k

F LIBRARY
"Wilson State
I University J

L

‘0

‘1

——

 

 

fl

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

fl i
_______1
Lg;
rim,

MSU Is An Affirmative Action/Equal Opportunity Inmnmion
cmma-oi

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Comp

Composition Analysis of the O'Neil Site Ceramics:
A Study of Raw Material Use
in Northwestern Lower Michigan

By

Janet Stouffer Dunn

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Anthropology

1992

Con

JJl/_ '. I,
5 ’1' 0")" 0

ABSTRACT

Composition Analysis of the O'Neil Site Ceramics:
A Study of Raw Material Use
in Northwestern Lower Michigan

By

Janet Stouffer Dunn

In this study the chemical compositions of prehistoric ceramic vessels from the
O'Neil Site of northwestern lower Michigan were determined by Instrumental Neutron
Activation Analysis (INAA) in order to test whether the residential patterns noted during
and after excavation were apparent in the pottery paste. A secondary purpose was to
determine whether an intensive sampling and analysis of pottery from a single intermittently
occupied site could provide meaningful information about the behavior of the prehistoric
inhabitants of the Upper Great Lakes.

The results of the analyses of pottery and clay samples from the O'Neil Site, as well
as clay and temper samples from the nearby Skegemog Point Site, suggest that simple
relationships between raw clays and prehistoric pottery in the Great Lakes do not exist.
Several possible explanations for this are considered, and the need for additional
information regarding the composition of geologic clays and archaeologically—derived clays
is discussed.

Copyright by
JANET STOUFFER DUNN
1992

As with mos
of many individu:
chair, William A‘
m Cornplcu'on in
When, Kcnne
PM)» aft cvider
iIISU'umcnm role
at the Michigan 5

Ofcrifical i
mm“ by Edv
Memorial Phoer
Mam“ for Res
PYOjCCt, who C(x

The may:
Michigan Menu

ACKNOWLEDGEMENTS

As with most research projects, this one has benefitted from the input and assistance
of many individuals. Foremost among these has been my advisor and thesis committee
chair, William A. Lovis, Jr., Ph.D., who not only guided this research, but also facilitated
its completion in many ways. In addition, the contributions of the other committee
members, Kenneth E. Lewis, Ph.D., Helen P. Pollard, Ph.D., and Norman J. Sauer,
Ph.D., are evident throughout this thesis. Michael Hambacher, Ph.D. also played an
instrumental role by kindly re-typing all of the known vessels in the O'Neil Site collection
at the Michigan State University Museum.

Of critical importance to the completion of this study was the technical support
provided by Edward A. Birdsall, Engineering Technician, Nuclear Reactor, Michigan
Memorial Phoenix Project. His assistance was facilitated by Philip A. Simpson, Assistant
Manager for Research Support Activities, Nuclear Reactor, Michigan Memorial Phoenix
Project, who coordinated the analysis at the reactor.

The analysis of the samples by INAA was performed with funds provided to the
Michigan Memorial Phoenix Project, University of Michigan by the U.S. Department of
Energy under a reactor sharing grant. Supplies used in the course of the analysis were
purchased with funds made available by the Michigan State University Museum,
Anthropology Division.

Thomas A. Vogel, Ph.D., Department of Geological Sciences, Michigan State
University, and Randall Schaetzl, Ph.D., Geography Department, Michigan State
University, graciously reviewed the initial results of the analysis and made useful
suggestions for further investigations. Kathy Welch, of the Center for Statistical
Consultation and Research, University of Michigan, provided assistance with the use of the

iv

S YSTAT staistical

 

Several frienc
Michigan have hci;
Minim-Davigv R
program also mad:
“13 presentation.

My. and

en“mgmlent of

SYSTAT statistical program

Several friends and colleagues at the Reproductive Sciences Program, University of
Michigan have helped me with their expertise and advice. Among these are Holly
Anderson-Davis, Ruth Lum, and Michael Muha. The administrative personnel of this
program also made available to me their computers and software for statistical analysis and
data presentation.

Finally, and most importantly, this research was helped by the quiet support and

encouragement of my husband, John David Dunn. For this support I am ever grateful.

 

List of Tables
List of Figures
Introduction
Background
Dcscriptior
Descfiptior
Experimental
EXperimel
HYPOIhesc
Sample

Colleen,

33mph“!
Method

TABLE OF CONTENTS

Page

List of Tables . ........................................................................... viii
List of Figures ........................................................................... ix
Introduction ............................................................................... 1
Background . .............................................................................. 2
Description of Problem ............................................................. 2
Description of Site ................................................................... 7
Experimental Approach and Hypotheses ........................................ 15
Experimental Approach .............................................................. 15
Hypotheses and Assumptions ...................................................... 15
Sample ...................................................................................... 19
Collection . ............................................................................ 19
Sampling Procedure ................................................................. 28
Method . ..................................................................................... 38
Sample Preparation .................................................................. 38
Analytical Approach ................................................................. 44
Results ...................................................................................... 48
Data Evaluation ....................................................................... 48
Data Analysis . ........................................................................ 52
Summary of Data . .................................................................... 72
Validity of Hypotheses .............................................................. 79
Summary ................................................................................... 82
Errors in Original Assumptions .................................................... 82
Significance of Results .............................................................. 90

vi

 

Future Dir

Notes ............
Appendix A
Concentratii
Appendix B
Concentrat
Appendix C
Concentrat
Appendix D
Centroid 1
Appendix E
K‘means
ADDendix F
Elements

TABLE OF CONTENTS (cont'd)

Future Directions .....................................................................

Appendix A

Concentration and Standard Deviation Results (All Elements) .................

Appendix B

Concentration and Standard Deviation Results (11 Elements) ..................

Appendix C

Concentration Values for 11 Elements, in Increasing Order ....................

Appendix D

Centroid Linkage Dendrograms Using 10 of 11 Elements ......................

Appendix E

K-means Clusters Using 11 Elements .............................................

Appendix F

Elements Detected by INAA and Used in Pottery Paste Analyses ..............

References

eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

93
95

98

130

138

144

155

172
177

 

Table l - Vessel.
Table 2 . Vesse
Table 3 - Sherd j
Table 4 - INA}
Table 5 - COHCer
Table 6 - 51mg}
Table 7 . Concem

DUPIiC:
Table 8 ‘ Multi;
Table 9 . Mean

LIST OF TABLES

Page
Table l - Vessels From the O'Neil Site (20CX18) ....................................... 20
Table 2 - Vessels and Sherd Information .................................................. 22-25
Table 3 - Sherd Provenience and Sampling Information ................................ 29—35
Table 4 - INAA Batch Information ......................................................... 39-41
Table 5 - Concentration of Elements in NBS Standard 1633a .......................... 43
Table 6 - Energy Lines and Half-Lives Used in Calculations ............................ 47
Table 7 - Concentration and Ratios of Selected Elements From
Duplicate Samples ................................................................. 61
Table 8 - Multiple Samples Taken From Single Vessels ................................. 67
Table 9 - Mean Iron Concentrations, in Increasing Order ............................... 89

Figure 1 . Map
HEW: Z - Ma;
Figure 3 - O'Ne
Figure 4 - Cenm
Figurc 5 _ Sche
Figure 6 _ Sch:
am 7 _ le
Figure 8 - pk“
HERE 9 . p10,
Fig“ 10 - P14

LIST OF FIGURES

Page
Figure 1 - Map Showing Location of O'Neil Site ....................................... 8
Figure 2 - Map of O'Neil Site .............................................................. 9
Figure 3 - O'Neil Site Showing Areas A and B ......................................... 10
Figure 4 - Centroid Linkage Dendrogram Using 11 Elements ......................... 53
Figure 5 - Schematic Dendrogram of Groups 1-6 ....................................... 55
Figure 6 - Schematic Dendrogram of Group 1 ........................................... 57
Figure 7 - Plots of Ratios for Samples 8-10 .............................................. 62
Figure 8 - Plots of Ratios for Samples 53-54 ............................................. 63
Figure 9 - Plots of Ratios for Samples 23-25 ............................................. 64
Figure 10 - Plots of Ratios for Samples 33-36 ........................................... 66

INTRODUCTION

For archaeologists the study of behavior involves extracting information from the
remainsofmaterialculture. 'I'niscanbeaccomplishedbyexaminingpattemsofartifacts
and land use at archaeological sites, or by intensively extracting information from
individualartifactsa'groupsofartifacts. Inthisstudythelatterapproachismkeninan
attempt to discern patterns of behavior relating to the manufacture, use, and discard of
pottery vessels, and the relationship of this behavior to raw material (clay) acquisition in the
early and late Late Woodland periods in northwestern lower Michigan.

