 (Plate 6, H). No food residue was found
within any of the vessels.

4. MISCELLANEOUS SHERDS

a) PAINTED SHERD (Plate 6, D). A shoulder sherd
ffith a positive painted black chevron pattern came from the
plow zone.

b) PROBLEMATIC SHERD (Plate 7, left in A). A
sherd with a marked concavo-convex cross-section, probablj
representing a node on the shoulder of a small vessel, was
found in the plow zone. The exterior of the sherd is dec-
orated with at least two narrow trailed lines encircling the

116 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

)ower half of the node.

E. Miscellaneous Artifacts

1. MISCELLANEOUS WORKED SHELL. Two pieces of
scored shell (Feature 5, level 2; unit 5, level 1) are the
only other shell artifacts from the site.

2. WORKED SLATE. Two small pieces of reddish slate
from level 1 of unit 10 appear to be ground around the edges.

3. IRON CONCRETIONS (Plate 4, B). Two iron con-
cretions may or may not be artifacts. Their provenience
is Feature 4 and the plow zone. There are no signs of
scratching or other indications of use.

THE ECONOMIC BASE

A. Plant and Animal Remains

Tables summarizing the range and quantity of particular
classes of animal species are included at the end of this
subsection. The distribution of some of these classes by
provenience unit are tabulated in Table 23. The distribution
of other species can be obtained through the Department
of Zoology at the University of Wisconsin. The mammals,
birds, and reptiles were identified by Elizabeth Pillaert.
Counts of individuals were obtained by taking the highest
number of any identifiable bone from either side of the
animal's body. Dr. David A. Baerreis identified the fresh-
water and terrestrial snails. The fish and fresh-water
mussels were identified by the writer. Counts of individual
mussels were obtained by taking the largest number of
right or left valves. Lefts and rights of ten fish elements
were identified for the large sample offish. These elements
were among the most easily identifiable and most prom-
inent numerically of the fish bones (e.g. dentary, articular,
premaxillary, maxillary, opercular). The largest number
of any left or right of these ten elements was used to
determine the minimum number of individuals. This method
provides a farily reliable count of the minimum number of
individuals. The actual number is undoubtedly higher.

The carbonized seed and nut remains were identified
by Mr. Leonard Blake. Thirty-seven corn grains, forty-
three hickory nuts, six acorns, and five hazelnuts were
identified. Table 23 contains the distribution by provenience
of corn grains and Hickory nut shells. Acorns were found

The Bornick Site 117

in: Feature2; Feature 3, level 2; unit 2, level 1; unit 14,
level 1; unit 18, level. Hazelnuts came from: Feature 3,
level 3; Feature 4; unit 2, levels 1 and 2. Measurement of
corn grains gave an estimated mean row number of 8.6
(of thev*jpe grains measured six were estimated to be 8- row
and three 10-row). A large amount of charcoal was also
obtained and is presently stored at the University of Wis-
consin in the Department of Anthropology.
B. Habitat Analysis

In order to reconstruct the exploitive pattern of pre-
historic peoples it is necessary to attempt to reconstruct
the composition of their natural surroundings. The method
used to interpret the composition of the local flora and
fauna during the period of site habitation is the 'scored
habitat preference' method (e.g. Cleland 1966). All of the
mammal, bird, and reptile species were given two points
for primary habitat preference and a single point for
secondary preferences. An indication of the composition
of the natural habitat at the site can be obtained by adding
the columns. The following table summarizes the per-
centages of the habitat preferences at the Bornick site.

Table '19. Habitat Preferences of Mammals, Birds, and iteptiles Expressed
in Percent

deciduous forest forest-edge grasslands
20 15 11

The percentages in Table 19 suggest the presence of
extensive waterways and marshland, open deciduous for-
ests, and some grasslands. This environmental setting is
essentially identical to that of the early historic period..

C. Exploitive Pattern

The animal bones and their depositional context reflect
a primary emphasis on the exploitation of aquatic re-
sources. Minimum pounds of meat have been estimated
for the total numbers of animals (Tables 20-22). Table 24
below contains the estimated minimum pounds of meat
obtained from each class of animal and the percent of
meat from animals by habitat preferences.

118 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

Table 20. Vertebrates Identified from the Bornick Site (Table prepared
by Elizabeth Pillaert)

-H O C

a c -H

Maranals 385 8?. 10 23 53.49 1151 98.04

Blarina brevicauda 1 0.23 1 2.33

Short-tailed Shrew

Citellus tridecemlineatua 1 0.23 1 2.33

Thirteen- lined Ground
Squirrel

Castor canadensis 11 2.49 1 2.33 39 3.32

Beaver

Peromyscus maniculatus 6 1.36 3 - 6.98

Prairie Deer Mouse

Ondatra zibethica 12 2.71 3 6.98 6 0.51

Kuskrat

Canis familiaris 11 2.49 1 2.33 9 0.77

Domestic Dog

Canis sp. 1 0.23 1 2.33 9 0.77

probably Domestic Dog

Procyon lotor 10 2.26 2 4.65 18 1.53

Raccoon

Kartes pennanti 1 0.23 1 2.33 7 0.60

Fisher

Lutra canadenais 2 0.45 1 2.33 13 1.11

River Otter

Cerus canadensis 7 1.58 1 2.33 350 29.81

Elk

Odocoileus virginianus 322 72.85 7 16.28 700 59.63

White-tailed Deer

Unidentified Bones 10044 -------------

Birds 38 8.60 14 32.56 18 1.53

Branta canadensis 2 0.45 1 2.33 6 0.53
Canada Goose

( Not considered a food item)

The Bornick Site 119

Table 20 <cont.)