Clay and pottery samples from the O'Neil Site, a stratified Late Woodland through
prouohistoric occupation site, as well as additional samples from a nearby site, were
analyzed compositionally in an attempt to discern whether habitation patterns at the site
could be seen in the concentrations of elements in the pottery samples. Intensive within-
siteelemental analysesofpotteryfromsingle sitesintheUpperGreatLakes have notbeen
previously performed. Hence this investigation was undertaken in part to determine
whether such an intra-site analysis could provide information about the use of clay
resources within a small region.

The results ofthis study indicate that compositional analyses ofpottery are far more
complex than originally believed. Not only do the individual components of pottery paste
(clay and temper) each contribute to the analytical results, but the natural variability of the
clay and tempering material also appear to affect the results. These factors may have little
impact on the outcomes of larger regional studies where the variability of raw materials
between regions may be far greater than that within the region. But for the analysis of
pottery and clay collected from closely-spaced sites, the natural range of variation of the
mwmatefialsusedmmanufacnnemepomerybccomesaucialmthemtapmmdonofme
results. Likewise, the effects of mixing varying amounts of clay and temper to produce the
final paste must be considered. These factors are detailed in the following sections with the
hope that future analyses of pottery paste will address these concerns.

1

BACKGROUND

Description of Problem

The interaction of human populations and the environments in which they find
themselves has recently been the subject of much anthropological study (e.g. Orlove 1980;
Vayda and McCay 1975; Moran 1979). In the Great Lakes this interest in the ecological
approach has often been reflected in archaeological studies dealing with food resources and
the carrying capacity of certain environments (Cleland 1966; Yarnell 1964). The emphasis
on food availability, procurement, preparation and storage is in large part a reflection of the
importance of food procurement in prehistory. However, it also reflects the relative
abundance of food-related artifacts in archaeological sites in the Upper Great Lakes region,
where projectile points, animal bone, floral remains, and in later sites, ceramic materials
dominate the prehistoric artifact assemblages of this region (Fitting 1975; Griffin 1983).

Non-food resources utilized by prehistoric hunter-gatherer groups have also been the
focus of research in the Great Lakes area. Predominant among these are studies of the
availability, acquisition and use of chert (e.g. Ellis 1969; Luedtke 197 6; Wahla 1981) which
was fashioned into projectile points, knives, scrapers and other tools (Frtting 1975). Others
have addressed the role of exotic materials in the material culture of the Great Lakes Indians,
and the importance of trade in the social and economic networks in prehistory (e.g. Griffin
1983; Brose, er al. 1985).

Only recently have the "everyday" utilitarian items such as clay cooking and storage
vessels been examined in terms of resource utilization. Although the prehistoric pottery of
the Great Lakes has long been examined on stylistic grounds in an attempt to establish
regional chronologies, to define cultural groups, and to identify spheres of social
interaction, only in recent years has the chemical composition of the pottery vessels been the
subject of serious inquiry (Trigger, er al. 1980; Clark 1991). In part, this stems from the

inability, until recently, to analyze clays for minute traces of elements which might
2

 

 

individual occupa
whenever vessels
COiflCidcnt Howe
allliyses of pom
in the analysis of

The study

fittest Pottery:

3

distinguish one clay source from another. It also results from the unstated assumption that
where clayisubiquitous —-asitisinMichiganandotherareaswithan abundanceof
glacial deposits (Dorr and Eschman 1977: 134) -- pottery-making took place in siru at
individualoccupation locales. Ithasbeen assumedthatpottersmanufactmed theirwares
whenever vessels were mded, and in areas where habitation sites and clay sources were
coincident. However, recent interest in regional studies, as well as successful elemental
analyses of pottery and other artifacts throughout the world. have have spin-red an interest
intheanalysisofdreseculturalmaterials.

The study of pottery floor the standpoint of resource utilization is of particular
interest. Pottery first made its appearance in the Upper Great Lakes region in the Early
Woodland period, or around 600 BC. (Fitting 1973; Griffin 1983), and is found in nearly
all Woodland sites in Michigan (Fitting 1975). Therefore, the presence of prehistoric
pottery in Michigan spans a period of at least 2,000 years. Together these factors permit
pottery sherds to be studied extensively over time and space, or intensively at a single
location or for a single period in prehistory. In addition pottery vessels can have fairly
short use-lives. It is assumed that since the prehistoric pottery of the Upper Great Lakes
was heavy and prone to breakage, it was probably used for short periods of time before
being discarded. In fact, the pottery sherds found in archaeological sites are likely to reflect
relatively short periods in the lifetimes of the potters who manufactured them. As such, the
study of pottery paste composition can provide a snapshot of time, preserving in the
ceramic matrix evidence of behaviors relating to raw material acquisition and its
modification into domestic implements.

Although pottery vessels were used extensively in the Great Lakes region
throughout prehistory, in the seventeenth and eighteenth centuries European metal kettles
began to replace pottery vessels in most Native American communities of this region
(Kinietz 1965), largely because the metal kettles were more durable and better able to
withstand the direct heat of cooking fires (Holman and Egan 1985: 63). Because metal

kettles so quick
sought for prom.
manuacmnng 0
Which each vessr
Indians.
Howeve
as from analog
south wesrcm U
is made from cl
Without m: be
of the South“

4

kettles so quickly replaced pottery vessels, little is known about the types of raw material
sought for pottery-making. the people who took part in the procurement, preparation, or
manufacturing of the vessels, the method of pottery manufacture, the specific use(s) to
whicheachvesselwasputorthesignificanceofpotteryintheculnneoftheGreatLakes
Indians.

However, some information can be gleaned from the artifacts themselves, as well
as from analogy with the pottery-making processes of the native porters of the
southwestern United States. We know, for example, that pottery in the Upper Great Lakes
ismadefromclayandatcmperingmaterialsuchasgrit. Itwasfiredatlowtemperatures,
without the benefit of kilns, resulting in a somewhat hard but brittle ware. Like the pottery
of the Southwest (Bunzel 1972), Great Lakes pottery is believed to have been made by
women. A unique 17th Century description of pottery making among the Huron by the
French explorer Sagard supports this assertion:

But as for our Huron and other peoples and sedentary nations, they had

(and they still have) the usage and the skill of making earthen pots, that they

bake on their hearth; these are very good and never break in the fire, even

though there is no water in them; but yet they cannot withstand humidity or

cold water for long without softening and breaking at the least blow that is

given them, otherwise they last a very long time. The Indian women make

them, taking suitable earth, which they clean and knead very well, mixing in

it a little sandstone [as a tempering material], then the mass being reduced to

a ball, they make a hole in it with the fist, which they enlarge continuously

while heating it inside with a little wooden paddle, as much and as long as is

necessary to complete them; these pots are made without feet and without

handles and are entirely round like a ball, except the mouth which projects

out a little (Kinietz 1965: 47).

However, the prehistoric pottery of the Upper Great Lakes differs from that of the
Southwest in its construction and decoration. Prehistoric pottery of the Southwest was
made by the coiling technique, and was frequently decorated with painted designs (Bunzel
1965). The late prehistoric pottery of the Great Lakes, on the other hand, was prepared

with a paddle-and-anvil technique (Kinietz 1965), and was decorated only with incisions

and other markir
other rrraterial to
Althougl
lakes region is r
”exchanged" bot
Brashler 1981).
adjacent and me
The anal
dealing with pot
ComPOSiIiOn of }
grounds. On at
31%, Which in .

5

and other markings which altered the surface of the pot, but which did not add pigment or
othermaterialtothe surfaceofthevessel.

AlthoughthemleofponeryintheprehistoricexchangenetworksoftheGreat
Lakes region is not well understood, certain elements have been shown to have been
"exchang " both within and between groups (see for example, Trigger, er a1. 1980;
Brashler 1981). In this way pottery styles have served as markers of interaction between
adjacent and more distant groups of people in prehistory.