30 %
.3

I 8 3

JS 2
^ a

Anas platyrhynchos

0.23

2.33

2 0.17

or A. rubripes

Mallard or Black Duck

Anaa dicors

0.68

2.33

:W-.':

Blue-winged Teal

Spatula clypeata

0.23

2.33

1 0.09

Shoveler

Aythya collaris

2.26

6.98

4 0.34

or A. af finis

Ring-necked Duck or

Lesser Scaup

Erismatura jamaicensis

0.23

2.33

1 0.09

Ruddy Duck

Anatidae

2.94

6.98

3 0.26

Ducks

Pedioecetes phasianellus

0.68

2.33

1 0.09

Sharp- tailed Grouse

EC topi s te s migratorius

0.23

2.33

Passenger Pigeon

Passeriforms

0.68

2.33

; *V".?

Songbirds

TTy*4 HATT+"T "Pi eH "R/Mioo

pj-C

UIULUtflllsX x XeU. i-XJIlfc; o

\>J

Reptiles

-2.49

9.30

5 0.43

Chelydra serpentina

0.23

2.33

3 0.26

Snapping Turtle

Emys blandingi

0.45

2.33

1 0.09

Blanding's Turtle

Chrysemys pi eta

1.36

2.33

1 0.09

Painted "Turtle

Snake sp.

0.45

2.33

Unidentified Bones

160

Amphibians

1.81

4.65

Rana catesbiana

1.81

4.65

_ _

Bullfrog

Unidentified Bones

wm.u.eii ox! .Leu djiivzo

^-7

totals

442

100.02

100.10

1174 100.00

120 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

Table 21. Pish Identified from the Bornick Site

SPECIES

mini mum estimated
no. of ^ Ibs. of

individ. meat

711 97.8 925.1

99.9'

Ictalurus sp.

143 17.9 71.5

7.7

Bullhead

Aplodinotis grunniens

213 30.0 319.5

34.5

Fresh- water Drum.

Micropteras sp.

160 22.5 240.0

25.9

Bass

Centrarchidae sp.

80 11.2 20.0

2.2

Crappie

Cottus bairdii

20 2.8 10.0

1.1

Mottled Sculpin

Moxostoma sp.

11 1.5 33.0

3.6

Redhorse

Ssox lucius

14 2.0 98.0

10.6

Northern Pike

Percidae sp.

4 0.6 1.6

0.2

Perch

Salvelinus sp.

11 1.5 55.0

5.9

Trout

Lepisosteus sp.

14 2.0 56.0

6.0

Gar

Unidentified Species 41 5.8 20.5 2.2

The Bornick Site

121

Table 22. Fresh-water Mussel and Snail Remains from the Bornick Site
1

A. Fresh-water Mussels

Ac tinonal a s carinata
Amblema costata
Laamigona costata
Ligumia recta
Pleurobema coccineum
Strophitus rugosus

Unidentified species
(Sphaeriidae - 7)

Hinges

minimum number of individuals

estimated Ibs. of meat

B. Fresh-water Snails

Amnicola limosa (Say)
Amnicola lustrica (Pilsbry)
Campeloma decisum (Say)
Gyraulus parvus (Say)
Helisoma anceps (Menke)
Helisoma campanulatum (Say)
Physa sp.

Pleurocera acutum (Raf.)
Valvata tricarinata (Say)
Unidentified juvenile

total individuals

number
47
2
11
21
15
28

201

22?
14

number
2
6
1
9
7
5
3
10
4
1

(provenience)

-.lit 18, lev. 1

(unit 11, lev. 1)
(unit 5, lev. 1)

( provenience of individuals listed in Table 21)

122 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

Table "2 3. Distribution of Fresh-Water Mussels, Corn and Nuts, and
Fresh-Water Snails at the Bornick Site

mussels

plants fresh-water snails

B .

unit/featur<

Actinonaias

Lasmigona c,

E
I

Pleurobema

Sphaeriidae

[Amblema cos.

hinges

Icorn grains

hickory nut

[acorn

Umnicola li,

[Amnicola lu,

byraulus pa,

Kelisoma an.