Theanalysisofponerypasteaddsanodra'dimensiontothelargebodyofdata
dealing with pottery structure and decoration. The determination of the chemical
composition of pottery matrices makes possible the comparison of vessels on non-stylistic
grounds. On a regional scale, this permits the assessment of clay and pottery fiom various
areas, which in turn can provide information about the movement of clay, pottery, or
porters throughout the area of study. This process has been successfully used in the Near
East (Bieber er al. 1976; Gunneweg and Mommsen 1990; Hancock, er a1. 1989) and in
parts of Mesoamerica (Arnold, er al. 1991; Mine, er al. 1989; Olin and Blackman 1989),
but its use in the Great Lakes area is rare. Trigger and colleagues (1980) used X-ray
Fluorescence to compare the pottery of sixteen Iroquois village sites in Ontario. The data
thus obtained suggest that the transporting of pets from one village to another may account
for the presence of stylistically foreign vessels at certain sites. Further analysis indicated
that intra-site patterning at the household level may reveal itself in the composition of
ceramic vessels.

More recently Clark (1991) examined the analytical results of Instrumental Neutron
Activation Analysis (INAA) of pottery and clay samples fiom over 30 late Late Woodland
sites in the Lake Superior region. His results imply two patterns of cultural interaction
within the Lake Superior basin. One of these involves the transfer of finished pots
throughout the region by exchange a importation. The other involves the local
manufacture of pottery of a particular style in various regions of the study area, implying

either the mover
(Clark 1991).

Work or
Great Lakes re g
within-site corn
mm the inter
northern Michig
Whether popula
Vessels locally 1
mt Vcssels mag
mmfm vair
Michigam as w
populadons_

A Seco;
cOultl be Used a

hat/e SquSSt-u.

6

either the movement of potters or the transfer of information over large geographic areas
(Clark 1991).

Work on ceramic composition analysis on a more local level is nearly absent in the
Great Lakes region. Although the analysis by Trigger, er al. (1980) consisted in part of
within-sitecompafisonsnootherwa'kofthisnamreislmown forthisarea. Forthis
reason, the intensive sampling and analysis of pottery fiom a single semi-stratified site in
northernMichiganwastmdertaken. 'I‘hebroadgoalsoftheprojectweretodetermine
whetherpopulationsresiding atthe siteon a seasonal basis manufacturedtheirpottery
vessels locally fiom nearby clays, and whether these vessels could be differentiamd from
the vessels made by short-term occupants of the site. The results of this analysis could
therefore provide information regarding the nature of resource (clay) utilization in northern
Michigan, as well as data related to the movement and settlement patterns of prehistoric
populations.

A secondary purpose of the project was to test whether this method of analysis
could be used at the site level in the Upper Great Lakes area. Studies of paste composition
have successfully been performed on materials derived fiom state level societies (for
example in Mesoamerica and the Near East), whose populations are relatively sedentary
and whose resource bases are well-defined. Paste composition has also provided
information of a regional nature on artifacts fiom band-level societies. However, assays of
pottery from a single site occupied by hunter-gatherers involved in seasonal migrations
have not been performed in this region. The results of an intensive intra-site sampling and
analysis program would therefore provide data regarding the variability in the paste
composition of vessels found at a single site. It would also prove useful in determining
whether, based on such variability, single vessels from one site could be considered
representative of the pottery from that site in larger regional studies of paste composition.

 

Description i

 

The site ct
site excavated by ;
Afllhmpology in l I
Called "Traverse C
Southern pom tr
PM from both t

551765 as an ldtal S}

 

and its manufacmrg

The O'Neil
MW in Charlevo;
archamlogieal invc

mood Creek. Ex:

Description of Site

The site chosen for this analysis was the O'Neil Site (20CX18), a partially stratified
site excavated by crews fiom the Michigan State University Museum, Division of
Anthropology in 1969 and 1971. The site lies along the shore of Lake Michigan in the so-
called "Traverse Corridor" (Levis 1973), midway between the loci of the Northern and
Southern pottery auditions of northwestern Lower Michigan. Because the site contained
pottery from both traditions, each of which spans the Late Woodland cultural sequence, it
serves as an ideal site for posing questions of behavior centered around clay procurement
anditsmanufacnrreintopotteryduringthelateWoodlandperiod.

The O'Neil Site is located in the NE quarter of the SE quarter of Section 7, T 33 N,
R 9 W in Charlevoix County, Michigan (Lovis 1973) (see Figure 1). At the time of the
archaeological investigation, the site was located on an active beach dune near the mouth of
Inwood Creek. Excavations at the site uncovered a partially-stratified Late Woodland
village site which contained ceramics, lithic material, bone and charred wood from the early
and late phases of the Late Woodland period, as well as some ceramics and European trade
items dating to the early historic period (Lovis 1973, 1991). Preceding the initial
excavation of the site, the area to be investigated was sectioned off into ten-foot by ten -foot
excavation units designated by unique unit numbers (see Figures 2 and 3) which are
retained in the catalogue numbers of the artifacts.

Initial work at the site revealed that most of the human activity occurred in one
region of the site, which was further subdivided into Areas A and B on the basis of the
degree of stratification noted during the excavations (see Figures 2 and 3). Area A was the
mac deeply-stratified area of the site and was overlain by a layer of loose, wind-blown
sand up to 0.5 feet thick. Beneath this was a humic zone of deep grey loamy sands 0.3-1.0
feet thick, which was designated Occupation Zone 1. Aboriginal materials and European

 

 

 

//// ,

//
, /

 

 

 

 

 

 

 

/

 

(After Levis 1973: Figure 5)

Figure l - Map Showing Location of O'Neil Site

35 .ezb .e .52 . N 2:3...

 

 

 

 

 

 

 

 

W110

N330 “'90
Wu; \

10

 

 

 

N360
W110

 

 

 

N330W90
W112 5

N430E0
El/2 E

Z

   
 
 
 
 
 
 
 
 
 
  

 

I AREAB

N390W20
5’— E112
nssowzo
‘— E112
N370W20
5—— am
Nsrsowzo
<—E&Wl/2

N340 W20
E 1/2

 

 

Figure 3 - O'Neil Site Showing Areas A and B

(After LOVE n.d.)

11

trade goods dating to the late Seventeenth and early Eighteenth Centtnies were recovered
form this zone. In the eastern parts of Area A an additional level could be discerned. It
was designated Occupation Zone 1b, and consisted of a light grey sand zone extending to
1.0 feet below the humus zone, with a sand lens separating the two zones. Immediately
below the first occupation zone was a layer of sand devoid of cultural material, followed by
alensofblackerganic material. Thisblacklenscontainedcultm'al materialandwas
designated Occupation Zone II. Beneath this zone was a thick layer of water-lain sands
containing black laminae. Carbon-14 analysis of carbonized organic materials incorporated
into these sands provided a date of 905 i 115 B.P., corrected to AD. 1073-1155 (N-
1268). This date is not compatible with the dates obtained for the underlying occupation
zone, and is thought to be caused by foreign materials introduced into the site during the
deposition of the lacustrine sands (Levis 1973: 24—34, 1991: 196-7).

The basal occupation level in Area A, which ranges in thickness from 0.05 to 0.3
feet, represents an intensive use of the site by Late Woodland period peoples. This level,
designated as Occupation Zone III, consists of a series of thin, grey-black sand layers
whose individual lenses were not possible to isolate. Two 14C dates, both of which were
compatible with associated cultural materials, were obtained from charred wood remains
collected from a hearth. These organic materials yielded dates of 740 :t: 100 B.P.,
corrected to AD. 1277 (M-2406) and 670 :l: 100 B.P., corrected to AD. 1283 (M-2405)
(Levis 1973: 34, I991: 196).

The stratigraphy of Area B is less complex, and is roughly the same as the upper
portions of Area A. This area was overlain by tan, wind-blown sands, under which lay a
dark grey loamy zone. Beneath the learn was an occupation zone consisting of black
sands. This zone yielded a 14c date of 430 :l: 100 B.P., corrected to AD. 1441 (M-
2398), for carbon associawd with a hearth designated as Feature 8. Based on the refitting
of sherds from this level and Occupation Zone 11 of Area A, it is thought that the occupation

12

zoneofAreaB maycorrespond totheintermediateoccupation zoneofAreaA (Lovis 1971:
35-39, 1991: 197).

Underlying the occupation zone of Area B were basal yelow-tan sands. Material
for a 14c analysis was obtained from Feature 3, a hearth which was intrusive into this
level. The date obtained for this material was 1000 1:140 B.P.(M2401), corrected to AD.
1004-1019, which was compatible with the materials found nearby in the occupation zone
of Area B (Lovis 1973: 35-39, 1991: 197). In addition to the excavations of Areas A and
B, a series of test pits were also dug to the northeast and southwest of the major occupation
area (Lovis 1973) (see Figure 2).