[Helisoma ca

[Pleurocera

[Valvata tri,

juvenile

H CO

H ft
0)

P 1

P 2

P 3

120

P 4

P 5

113

245

-120

x - indicates the presence of less than a teaspoon of nut shell

t - represents one teaspoon of nut shell

+ - indicates less than one gram of fish bone

The Bornick Site 123

Table 24. Minimum Pounds of Meat by Animal Type and
Percent of Meat from Animals by Habitat
Preferences

1. animal type marrreals birds reptiles fish mussels to^al
minimum Ibs. 1151 18 ! 5 925' 1 2 *- 2113 Ibs,

2. habitat type aquatic dec, forest forest-ed/re ^rr^-ss lands

49 12 31

The breakdown of pounds of meat by habitat type was

obtained by distributing the pounds of meat from a particular

species according to its primary and secondary habitat

preferences. Although it is difficult to estimate the amount

of energy involved in exploiting an aquatic habitat compared

to a forest and forest-edge, it would seem that a much

.greater amount of effort was expended obtaining the small

mammals, fish, and birds from the waterways. The amount

of effort would depend of course on the techniques of

.exploitation. Harvesting of at least the fish seems to be

indicated by the very small size of the fish bones analyzed

;and the flat net-sinker. Small mesh nets and traps would

ihave been especially effective during the spring, when most

| of the fish species at the site spawn. If the interpretation

of bone and antler projectile points as fishing weapons is

correct, this would indicate an even greater expenditure of

energy on the exploitation of fish.

Meat and plant foods were not the only sources of calories
jat the Bornick site. The large numbers of smashed bone
strewn throughout the site probably indicates the practice
| of making bone grease (e.g. Leechman 1951). Ninety-six
percent of the mammal, bird, and reptile bones at the
Bornick site are too small to identify. This practice is also
indicated by the kinds of bones not smashed. These consist
almost exclusively for the large herbivores of lower leg
bones, which lack meat and probably were too small to
bother smashing It is interesting to mention in this con-
text, that the Hickory nut shells from the site were also
smashed into small pieces. Leonard Blake has suggested,
;that this "may indicate the practice described by Swanton
and others of throwing finely pounded hickorj nuts into
boiling water and then skimming off the resulting oil from
the top of the water" (pers. comm.). Both of these practices
may indicate that sufficient amounts of fat were not being

124 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

obtained from the fish, mussels, and birds. Similar prac-
tices were conspicuous at the large Walker -Hooper village
site.

The importance of other kinds of food is difficult to
determine. Although only thirty-seven corn grains were found
at the site, their scattered distribution throughout the site
wquld seem to indicate that corn may have been quite
common.

SITE FUNCTIONS

All items except storage pits, unworked flakes, ceramic
waste, and miniature vessels were grouped together within
their activity classes. The percentages of these classes are
given in Table 25 below.

Table 25. General Activity Emphasis at the Bornick Site

activity

1. food procurement 13.7%

2. food storage 26.9%

3. fabricating and processing 55.6%

4. recreation and ornamental 3.8%

100.0%

LOCAL CHRONOLOGY AND EXTERNAL RELATIONSHIPS

A single radiocarbon date was obtained on wood char-
coal taken from level 3 of Feature 3. The date was 660
50 B.P. 1290 A.D. (WIS-288). This age determination
is close to five radiocarbon dates for the Walker -Hooper
site. The mean of these five dates ranged between A.D.
1200 and A.D. 1240. In order to further test the degree of
similarity or dissimilarity between the two sites indices of
similarity were calculated using ceramics as the single
most extensive artifact category at the two sites. Because
there seemed to be evidence for an extended occupation at
the Walker-Hooper site, the five areas of the site exca-
vated were retained as separate entities. Large differences
in the indices of similarity would support an argument for
appreciable temporal and/or cultural differences between

The Bornick Site 125

the excavated areas of the two sites.

The method used to calculate numerical indices of
similarity was based on ordination methods previously
employed in ecological studies (e.g. Goodall, 1954; Bray
and Curtis 1957; Curtis 1959) and in soils (Hole and
Hironaka 1960; Bidwell and Hole 1964). One hundred and
twenty-one ceramic variables were chosen for comparison
between the five excavated areas of the Walker -Hooper
site and between these and the Bornick site. Variables
selected were those previously described in Classes I-
XVII. An additional variable 'plain' was added where
relevant. As a control, 124 variables were chosen for
comparison between the Grand River phase excavations
(the Walker-Hooper and Bornick sites) and two areas
previously excavated and described by the writer from
the Orr phase Midway site (Gibbon 1970).

The data were scaled by using proportions rather than
rank correlations of frequencies in each class and then
interjected into the following formulas:

Index of Similarity s I- (2W)/(A*B) X 100

Index of Dissimilarity- D-100 - I.

The meanings of the symbols and the method of scaling
the data are described by Hole and Hironaka (1960). Each
of the variables was treated as an independent attribute
and each class was given equal weight. However, equal
value is usually not given each class or variable by
archaeologists in constructing pottery "types." Methods
of weighting are discussed by Hole and Hironaka (1960)
and by Bidwell and Hole (1964). Although clusters of
variables on units (i.e. pots) are more informative and also
can be compared by similar methods, structural factors
were discounted because of the fragmented nature of the
pots. The calculated indices of similarity are given in the
lower left portion of the matrix (Table 26) and the indices
of dissimilarity are shown in the upper right part of the
same table.

The taxonomic dendrogram in Figure 6 illustrates one
method of arraying the excavated areas in a hierarchical
classification according to the weighted variable group
method of Sokal and Michener (1958). Capital letters rep-
resent primary groups established at the initial approxi-
mation. As anticipated the static relationships represented
in the dendrogram segregate themselves into two separate

126 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

clusters, a Grand River and an Orr. One of the conclusions
stemming from an earlier analysis of the Midway site was
that the two areas excavated represent samples of a single
occupation unit over a relatively brief period of time. If
this conclusion is correct and i the Walker-Hooper site has
a single Grand River phase component as concluded in a
previous study (Gibbon 1969), the level of association of
areas within the two phases generates the testable hypo-
thesis, that most Oneota components or large excavated
areas within a site in Wisconsin would associate between
80-85 on the dendrogram. This is a measure of intra-
culture ceramic variability, resulting, perhaps, from
residence rules, marriage patterns, and activity differen-
tiation.

Despite its location in a widely different geographical
area, the Bornick site ceramic sample is actually closer
to some Walker-Hooper site areas (i.e. Group A), then
some are. to each other (i.e. Group A and B). The num-
erical values and the dendrogram sustain the conclusion,
that the site is closely related to the Walker-Hooper Grand
River phase type- site and occurs in or very near its span
of occupation.

SUMMARY

The Borcnick site is a small, probably single family
habitation site near White Lake and the Grand Marsh in
Marquette County, Wisconsin. It is clearly a component
of the Grand River phase, whose heartland is around the
Grand Marsh and along the Fox and Grand Rivers in Green
Lake and Marquette Counties. The isolated position of the
site in woodland well back from the marsh and major water-
ways suggests a defensive and/or winter setting. The
location of the large, stockaded Walker-Hooper site, which
is some distance up the Grand River, a small tributary of
the Fox River, may also be defensive. A wide range of male
and female work activities are evident, particularly fab-
ricating and processing activities and food storage. The
main subsistence activities were the exploitation of aquatic
resources and the hunting of white-tailed deer. Horticul-
ture may also have been a primary subsistence activity of
the inhabitants, but its importance is difficult to determine
from the data. A. single radiocarbon determination (WIS-288)

The Bornick Site

127

ct SS
O H-

Bornick

ct
O

l-t

00
IN)

00
Ox

3KM

H-
?

-J

ON
00

-N3

ct ct

4=-

<ff

0> 4

CO H-

ON
vO

00
ON

00
O

ct t-

CD
4

^
o

[

IN)

ON
U)

ON
4=-

(D
4

H-
O

-xj

-O

00
U)

ro
o

4=-

-N3

^
H

-^3

IN)

4=-

IN)

ro
ro

Bornic

00
Ox

U>
O

4=-

U)
OX

U)
Ox

VjJ

r\>

4=-

Q)
H-
Ct

128 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

places the site in the latter half of the thirteenth century
(A.D. 1290 50).

Grand HIver Phase

Orr Phase

a + b
total
100

4 Bornick

2 M-

Midway
2 M-l

85
80

(

1 ,

(A)

(B)

65
60

Figure 6 . Dendrogram Showing Increasing Degrees of
Similarity between Individual Excavation Areas
(1-5 represent the five excavated areas at the
Walker-Hooper site)

The Bornick Site

129

APPENDIX A

Table 21. Attribute Associations and Measurements of Vessels
(all vessels are plain)

mean width

Class

in mm.

oj a

-H

-P O

V-P

*S,

fc f-i

2 t,

H -H

XII

0,
H

: -P rH

X
0)

o> o

4 s rH

c o

H O

1 5

P 4

1 fe

a. part also in level 1 of units 14 and 22; b. part in level 1 of unit
17; c. part in level 1 of unit 22 ; - = absent information
colors - 1 . yellowish-brown, 2, brown, 3 reddish- brown, 4, gray,
5. black

Table 28. Attribute Associations and Measurements of Miniature Vessels

-p

Class

mean width
in mm.

diameter of
rim (cm.) j

M
M
M

-p

X
M

M
M

should.

H
r-t

17
Sf

2
1
1

6
6
6

1
2

14
9
9

1
1
1

1
1

3
5
5

4
4
4

3
3

2
1
1

fra

1
4
7
9
14

1
1
1
1

7
1
2

6
6
6
6

5
11
10

1
1
1
1
1

1
1
1

3
3
3

2
1

6
6
6

2
2

5
4

1
1

5
5

= undecorated; - = absent information; colors the same as in Table 25

130 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

Table 29. Weight of Shell Tempered Sherds, Distribution of Round Shell
Tempered Necks, Number of Plain Shell Tempered Sherds, and
Decorated Shoulder Sherd Techniques and Elements from the
Bornick Site

unit/feature

-P

round necks
. (with)
plain #
shoulders

decorated shoulder sherds

no.