The depositional and occupational history of the site is one of dune formation
through wind-blown and water-lain sand, interspersed by depositions of cultural materials.
It is thought (Lovis 1973, 1991) that by AD. 1000, parts of Area B were stable, and that
afterthistimeoccupationstookplaceinAreaB andpartsofAreaA. Some timepriorto
A.D. lZWfinthasmbflizafionofAmaAoccmredleadingtotheoccupationofthisareaas
well as of Area B. This was followed by the inundation of Area A and the deposition of
water-bome sands; during this phase no occupations took place in Area A. Subsequently
there was a limited occupation of Area A, after which the area was covered by wind-blown
sands. By the fifteenth century, Areas A and B had become somewhat stabilized, and were
re-occupied. Further wind activity precluded the complete stabilization and formation of
discrete layers of occupation. The later occupations of the site in the ensuing years, along
with the deposition of sand and organic material, covered the site with a modern sand loam.
The final stabilization of Area A occurred in the late seventeenth and early eighteenth
centuries, when groups of historic Native Americans visited the site (Lovis 1973: 39-40,
1991: 197-9). Throughout its occupational history, the site was inhabited on a seasonal
basis (probably dtu'ing the warm season) for varying lengths of time. No evidence of
permanent, year-round habitations were recovered at the site (Lovis 197 3, 1991).

13

Theceramic assemblagefromthesiteoonsistsofpotteryfromtheearlyandlate
LateWoodlandperiod. 'I'heearlylateWoodlandis characterized by Mackinac Ware
vessels and by Skegemog Ware vessels; the late Late Woodland is represenwd by Juntunen
Ware pottery as well as by Traverse Ware vessels (Lovis 1971). The Mackinac and
Juntunen Wares represent wares whose styles reflect the Northern pottery tradition of
upper Michigan. The type site for these wares is the Juntunen Site, located on Bois Blanc
Island in the Straits of Mackinac (McPherron 1967). The Southern pottery tradition is
illustrated at the O'Neil site by the Skegemog and Traverse Wares, both of which were
named for pottery found at the Skegemog Point Site in Grand Traverse County, Michigan
(Hambacher 1992).

A recent re-analysis of the O'Neil Site artifacts from a spatial point of view (Lovis
1991) indicates that the site was occupied in two different ways. The occupations of the
site by groups manufacturing Mackinac, Juntunen and Traverse Ware pottery consisted of
residential, or long-term habitations containing areas related to domestic and/or core-
reduction activities. The occupations of the site by groups using Skegemog Ware vessels,
on the other hand, are characterized by "logistic," or extractive camps. Although all of
groups are thought to have visited the site in part to collect chert from the nearby Pi-wan-
go—ning Quarry, the groups responsible for the Skegemog Ware sherds at the O'Neil Site
were thought to have stayed at this site only long enough to extract the chert and to perform
some basic core-reduction activities. The other occupations of the site are believed to have
been longer in duration and are thought to have encompassed more activities than the
Skegemog occupations. Furthermore, it is presumed that dming the Mackinac, Juntunen
and Traverse occupations, the site would have been utilized by more diverse groups of
individuals, probably representing family groups rather than age- and gender-specific
groups which most likely utilized the site during the logistic forays (Lovis 1991).

Because the two difi‘erent forms of occupation correspond to different pottery
types, this site is ideal for examining the relationship between residency and pottery

 

mnufacnring lx'l
potta'y traditiom

cxamintd from it

pmwidcs the motel
rtlitivcly small n:
Sampling of the pl
set of Salnplcs for

and Utilization

 

14

manufacturing behavior. In addition, the presence of four ceramic types representing two
pottery traditions as well as two time periods permits the paste composition results to be
examined from several points of reference. Also, the presence of raw clays at the site
provides the necessary baseline against which to compare the pottery samples. Finally, the
relatively small number of identifiable vessels found at the site permits the intensive
sampling of the pottery vessels from this site. In short, the O'Neil Site provides an ideal
set of samples for examining, within one site, the relationship between clay procurement

and utilization.

 

l
EXPE

Experiment
The anal}

Itsidcnce and the

This method is ca

 

53mph mmix. It
dim a large nut
die identified ves
Chm“ a“Elysis it

Widens of

Hypotheses
PRVious
two different wa
Trims: and In
Marion (

rtPitts

flit Sit: {01, Stay

EXPERIMENTAL APPROACH AND HYPOTHESES

Experimental Approach

The analytical method chosen for exploring the relationship between population
residence and the use of resources was Instrumental Neutron Activation Analysis (INAA).
'I'hismethodiscapableofdetectingandquantifyingasmanyasSOdifi‘erentelementsinthe
samplematrix. Italsocombineseaseofanalysiswith analyticalaccm'acyandtheabilityto
detect a large number of elements nearly simultaneously. Using this method, a majority of
the identified vessels from the O'Neil Site were analyzed and the data were subjected to
cluster analysis in order to determine the statistical groupings based on the elemental

compositions of the samples.

Hypotheses and Assumptions

Previous work at the O'Neil Site (Lovis 1991) suggested that the site was utilized in
two difi'erent ways: as a seasonally-occupied residential site (represented by the Mackinac,
Traverse and Juntunen pottery), and as a ”logistic" occupation of considerable shorter
duration (represented by the Skegemog pottery). This suggests that the people who utilized
thesiteforseasonalhabitationusedmostoralloftheresourcesavailabletothematornear
the site, including local clays. During shorter "logistic" occupations, however, it is
expected that a more selective use of resom‘ces would have occurred, and that any clay
vessels used at the site would have been manufactured elsewhere and brought to the site as
needed. Further, this pattern is expected to have remained the same for potters of different
pottery traditions throughout the Late Woodland period. These suppositions can be re-
stated as follows:

Groups occupying the site for long periods of time (residential occupants)
would have manufactured their pottery locally from locally-obtained clays

15

16
and temper, while short-term occupants (logistic occupants) would not have

made their pottery locally, but would have brought them to the O'Neil Site

fiom other locations.

Residential occupants of the site at any point in the Late Woodland period

would have made their pottery locally, while logistic occupants during any

sub-phase of the Late Woodland would have made their pottery elsewhere.

This difi‘erential pattern ofpottery-making based on the use ofthe site is

discernible in the chemical composition of ceramic pastes.

In addition, several assumptions about the nature of the materials to be analyzed, as well as
presumptions about the behavior of the prehistoric potters are necessary. The first is that
local and non-local clays are compositionally different fiom each other, permitting
differentiation between local and non-local pottery. Although this assumption seems
obvious, in fact it has not been tested for small areas (less than 100 mile radius) in the
Upper Great Lakes area. For this reason, clays collected during archaeological excavations
of the O'Neil Site and the nearby Skegemog Point Site, a Late Woodland occupation
situated along Lake Skegemog about 40 miles to the southwest of the O'Neil Site
(Hambacher 1992), were analyzed by INAA in order to determine the relative differences in
the composition of clay fiom a nearby site.

The second assumption is that the clays collected during the archaeological
excavations are representative of the clays used to manufacture prehistoric pottery vessels.
Thisassumesthat: 1)theclaysom'cesremainedthesametluoughouttheLateWoodland
period, and were not exhausted during this time, and 2) the clay deposits within the
procmementareaofaparticularsite wereusedin the samemannerandtothe sameextent
throughout the Late Woodland period. It further assumes that the clays collected at the
O'Neil Site and the Skegemog Point Site are native to these respective areas, and were not
simply brought to the site and discarded there. In spite of these assumptions, it is believed
that the use of archaeological clays from the sites more truly represents the actual
composition of the clays used in the manufacture of Late Woodland pottery, since these

clays would hit:

 

manufacture.
about the manufa

waltriscomrnon

 

 

the composition c
Clements to the p:
”31- These assur
1973) was suffici
ShOwn later. this :
“it paste. and in ,
Wading imposs
Occ“pardons iden.
idmfi'y the atthe
Wyn With sim-

Gil/en [ht

338:1

l7

clays would have been obtained and prepared by the individuals involved in the pottery
manufacture.

Ancillary assumptions regarding the nature of the pottery paste include assumptions
about the manufacturing processes and the final composition of the ceramics. Although
wateriscommonlyaddedtotheplastic material, thisisassumednottosignificantly alter
the composition of the clay. Similarly, the addition of temper was assumed not to add
elements to the paste in significant amounts, and was thought to only minimally "dilute" the
clay. These assumptions stemmed fiom the belief that the grit temper in the paste (Lovis
1973) was sufficiently large so as to permit sampling which avoided the temper. As will be
shown later, this assumption was not correct; temper was very well dispersed throughout
thepaste,andinmostcaseswas sofinelydividedastomakeavoidingitinthe sampling
procedure impossible. Finally, it was also assumed that the logistic and residential
occupations identified by Lovis (1991) were correct, and that the pottery sherds used to
identify the archwological culture of the occupations could be re-identified by another
analyst with similar outcomes.