0) H
03 H
J M

*:"

sherd
design
elements

P 1

NVC

P 2

NVC

383

255

'87

454

144

P 4

NVC

3*3

100 a

P 5

128

113

41 4

1*5

454

=65

~ _

244

10 9 a

170

203

255

111 a

284

198

100

The Bornick Site

131

(Table 29 - Continued)

unit/feature

! i
I"

0) x-^

*$
15

[plain
| shoulders |

1
a

decorated shoulder sherds

no.

0) H

0) >

at H
H H

00,

r4|M

sherd
design
elements

159

573

273

556

154

113

7^3

271

941

411

l a

624

210

947

339

1429

331

132 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

Table JO Attribute Associations of Unattached Shell Tempered Lips
and Rims from the Bornick Site (all measurements in mm.)

unattached rims

unattached lips

-P
I

-P

r-l

o a

1 a

1 9

1 ^

1 b

The Bornick Site

133

(Table 30 - Continued)

unattached rims

unattached lips

i-^

I
i

H
0)

rH
O

M
M

3
O

M
M

01
0)

rH
o

0)
CO

0)
00
4

-a

M
M

CO
CO

rH
O

M
M

X
o

rim thickne

lip thickne

H
rH

M
M

r- 1

lip thickne

C 1

1 1

1 a

134 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

(Tafole 30 - Continued)

unattached rims

a - classes VII and VIII

b - classes IX and X

tt>

0)
B>

M
P

M
M

rim thickne

Lip thickne

c - one sherd with hole
drilled in from
exterior

- undec orated

- - absent information

2
2
2
2

1
1
1

25
30
27
29

7
8
8

8
8

4
6
5

note: the width of most
lip fragments was not
measured.

1 D

Table 31 . Measures of Dispersion for Unattached Rims from Large Jart

1 5

mm i l

1 . rim height 24.0 -0.27 9.67 0.60

2. lip width

3. rim width

4. neck width

5.0 -0.49

7.1
7.7

U 3

93.

h<z

5-51

1.20

0.47

-1.05 1.57 -0.16

1.45 3-8

2.46 4-10

o3
56
57
53

0.28 2.17 0.53 ^.70 4-14 43

The Bornick Site

135

Table 32 . Weight of Stone Debitage, Fire-Cracked Rock,
and Miscellaneous Stone Debitage from the
Bornick Site

;

unworked.

unworked

chert

qunrtzite
flakes

chert
chirks

-P

H
4*

.H w

o
o

* .

w ^

r .

4-5

^_(

4->

^V^

i i

r-i

O O G

-P

G 5i <H

i-t

r-i

F 1

6.5

F 2

18.0

F 3

50.7

78.1

15.

F 4

41.1

P 5

8.1

5.2

23.7

1
2.

53.1
7.0

8
1

18.4

1.9

33.1

32.6

1.5

17.9

1 !

3^8

57.7

13.

31.3

6c 2

33.2

51.1

4.5

12.8

52.4

23-9

8.0

139.2

17.3

9.5

136 WISCONSIN ARCHEOLOGIST

Vol. 52, JNO. 3

Table 33. Provenience of Sub-Categories of Utilized
Flakes from the Bornick Site

side retouch

end retouch pide+end

. <1>

uni facial

bifacial

uni facial

bifa.unif.

bia.

single

both

single

sin.

both

si bojsin. [both

xl*

r 1

4-^

; ^

"X^

f-\

. *

F I

F 2

F 3

F 4

P 5

1^-

BIBLIOGRAPHY

Bidwell, O.W. and F.D. Hole. 1964. An Experiment in the
Numerical Classification of Some Kansas Soils. SOIL
SCIENCE SOCIETY OF AMERICA PROCEEDINGS, 28
(No. 2): 263-268.

Binford, Lewis R. and G.I. Quimby. 1963. Indian sites and
chipped stone materials in the northern Lake Michigan

The Bornick Site 137

area. FIELDIANA ANTHROPOLOGY, 36(No. 12):277-307.

Bray, J.R. and J.T. Curtis. 1957. An Ordinationof the upland
forest communities of southern Wisconsin. ECOLOGY
MONOGRAPH 27: 325-349.

Cleland, C.E. 1966. THE PREHISTORIC ANIMAL ECOLOGY
AND ETHNOZOOLOGY OF THE UPPER GREAT LAKES
REGION. Anthropological Papers of the Museum of
Anthropology, University of Michigan, No. 29. Ann
Arbor.

Cook, Ina. 1851. Land Survey of Montello Township, Mar-
quette County. MS, Office of the Clerk of the Commis-
sioners of Public Lands, Madison, Wisconsin.

Curtis, John T. 1959. THE VEGETATION OF WISCONSIN.
University of Wisconsin Press, Madison, Wisconsin.

Gibbon, Guy E. 1969. The Walker-Hooper and Bornick
Sites: Two Grand River Phase Oneota Sites in Central
Wisconsin. Unpublished Ph. D. Dissertation, University
of Wisconsin, Madison.

Gibbon, Guy E. 1970. The Midway Village Site. THE WIS-
CONSIN ARCHEOLOGIST, 51 (3): 79-162.

Goodall, D.W. 1954. Vegetational Classification and Veg-
etational Continua. In FESTSCHRIFT FURERWIN AICH-
INGER ZUM 60 GEBURTSTAG. Sonderfolfe der Schrif-
tenreihe. Angewandte Pflanzen-soziologie. 1:168-182.
2 vol.

Griffin, J.W. 1946. The Upper Mississippi Occupation at
the Fisher Site, Will County, Illinois. Master's thesis,
University of Chicago, Chicago.

Hall, Robert L. 1962. THE ARCHEOLOGY OF CARCA-
JOU POINT. University of Wisconsin Press, Madison.

Hole, F.D. and M. Hironaka. 1960. An experiment in or-
dination of some soil profiles. SOIL SCIENCE SOCIETY
OF AMERICA PROCEEDINGS, 24:309-312.

Hurley, William H. 1966. The Silver Creek sites (47-Mo-
1 to Mo-5): a complex of five < Woodland site locations
in Monroe County, Wisconsin. Master's thesis, Uni-
versity of Wisconsin, Madison.

Leechman, Douglas. 1951. Bone Grease. AMERICAN AN-
TIQUITY, 16: 355-356.

Sokal, R.R. and C.D.A. Michener. 1958. A Statistical
Method for Evaluating Systematic Relationships. UNI-
VERSITY OF KANSAS SCIENCE BULLETIN, 38: 1409-
1438.

138 WISCONSIN ARCHEOLOGIST

Vol. 52, No. 3

A CACHE FROM JACKSON COUNTY
David Kriesel

A cache of 74 blades was found in a sand bank near the Black River, North

Bend Township, Jackson County. They are of quartzite and range in size from

1/4 to 7 1/2 inches. They were lying one on top of another stacked like

ordwood. All are in perfect condition and show no signs of use. Such caches

e of a Late Archaic tradition dating some time during the first millennium B.C.

The Bookshelf 139

THE BOOKSHELF

THE NAOMIKONG POINT SITE AND THE DIMEN-
SIONS OF LAUREL IN THE LAKE SUPERIOR REGION
by Donald E. Janzen Anthropological Papers, Museum of
An thro* sy University of Michigan, No. 36, 17 Tables,
21 ^AIGO. 141 Pages. Ann Arbor, 1966. $3.

fhir monograph will be of much interest to students
of Wisconsin prehistory. The Laurel material from Naom-
ikong Point as well as that from two sites on the Door
Peninsula of Wisconsin (Mason, 1966 and 1967) are southern
examples of what has been called "Lake 'Forest Middle
Woodland" or "Northern Middle Woddland." Laurel ex-
tends eastward from central Saskatchewan at Squaw Rapids
(a site briefly i.oted by the reviewer) to northern Ontario.
Point Peninsula II of New York State shows some cultural
similarities. Laurel .culture seems to have been an adapt-
ation of Middle Woodland peoples to a Lake Forest ecology.
Hopewell, such as that of Illinois, probably influenced Laurel
at least along the southern edges.

Janzen's site is located to the west of twin points on
Naomikong Point on Lake Superior's south shore in the
Upper Peninsula of Michigan, Chippev Cor^ r Two ex-
cavation seasons were spent at the t^ue, 1 and 1967,
with a preliminary survey test ma^e during the previous
year; two of the seasons were sponsored b\ tjio National
Science Foundation. The author directed the Hst season's
work.

The report is divided into five parts: Introduction,
excavation, artifact description, interpret ~Lou, and Northern
Middle Woodland relationships. In addition there are two
appendices: I. Chi-Square Tables for Ceramic Attribute
Associations, and II. Correlation Coefficients over .80
for Artifact and Non- Artifact Relationships.

The Introduction briefly covers the "History of In-
vestigations" and adequately describes the geology, flora
and fauna, geography, and prehistoric physical setting. A
map is included to show the "Biotic Provinces of the
Upper Great Lakes," the same used in other Michigan
reports (note remarks regarding this map in earlier Book-
shelf reviews concerning Wisconsin) and places the site
in the Canadian biotic province, a Lake Forest floral
assemblage. It is suggested that the earliest possible date

140 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

for occupation of the shoreline site is 250 B.C., since the
present shoreline came into existence around that date.
Later the author puts the site in the upper limits of Laurel's
temporal span from the time of Christ to about A.D. 450.
A radiocarbon date of A.D. 430 (M-2055) was obtained
from charred "food" on a Laurel pseudo-scallop shell
Naomikong Point variety vessel.

Although a small amount of Late Woodland cultural
material was present, as well as historic Chippewa traces
dating around 1850, the majority of the artifacts could be
classified as Laurel, and the analysis of the latter is the
basis of the report. No other Laurel site matches the yield
of ceramics at Naomikong Point. Large areas of the site
were excavated to reveal "group size, subsistence pat-
tern and settlement pattern, and exploitation of natural
resources."