Given the foregoing, the following hypotheses were established for this study:

1: There are no significant differences between the chemical
compositions of the pottery from each of the residential occupations
(i.e. the Mackinac, Traverse and Juntunen pottery).

2: There are no significant differences between the chemical
compositions of the local (O'Neil Site) clay and the pottery from the
residential occupations (i.e. the Mackinac, Traverse and Juntunen
ponery).

3: There is a significant difference between the chemical compositions
of the non-local (Skegemog Point Site) clay and the pottery fiom the
residential occupations (i.e. the Mackinac, Traverse and Juntunen
pom).

 

The validity of tl
MMhm

Site.

 

18

4: There is a significant difi’erence between the chemical compositions
of the pottery from the logistic occupations (i.e. the Skegemog Ware
pottery) and the pottery from the residential occupations (i.e. the

' , Traverse and Juntunen pottery)

The validity of these hypotheses were examined using a total of 52 samples of pottery and
clay m the O'Neil Site, and 7 samples of clay and temper from the nearby Skegemog

Site.

Collection

The oer
dttaii (Inns 1'
researchers 91'
which 777 had
50). Each 0ft
and were subs
“33°15 (Lovis
is given in Ta]
Site (Lovis 19'
number 10 at l
indWits vegs,
however, lac}

The q
the Late Woo

(See Table 1).

SAMPLE

Collection

The ceramic assemblage form the O'Neil Site has been previously described in
detail (Lovis 1973). During the excavation of the site by Michigan State University
researchers 9194 Late Woodland and prom-historic pottery sherds were recovered, of
which 777 had decoration or surface treatment enabling categorization (Lovis 1973: 41,
50). Each of these 777 sherds we identified with an MSU Museum catalogue number 1.
and were subsequently sorted into groups representing a minimum of approximately 80
vessels (Lovis 197 3: 50). The distribution of identifiable vessels reported by Lovis (1973)
is given in Table 1. Although a minimum of 79 vessels have been reported for the O'Neil
site (Lovis 1973), it is believed that subsequent work on this collection expanded this
number to at least 92 vessels, since the Minimum Vessel Sheets 2 for the O'Neil Site
includes vessels with numbers ranging from 1 to 92 (Lovis n.d.). This same document,
however, lacks information for vessel numbers 22, 27, 3, 45, 46, and 89.

The ceramic vessels represented at the O'Neil site consist primarily of pottery fiom
the Late Woodland period, with some additional pottery from a prom-historic occupation
(see Table 1). Within the Late Woodland period, vessels were attributed to the early and
the late phases of the Late Woodland period, as well as to both the Northern and Southern
pottery making traditions in Michigan (Lovis 1973). The Mackinac Ware vessels,
representing the Northern pottery tradition of the early Late Woodland period at the O'Neil
Site, are believed to date to around AD. 800 - 1,000 (Lovis 1973: 59). The Skegemog
Ware vessels hour this site are thought to be roughly contemporaneous with the Mackinac
Ware, and represent the Southern tradition of pottery in the early Late Woodland (Lovis
1973: 63).

The Juntunen Ware pottery at the O'Neil Site is a Northern pottery type. Its

incidence at the site is dated to AD. 1,200 - 1,300 as well as to the fifteenth century AD.
19

[Ill

 

iilil-llilli

EFF Couture:— 7.3.x:

: sated; CLUECOZ N
. . 3: 29003 “364
CQNntCCMu Q ( ~

N

325 a:

 

— 80:0t-050“

 

To a: \ Car-=39. — 5%)
H e \ .2.
. a N UG-LQL Khuuuauah an t:
anvil-P lb .

5

WEEK. 0.3%; V

 

i!

. t Enoch.
Aanuflnv @Nv 02.0 .3770 a... Ear-bu Evacntxv H

 

 

O\
N

225 8. e223 mmEb

 

F‘NF‘N

n 25.

a 25

m 25.
A..§eaa..v a on»...
coca 5;er .oaeé H 25.
E3502:

 

Anew ”m8" «fired

.0 .2: has \ .u a: 22

32.550 :02.0
$860

 

ea a5 ease

82.82 .Q.<

@5883 8a..— 8a—

chasm

Oswv—tv—tv-t

.§< cage—h.
355m $3.5.
eon—£30m 883$.
3383:: 823:.
85055 9.83%.
mmmgé .

 

Cbu "3% «30d

.Q.< .U 52 d cam—8.3 .D.<

3683 23 a:

50582

t‘fl'

n2. d wan cos—“5:.
382:5 .3054 5555—.
752322

 

So "at: ”sea

Ba? 8582 8 08am .cnaofiefiaoo3 33 2.30

 

Eufizom

MN

358 mmaomoam
ae— znahm moaomoam
005on

 

an £63 £66

oouToow .Q.<

 

3683 o3 :5

E0582

v
m

Bodega: 8582
33582: 8:282
375654

 

 

gut—0&0“

825
02m .623

 

Etch

 

85.5...
Ease.—

 

£089»

 

E

8ng as can =oz.o 2: sou..— 288> . _ can...

 

on: 933

 

 

 

 

(lovis 1991'. 2C9
and spans the P-
The abo
archaeological ti
compare the bet
In order to Sim;
Traverse or Jur
SUI>gmup of tl
genera] pone:

0“ Compositit

21

(Lovis 1991: 207). The Traverse Ware pottery, in turn, represents a Southern pottery type
and spans the Period from AD. 1,200 - 1,500 (Lovis 1973: 85).

The above four pottery types representing two cultural traditions as well as two
archaeological time periods were used for the paste composition analysis in an attempt to
compare the behavior involved in pottery making, use, and discard of these four groups.
In order to simplify the data analysis, only the pottery type (Skegemog, Mackinac.
Traverse or Juntunen) was recorded for each vessel number, the variety being considered a
sub—group of the general pottery type. As will be shown in subsequent chapters, even this
general pottery classification proved to be too fine a distinction for the classification based
on composition alone.

After completing an initial inventory of all of the O'Neil pottery available at the time
of the project, a list of the vessel number and cultural designation was compiled. The
original cultural designations for 33 of the 92 vessels was determined using the original
Minimum Vessel Sheets (Iovis n.d.) along with photographs of sherds with type/variety
designations (Lovis 1973). Because complete cultural designations for each of the original
vessels was not available, the vessels were re-typed by Michael Hambacher, whose
designations of the known vessels (Michael Hambacher, personal communication 1991)
closely matched those proposed by Lovis (1973) in the original analysis. However, four
early Late Woodland vessel (Vessel 17, corresponding to Sample 13; Vessel 28,
corresponding to Sample 5; Vessel 32, corresponding to Sample 6; and Vessel 53,
corresponding to Sample 14) were identified differently by Hambacher and Lovis. These
dual designations are maintained throughout this study. A listing of the pottery type
designations given by both Lovis and Hambacher can be found in Table 2.

Pottery types were established by Hambacher for 54 of the 92 O'Neil vessels.
Another 6 vessels could only be given tentative type designations, 6 could only be
identified (or tentatively identified) within the early or late Late Woodland period, and 15
could not be identified at all (Michael Hambacher, personal communication 1991).

‘rl‘ultlL h

 

 

 

 

 

 

new—QCEV. .3: _Umau> Dinosaur—L. ~ N 023: N N M
Ca: .558 £th Eocene—:5: 3.9.: ..o.: erect! b 535:2 535:2
00.07. Unfit—t can); MAKE. 0&3: Q~QEQM4 7.3.83

 

 

:O-uflELAVhp-n nut-03W gel-w Giovanna-ex, l N thnutht

22

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2002088 20: 20mm0> 8052330 2202.0 80523.50 2202.0 0:0: em
2002059. 20: 20m00> 20820003 082 082 3222300 0:0: mm
02082288 88800.25 0 72 0:0: mm
8:382 8:382 N2 2N
2.0222088 20: 20m00> 208220003 082 082 0:0: on
2228:; 09 2:0an 2002.288 20: 20$0> 8:382 8:382 0:0: «02
2002.282 00: 20mm0> 80080.2. 0:0: 222
mofiowoxm 8:382 m2 2.2

8:382 2 2 02

wofiowoxm 2. n2

82mg; 20: 2ommo> 5:852 22.22800 0:0: 2
0282288 00: 2080> 0:0: 2
2.022288 00: 2omm0> 208220003 082 .2280 0:0: N2
i003 cease; om :

5:85;. 5:85;. we 3

2.02.228m 00: 20mm0> 80080.2. 0:0: a
w080m002w m050w002m m 22

20022253 20: 2080> mofiowoxm 2523.022 0:0: 2.
wofiowoxm M080w002m N o

5:35;. mm m

w080m002m w080w002m 2 20

080802. on m

00098.2. 222 N

2002.280 20: 28m0> 00008.22. 0:0: 2

22%: £5.00 .822 .0053535 882 920.: 02.55 03:22.72 03:52
8202 00.2.2. 083 22.2.2. 083 0.055 .0082»

 

528500.25 2:025 20:: 0.20082» . a 038.2.

 

 

. :Ec.‘ 6....)- thcea ...... m a a. 4 . .1
:00. . 2:5 35$ 5 22:. 0.25: \ 2 E v \ 209.3 \
8.372

:0. =53 N 0338

 

 

23

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.02088 :0: 20mm0> 55052:: 0:0: we
20022280 20: 2080> 95052:: 0:0: 2.2g
0282288 5288.825 02 0:0: 3.
0282288 58:80.82 02 0:0: 5.
0838.2. 2.2 3.
2:022:80 20: 2030> 55052:: 0:0: 9.
3&8» 20: 20mm0> 08028.2. 08038.2. 0:0: 9
2002528.... 20: 200m0> 0888.2. 0898.2. 0:0: 22.
2.02.28: 20: 2080> 55052:: 0:0: 9.
2002.28“ 20: 29.002, 08038.2. 0:0: 0m
2002.283 :0: 2020> 08220003 082 082 2328052 0:0: mm
[5:88; 5:882. em hm
0.80.8.2. 0802: .2. 22 em
5:85; 508:3 mm mm
2.02.2280 00: 2800> 5652:: 0:0: 32
0282288 5288823 02 0:0: mm
88.032: 09:00.35 0 mm
5:882. :0:8::M mm 2m
2282:0288 02:: 0022 2000: .0: 038m 5:85: 5:882. av em
2002,28.“ 20: 28m0> 8052380 2202.0 80:22:56 2202.0 0:0: on
wwEowoxm 8:382 m ”N
0282288 5288.525 02 0:0: R
8:382 8:382 02 on
5:02:02 5:88; 2n mm

2232 .5800 .22. £052.85 852 :20... 2.2.522 .3552 85:52

.802 0.2.2.2. 0.83 22.2.2. 0.83 0.085 88>

 

3.22.03 N 035—.

 

  

   

 

 

 

 

 

0.2M2\=fla>a 2o: 2cmmu> -
:00— .EEcu .nnuiuwtuacEEC FN¢~ rt... 0.35
N . .22 L
08072 g 82.25 9.52.» 22.2.5 0.23: QUHMNW késfiz
2.0.30;

 

 

 

 

 

 

 

 

 

A‘.~:°UV

N MiG-uh.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2:02:88 :0: 20mm0> 08028.2. 08022.2. 0:0: N2.

2:02:88 :0: 20mm0> 08028.2. 0:0: 22.
2:02.28: :0: 20mm0> 55052:: 0:0: o:
8:382 8:382 2 ac

2002.28,: :0: 2080> w080me2m 22.2280: 0:0: we
2:02.280 :0: 20mmo> 08028.2. 22.22000: 0:0: 5
2:02:28: :0: 20: 00> 85052:: 0:0: 8
2:02:28: :0: 20mm0> 55052:: 0:0: mm:
2:02:28“ :0: 28m0> 0898.2. 0808.: .2. 0:0: 3
2:02.28“ :0: 20000> 95052:: 0:0: no
08032.2. 3 a

2:02.28; :0: 28m0> 08028.2. 0:0: 20
2002.28: :0: 28m0> 580:0 886:2 :0:>0:U 250282 0:0: 8
2:02:80 :0: 28m0> 0898.2. 08028.2. 0:0: am
0282296 :0: 2030> 0:0: mm
m080w002m wofiowoxm o2 - 22 um

2:02.280 :0: 28m0> mofiowoxm 23280: 0:0: on
2:02.283 :0: 23m0> :Bm.52:= 0:0: mm
0898.2. 2 E

wofiowoxw 8:382 :2 mm

genial am 8

w080m82m 2. 2m

2830:8880: 200m: :0: 838m 080387 on - 02 on
0282282: :0: 2800> 0:0: 9.

2232 .888 .20: £058.83: 88— :8: 02.5.22 :0:—:52 :0:—:52
8:02 22.2.2. 0.8.5 0:3. 0.83 02:85 88>

 

 

 

 

 

25:08 N 02:25.2.

 

 

 

 

 

 

 

 

I! .. .. _ a..fl;lu~d=: ‘ \ 05‘s: \ 3
.5353. .3: .0mmu> — Mczawuxv‘ Azimoaw x \ 0.5: \ Mn 4
:oo— .EEe imam Coco-550:» at: :3... £39: Len—=32 L353?
332 09:. 003$ «EAR. 903$ 235$. 39%)

 

 

1‘.

«ELIOUV N Q—Qfirfi

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3388 00: ~800> :0::0::_. :0:::::~ 0:0: «0
00—980 .0: $0005 Baas—0:: AN 093 _.:0£m8_0::_. 0:0: 3
55:3. 5:85;. Vn - mm o0

28:08 8:88:85 02 0:0: 3
33:80 :0: ~0mmo> :0::82. 0:0: xx
5523. On 5

00—980 00: ~080> 5505:: 0:0: 8
00:025. mm - mm mm

:0:::::~ :0::83 on SW

@0383 00: ~800> 5505:: a mm
@0183 00: _ommo> 058003 83 08— 3.3080 0:0: mm
:0::83 \ ‘ mm 5

00—0080 .0: ~800> 220003 003 0:: >368: 0:0: aw
0fi0>flh mm 2.

0328; :0: _0000> 0:0: E.
00:025. 00:90:. an E.

0328.»: 00: ~00mo> 0:0: or
00—0880 00: ~0000> 5505?: 0:0: nh
003800 00: _0mmo> 565:: 0:0: E.
33:80 00: _0000> mofiowoxm 3:38: 0:0: 2.

:03 .5500 .0000 £053.55 Ana“ 90.: 030‘: LEE—.2 :0:—.52
8.02 003. 95$ 09¢. 0.55 0.05—um .008>

 

8.253 N 0.8:.

 

 

Undoubtedly (I
”other" catcgm
and to the “um
Emily, S VtSE
been othmis
shcnis and th
Miami sugg
“mbfi‘S, bu:
Vessel" cour

Bee:
‘0 sample an
mm W6
dCK'Tmine x

were 13mm

26

Undoubtedly many of these 27 difficult-to-identify vessels correspond to Lovis's (197 3)
"other" category which contained 29 vessels represented by sherds too small to classify,
and to the ”unclassified" category, which contained 7 vessels of 5 "types" (see Table 1).
Finally, 5 vessels were not available for sampling due to either being on display, or having
been otherwise removed from the collection. For an additional 6 vessels both the pottery
sherds and the Minimum Vessel Sheets corresponding to these vessels could not be
located, suggesting that these may have been "vessels" which were assigned vessel
numbers, but which were later found to be subsumed by other pots in the ”minimum
vesse " count.

Because the collection represented a small number of vessels, an attempt was made
to sample as many of the identified containers as possible. Of the 54 vessels with known
pottery types, 37 were sampled at least once (see Tables 2 and 3). The criteria used to
determine whether a vessel would be sampled was four-fold First, vessels from Area A
were preferred over those from Area B, since the former was the stratified area of the site,
andthe latterwas not. Likewisevessels fromareaAorB werepreferredoverthosefrom
the test pit area, since little habitation information was available from these test pits.
Second, where practical, a preference was given to vessel whose sherds had complete unit
and level information recorded on the Minimum Vessel Sheets or on the Museum catalogue
cards (Michigan State University Museum n.d.). Third, a representative number of
samples from each of the four pottery types was required. For this reason, samples were
taken from all of the Skegemog and all but one of the Mackinac vessels. (The Mackinac
vessel designated as vessel 19 was not sampled simply because the single sherd
representing this vessel was too small, and sampling would have destroyed the specimen.)
Finally, as far as practicable, the number of vessels sampled from each of the four pottery
types was kept relatively equal. Thus not all of the Traverse vessels were sampled, since
the available vessels of other types was much lower than that available for the Traverse

vessels.