The author feels there is "very little of the right kind
of data from other archaeological sites to test this model"
or probing the cultural dimension. Attempts were made to
correlate task areas with artifact and non-artifact distri-
bution. Five 20 by 20 foot units were used for this purpose,
either contiguous, or connected with lateral trenches, some
extending into the beach. No instances of stratification sep-
arating midden zones occurred except one discontinuity. A
total of 47 features were identified, described and their
contents listed, and their locations plotted on an excavation
map. The author believes the "post molds and features...
may represent the fish smoking areas of the site."

Ceramic descriptions were made so both, attribute and
typological analysis approaches would be useful. The author
criticizes J.V. Wright's "Laurel Tradition and the Middle
Woodland Period" (National Museum of Canada, Bulletin
217) for using only an attribute analysis. A table is included
showing the frequency and distribution by excavation unit of
the decorated body sherd attributes. He makes a valuable
observation that there are idiosyncrasies of degrees of
variation in single pots regarding variation in lip, rim, and
body thickness, as well as rim profile, that would foil the
archaeologist in pottery classification. Only partially re-
constructable vessels were recovered at the site. These
reveal a conical shape with rim diameter about three-
fourths of vessel height. Five ceramic types and one
variety were described in the Middle Woodland assem-

The Bookshelf 141

blage with type designated as Laurel Psuedo- scallop, Lau-
rel Linear Stamp, Laurel Plain, Laurel Dentate, Laurel
Incised; and Laurel Pseudo-scallop shell-Naomikong Point
Variety. The Late Woodland Pottery was Mackinac Ware.
The author's hypothesis is "that dentate and simple tool-
decorated pottery represent a cultural change through
time, and that the latter succeeds the former."

The stone industry is divided into finished stone arti-
facts and chippage debris that is also analyzed. Of interest
are the six small unworked obsidian flakes found in the
midden that were obtained from Yellowstdne Park's Ob-
sidian Cliff; all other material is derived locally.

Only 49 projectile points were recovered. These were
side-notched and stemmed, one large ^nyder's point of
Illinois Hopewell and diagnostic side-notched points with a
flaring convex base that resembles the Stott Corner-notched
point of Manitoba. There were also Late Woodland triang-
ular points present. Other stone artifacts were preforms
(more commonly called bifacial blades), scrapers (with
nearly two thousand examples and only a sample adequately
considered in the analysis), five drills, retouched flakes,
net sinkers, hammerstones, celts, and pestles. Copper
awls, beads, cones and possible fish hooks and a possible
tip of a bone harpoon were other non-lithic artifacts.

Faunal remains revealed that mammal bone exceeded
fish bone by a ratio of 15 to 1 in weight, yet the large
number of net sinkers suggested a fishing site and that the
fish bone had disappeared in the acid soil. The author
concludes regarding settlement pattern that "rather than
a community-wide cooperative fishing effort, it appears
more likely that each household, or extended family, acted
as an independent unit with net sinkers (or nets) kept in the
; vicinity of their living areas . . .this type of fishing par-
allels the Chippewa fishing practices."

In the last section the site is compared with other cul-
tures of the Northern Middle Woodland: Laurel, North
Bay, Saugeen, and Point Peninsula. There is an inter-
action of these people as well as influence from the Hope-
well interaction sphere. All Laurel sites are adjacent to
either lakes or rivers suggesting favorable fishing lo-
cations. There is a high frequency of side and end scrapers
substantiating that unifacial tools are more frequent at
fishing stations. All Laurel sites have "a ceramic assem-

142 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

blage which derives its decorative elements from a common
attribute pool." Stone material is converted into finished
artifacts on the sites. Sites are of great variability in size
suggesting seasonal economy with large fishing camps
during spawning season.

The excavation of this important site has been of high
value to Great Lakes archaeology and the analysis and
report are models to follow.

Thomas F. Kehoe,
Milwaukee Public Museum

THE PREHISTORY OF THE BURNT BLUFF AREA, As-
sembled by James E. Fitting. Anthropological Papers, No.
34, Museum of Anthropology, University of Michigan, Ann
Arbor, 1968. 140 pp., $3.

Burnt Bluff is on the southwestern tip of the Garden
Peninsula of Lake Michigan's northern shore. Pictographs
and caves on the imposing limestone cliffs drew the
attention of both summer visitors and archaeologists, and
stimulated the landowner of the Burnt Bluff area to build
stairs giving access to the prehistoric remains. These
stairs facilitated the fieldwork conducted in 1963 and 1965
by University of Michigan crews; the report does not state,
but one may guess, that the stairs increase the likelihood
of ignorant visitors injuring the remains, so that a measure
of salvage was involved in the Michigan crew's work.