27

The total number of vessels sampled is as follows:

Skegemog 6
Skegemog/ "Problematic" 1
Mackinac / Skegemog 3
Mackinac 4
Traverse 12
Juntunen 1 1

For most of these vessels, samples were removed from only one sherd. However, in the
case of Vessel 57 (Samples 8-10), Vessel 50 (Samples 19-20), Vessel 85 (Samples 23-
25), Vessel 90 (Samples 53-54) and Vessel 11(Samples 29-30), samples were taken from
multiple sherds both to test for the reproducibility of the INAA results across a given vessel
(e.g. Vessels 57, 50, 85, and 90) and to check for the effect of diagenesis in the case of
vessels whose sherds were found in multiple locations (e.g. Vessels 50, 85, and 90). In
addition, because the temper pieces were so large in one vessel (Vessel 11) it was possible
to isolate some of the temper fragments and collect them as a separate sample (Sample 30)
which could then be compared to the corresponding paste sample from the same vessel
(Sample 29). Likewise, when vessel 52 was sampled, it was noted that most of the
resulting sample consisted of untcmpered clay. This "pm-e" clay was collected as Sample
32.

In addition, three clay nodules collected from the O'Neil Site during the 1969-71
excavations were also sampled in duplicate or triplicate (Samples 33-35, 37-40). These
dried nodules proved to be very sandy and broke apart easily when slight pressure was
applied. "Sand" collected from just beneath one of the nodules during the original O'Neil
Site excavations was also sampled for this analysis (Sample 36). Four prepared clay
nodules collected from the Skegemog Point Site during the this site's original excavations
in the 1960's (Michigan State University Museum n.d.) were also sampled in singletons or

 

 

duplical
Sktgcrr

49 (Jur
of 100
pmV'ldt

the pot
check
each k
We
“limb
Table

28

duplicates (Samples 41-43, 45-47, 60). Finally, one piece of presumed temper from the
Skegemog Point Site was also sampled (Sample 44).

A total of 60 pottery, temper, clay and sand samples were collected. Two of these
samples (Sample 8 and Sample 53) were run in duplicate (hereafter designated as S8N1/2
and SS3N1/2, respectively) to test the homogeneity of ground samples and the
reproducibility of the INAA results. Three other samples were not analyzed by INAA.
Samples 19 and 20 (Traverse Ware) were not run due to suspected contamination of the
sample by the adhesive and India ink used in the curation of the sampled sherds. Sample
49 (Juntunen ware) was not run due to insufficient material having been collected. (A total
of 100 - 200 milligrams (mg) of sample was required for the INAA analysis. Sample 49
provided only 46 mg of material.)

In addition to the pottery, clay and temper, each lot of dental bits used to remove
the pottery and clay samples from the artifacts were also analyzed by INAA in order to
check for possible contamination of the samples with filings from the bits. Two bits from
each lot were sampled by breaking off the tips of the bits and running these in a separate
INAA batch (see Chapter 5, Method). A complete listing of sample number, vessel
number, pottery type, sherd provenience, INAA batch, and bit lot number can be found in
Table 3.

Sampling Procedure

The artifacts used in this analysis were sampled by grinding off material from the
inner surfaces of the sherds with Tungsten-Vanadium hardened steel dental bits (Pfingst &
Company, Inc., South Plainfield, NJ). The dental bits, which were cleaned with disnlled
water and powder-free tissues, were used in conjunction with a Dremel Tool for the sample
collection. The sherds were prepared for sampling by first removing 1-2 millimeters (mm)
of surface material in a 1-2 square centimeter area and discarding this material. The

 

 

 

 

 

 

+5.... accz ~ “Qt-(q r um“. L * U.~:Q.—-A\:ahl- ‘1 S...

:O-aflshehn: MC=QEQm 3:6 99:92:20th 3.825%. n R4 bséfirh

 

29

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$305
2 80
22 be, pa; 8: 333. m 35. an? 82 m 8:382 030% a N:
:0» 83-80:: 0::
52: as. 22 be. use a: . £094 a E2 02: 092 see 85%: 0.8.83 8 :
a: £83. 25525 83 :22 m moaowem S0303 R S
a: 0083. m es. :9: 82 < 0.080030 Slamfle R a
a: :33 m .22 8%? < 08509.0 220% R E
: 25:
case 5:0 088 00> a: .. ”was. .980 EB :52 < 085030 «.8080 E a
.12.. no.5
8: «was. 322 8b: :an m moaommm 0.00% mm 0
I. woxm
8: _. 2303 N 32 E 82 0: 02.282 Spawn an n
3—
s: :83 N :32 :3: 02.2 080 0°50on 32mg 2 a.
:8 323 < 250 so:
a: .. 0803. m 3.2 on; :22 m: "8.000% 0.8%on 0 m
m:
22 0232 2S 5: 0803. m :32 9.8 82 as: 05009.0 09.003... 0 N
A8: 00
8200
:8: _. 0803. 22.2 at: 802 < 38000010 meals: v :
Base: 33 t. .05.. :5 32 25. 35 e n
E. 882 8:. :0 a. 8:25:80 08:5. 83$ :02 > :6

 

 

 

 

 

 

 

00:00:85 90.0.0.6 000 0000.00.35 0,—0.5 . n 030,—.

 

:13 Che: E...) [Eula Er- ... abavvvh u. 1:...— _ CV3. Inc;

. h c h \
nee—AHEQW u» 2%- ~ 4:. -.u>9w~ ~ :2: NQQLE
.95.... BuOZ NGL< an: _ L. UbiQ~CQ>OLA~ accumsh< ~

Anv.u:°.0v nu Mann-urn.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.5982 «o : BEN .80
mesa ”was 53 era as at _, mans 532 23 Emz < 0325 wawox R a
a 25w
EE 2588 2:, a: * mans 980 EB oaz < 0523 230% fi. 8
558 Mm
C385: 0:338 God x5 3?: 5:3:
3 835828 2953 BS 8: 38 .. £me a 25w .80 2.3 232 < 0525 93%: on on
938:: oswogao 89a is «.85 E 25R
3 BEESSO 03:33 B: 8: an * mama. d=80 83 932 < 8523 «2.8% on a
6:8 baton
9% $8 9 a: 03853 = BEN .80
EB :08on a“; 38 _. mans 522 23 2.2 < 8525 0.3.83 8 M:
at» $8 8 a: 2953 a BEN .80
PE 22:85 was 5: _., ”an? 532 23 22 < 8359 233m 3 2
ca 5
2.08 bosom 8E
22 “232:8 23 38 ”Sam 8 8. .N BE SE 932 ace 8:282 one: on 2
in Saginaw S m
55 22 be, pa; 8.: _. 3an m 5,2 83 82 m 8:282 was: 3 2
= 25a .80 t. we;
22 “23088 S3 E: wmwmmm A5 83 ommz < @2382 0.2.83 mm E
E» Baggage“ =3 9 1 ”ea
53 22 a? 2:; an .. 3%? m E2 83 cmmz m 32:82 339% S 2
2.955 a “3 t .23 E5 82 2:... n .26 * t
E 882 82 a: $8253.... .352 23$ am: > :5

 

 

 

 

25.33 m 035—.

 

 

 

 

. amg any-“Pu tctcu Eda! ' a(:u(\ h h
fiv-QEQW xx acxn fl LL. T8904 ~ 3:: Trent; .9th
.11.. R367— IOL< s:- _ ... Outomflebahan goat-Lax ~ 9.5?»

 

 

 

 

 

 

Anv.u-h°9V Mu Q-n-flrhn

 

31

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5:036:00 3:3 ”8mm $02.0 BE 8%
Sac .232 30 8322.. _ Nag ... wNmeo .3 a N 8m 2 2 use mozb 23.8% >20 mm
>20 mo ziufitm 35:8 038$
”:83 0382 80:3 59:8 0: 9E8
“882 5? 22a 28 29222 an ... NanNo 2. E2 83 onmz m 0228.2. 0.8.83 Nn Nm
Snag 3,62%me a: m
55 22a 28 23222 an .. wanNo N $5 83 82 m 0225 N.om.§m N a
oN 2.5% No ENE Spa N 2:32 N: m 53$
:23 c5 cocaabwac £82 an .. NNN mNo N sag ON? 82 m 222,5. 3.83“ : on
NN 2:32. N: m
2.3 2: 2 EN Egai at ... NNmeo N sag ON? 82 m 2225 Nina : NN
89:8 35.2-8388
.23 £3 255 .568. a: . NanNo N E2 an? 82 m 82:5 $232 E ”N
Homes Bummbwafi 9
-8508 .23 22a 28 22:22 a: . wNNnNe 2.8 we so on? 9.2 m 0225; 328m No NN
“0&an @8669“. 9
-asaoa 55 22a 28 2932 an * $38 2.8 8.2 2o 83 82 m 02E“; 2.2 26: m 8
59:2 mo v8 20 525
Eva 092 223 22a 28 2.8 * wNmnNe 2.8 80 83 cmmz m 8E5 $2.83 3 N
282 Mo = BEN .80
883 032 223 22a 28 e: .. wNNmNo no. 83 82 < 8525 22.8% 3 «N
“£82 .3 a BEN .80
“82.2 092 223 223 28 an _. NNNnNe .29 83 82 < 022,2... 22.83. 3 «N
3.25% a .3 t ES 2:: 82 25. n .25 t a
E 882 8.2 .E + 8:2..an .852 22$ am: > :5

 

 

 

 

 

 

 

 

8.253 n 035,—.