This volume on the prehistory of the Burnt Bluff area
comprises two excavation reports, a discussion of the
geology of the area, descriptions of textiles and rock
paintings from the Burnt Bluff Site, and a survey of northern
Lake Michigan lithic industries, plus introduction and
summary by the volume's assembler, James E. Fitting. The
publication's printing is clear and attractive, and the plates
are good except for those illustrating Fitting's paper on the
lithic industries, .where too much contrast in the photographs
obscures the outlines of the artifacts. Michigan's proof-
reader let slip several spelling errors, but the meaning of
the text is not affected.

"The Spider Cave Site," by Charles E. Cleland and G.
Richard Peske, will most interest readers of this journal.
The cave, ninety feet long by thirty feet deep, is twenty
feet above the lake in a bluff 150 feet high. Geological study

The Bookshelf 143

suggests the cave was formed some 4000 years ago. Cleland
and Peske made test excavations in the cave in 1963,
supplementing their finds with the collection from the cave
lent by the landowner, Mr. Henry Lang. Since the rocky
debris on the floor of the cave showed no stratigraphy, Mr.
Lang's material could be massed with the archaeologists'
in their analysis of the cave's occupation.

Nearly all the artifacts from Spider Cave were pro-
jectile points. Many of the points had shattered tips, and
the flakes detached upon impact were found close to several
of the tip-shattered points, demonstrating that the points
had been shot into the cave. This circumstance, with the
preponderance of Middle Woodland points and lack of
ordinary habitation remains, led Cleland and Peske to
conclude that tL^ cave had mystical significance to the
Middle Woodland Indians, that they shot projectiles into it
as they passed in their canoes, as Indians were observed to
do in the eighteenth century when they passed a picto-
graph-adorned cliff on Crooked Lake, Minnesota.

Dating for Spider Cave could only be performed by
comparison of the points from its floor with specimens
excavated from stratified occupation sites in the region,
particularly the sites described by Mason on the Door
Peninsula. A Middle Woodland period for most of the
points is supported by the presence of socketed toggle head
harpoons, which Mason believes to be diagnostic of Middle
Woodland on upper Lake Michigan. Cleland and Peske
divide the bulk of the Spider Cave points into six types,
some including more than one variety. They use form as their
principal criterion, and give each type a local name, though
they point out comparisons with other named types. In an
extended discussion of the points' metric attributes, the
authors explain their weighting of form over measure-
ments, in classification. Their analysis is clear and useful,
if not particularly ambitious.

Donald E. Janzen's report on "Excavations and Survey
at Burnt Bluff in 1965" concentrates on the work at cave
B-95. Only fourteen artifacts were found in Janzen's in-
vestigation of this inhabited shelter, but the inventory
includes a bark rope, two bark fabric fragments (one
plaited, one twined), a fragment of European wool cloth,
four wooden artifacts (all of problematic function), and
two projectile points (one corner-notched, one side-notched).

144 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

In addition, human bones attributable to two infants, a child,
and three young adults were recovered. Faunal analysis of
other bones, conducted to the meticulous depth we have come
to expect of Cleland and the other younger archaeologists
from Michigan, revealed that the shelter was probably
used for a short period in the spring by a small group of
hunter-fishers.

"The Burnt Bluff Rock Paintings" are discussed by
Douglas W. Lugthart. Relying heavily upon Dewdney and
Kidd's INDIAN ROCK PAINTINGS OF THE GREAT LAKES
(Toronto, 1962), Lugthart considers the paintings to be of
Ojibwa origin. Several motifs seems referable to Mide-
wiwin symbols, but others are either too simple or too
poorly preserved to be thus interpreted. Two paintings
showing human figures with curved lines extending from
phallic projections are denoted by Lugthart as men urinating,
but could as likely be men ejaculating or men dancing wearing
breechcloths decorated with an animal tail. As Lugthart
realizes, the only certain conclusion to be drawn from these
pictographs is that to our eyes their meaning is ambiguous.

Fitting's paper on "Northern Lake Michigan Lithic
Industries" compares ten sites, five from the Burnt Bluff
area and five elsewhere in Michigan. Placing lithic analyses
in the context of ecologoical settings, Fitting uses his
familiar contrasts between the Chippewas' northern hunting -
fishing economy and the hunting- fishing- farming village-
based life of groups such as the Ottawa in the Carolinian
biotic province, assigning the ten small sites to one or the
other pattern insofar as the meager data permit. The
paper is a fine example of how archaeological data can be
viewed as, in Gordon Childe's metaphor, the fossils of human
societies, not merely objects for classification.

These four papers, with Prahl and Farrand's discussion
of the postglacial history of the Burnt Bluff area and
Volney Jones' short report on the fabrics discovered by
Janzen, add up to a stimulating review of the prehistory of
a typical northern Great lakes region.

Alice B. Kehoe,
Marquette University

1971 Award Winners

145

p I f 1 1

* *

| * S-

l|3|

?ila f i

"I!

146 WISCONSIN ARCHEOLOGIST Vol. 52, No. 3

BOOKS RECEIVED

DRAWING ARCHAEOLOGICAL FINDS by Conant Brodribb.
Association Press, N.Y. 1971. Price: $4.95.