 

 

 

 

 

. —.E_:E=m % .9—
:L @332 aunt. z:

 

Ape-25v mu 01.6,!

32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

59:8 2:8 6:3 2 8:30: cam ommm
3:22:80 352 :E 2 2a .5 a $0.: 0:: 2: <2 25 .0:
35:02: so: 320 025.. N :82 * :Nme: % E2 ”85:38 0:80:02: 8.82.:an :20 Ne
:0:—:8 080m 55 2 88:0: cam cmmw
3:033:00 €53. b3 n 2% .E a 0.0.0.0 A85 .0: $2 25 .E
“9:032: so: 320 028:: 2 :82 . NNme: m E2 0050350 0:50:02: 8.32an :20 G
3:055:00
:23 be, 6:: :ozb 88: :N 8: 2:: 25
=8: 3 2.8.6:: :28: 2:88. N :82 .. :Nme: N E2 8:: 82 :0:. :20 39% :20 8
3:033:00
3:3 b? ”25 3.02.0 89a wN :0: 3E 88
:82 B 8226 :28: 2:55. 2 :82 .. :NwImNo N E2 8:: 82 :2: 32.0 $.%% :20 :m
5:033:00 3:3 SE 2mm
”2:: :20 so: .295: :20: N :82 * :NNnN: E 25: .80 $3 8:: :8: $20 28.8% :20 mm
3:85:00 3:2 3:: can
62: :20 :5: .295: :20. 2 :52 .. ”SIR: E 25: .80 $3 83: so: :20 336% :20 R
2a :20 so: 2:: 2a
320 58:2. :55: 2.2 82m: .23.. .. NNwImN: .3 .: N 8: 22 :0:. :35 23.8% 2% 8
3:25:00 3:2 ”03m sozb 8E 2mm
82: ..2::§ :2: 05:29. m :82 ... :Nme: .2 .: N 8|: :5: :2: $20 23.8% :20 mm
3:033:00 has: ”cam mozb “E: 85
88: ..2::§ :20 83:22. N :82 .. WNNIWN: .2 .: N 8m 22 so: :20 23.8% :20 %
Basso, n .3 t. ES 2:: 8.2 0:? w .30 u t
E 882 :02 z: +022an 32:2 25$ 0:2 > _:m

 

20.2.03 m 030,—.

 

 

A ELF-Gov n 395-.

 

33

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C0588 :02: 5055 80.8 $005
88: 82:88 8.2 be, 28» 2.8 .. 888 22 258 .80 83 2282 < 8828: 28.88 8 8
$838: 09: 00: 09.: .0: : 0:0N .000 LI!
:8: b? 8223 302 2. wmwnmc 20302 2.3 052 < 583:2 n28mfi0¢m cm 3
0:00 bosom
209:8 “a 0:008 n 0:0N .000
22022:: £23 :8: 80> 82> 2:2: _.. wmwlnno :0302 on? 052 < 5:85;. 0.8.263” 02 8.
5:83 0:28.55: 0: $2 32
88 2: 88808 8:8 222m 2: <22 28 2:
.. N 820 2828:. N :82 . 888 2 E2 8822:2182 8882M 8.8.88 50 2.2.
8582 0:08.50: 0: $2 32
202mm .5 wofiowoxm 2:8: 3:5 .5 <2 0.8 .5
.. N >220 828:. 2 :82 .. 888 2 E2 888223 88808 8.8.88 .220 82.
59:2 2:02am: 0: v: 3m
222m 2: 88808 88: 222m .8 <22 28 2:
.. 22.2. 2220 838:. 2 :82 .. 888 83882 888262 88896 28.88 2220 8
£083 :2 :00an 002:: m. 0.: v: 2252
88:8 222m .0: 888:8 82: 8.22.222 38 222.0. .2: <22 28 2:
Lures: w:to&:2 C02 3008.. 109:8: .2. £me m .32 380»...me mofiowoxm mo.mm.nmwn .2025. 3
a. E X 02252
2% .5 38038 5.0.0.0 300 A028 .5 $2 28 .5
80¢ :33 C0. 3020.. 2 02:2: .. «3% m 2o>02 wofiowoximv moaomoxlm 86038“ 320 9.
22:88 22 23 t ES 28: 8:2 25. 8 .80 22 2»
E 822 82 .2: 2.882822: 8225 22$ 0M2 > an

 

222.882 8 2.28:.

 

 

 

  

+ 00:

AULEGUV n 0.3-2'5

 

34

 

 

 

 

 

 

 

 

 

 

 

 

Sm om
0:0 .5 00800020 800 0:0 .5 $2 0:0 .5
.. :20 828:. 2 :82 .. 888 N E2 888220 88822: 8.2.88 :20 8
083 :0 80: 8 Em :08:
”00:83 :0 88 0.0 .82.: E E:
-088 800:. 20:3 08: ::0> 083 8.: 2. 0320 0.32 00: 0:0 00 3 082 m 5:58: 02.00% 3: an
0285
:20 :02: :83 08 0:25 :00: 2. 08030 0:0N 00m 020 00>]: 082 m :0::83. 00308 3 mm
.2: 222> :8 22: N: 3
08: :03 00m :38 08>: :00: ... 08:30 :0 08: 03$ 082 m 58:80. «.2003 m mm
.8» 00 8008 828:: N} m
030:0, :83 00m :38 8:3 8:: 2. £030 :2¢ 88:: on? 009: .5 5:88;. 000003.” hm 0n
:0» .00 8008 3005 BR: 08m
0380: :03 00: :38 08>: :00: ... 08:30 8 8.80: 035 022 m 5:882: 0.3.003” Hm mm
0000.8: :0 SEN; :0:: 8 0:800 3005
00.: \B 00: 3038—8 0085: :00: _. M8030 = 0:0N .000 00? 08:2 < 5:882: 0.0000va 0a 00
8:80. 02:: :83
00: 203820: 0088: :00: 2. 300.00 $5058: 095 032 m :0::82. 0.08003.“ 00 QEmn
.0:00 80:00
000:: :000: 80:: 0.005 a 0:0N .000
22:88 8.2 2:822 2:3 :28: .. 888 8:62 83 :82 < 8228:: 28.8% 8 8
.0:00 80:00
00:80. :0228 30:85 80: 3005 mm 0:22:20":
80.0 008880 08: ::0> 0.83 :00: _.. 0830 E 0:0N .000 03> 032 < 5:88;. 2.00% mm :0
02:88 .2: .3 t ES .80 88 8:: .2: .80 22 8
E 882 85 2: : 88252:: 88.3: 22$ 28:2 > 2::

 

 

 

 

 

 

 

 

 

 

 

3.2.000 m 030,—.

 

 

 

 

 

 

 

I - J34. $3.14!?“ 12.1.0.2: :0: _ <\Z _ 20:0._:E<.:Z ~ £00 .500 ~ 0.000.065: \ :0 \ ~4m \
USQEQW k .0; ~ .7: 3.5.- ~ FED ~09}; 2:5 k .30 a. .2
5:9... 20qu IOL< 0:- H .9 00:0..8030: uUflhm0L< ~ 0.03)) 3W: x, ‘0‘“

 

 

 

A—UouI-Avnvv Mu QIAHRNrH.

35

.200: 00000808800 080805 00:00:85 000 ..00080_:0.8.. A08: 006: 000 00:. 200.00 00800000

.203 00000808800 080803 0008:8000 000 00800000 “00th 00:3 000 0080002

A000: 0388 800,000 800.: .0: 0000020 03300800 000: 3: 302

 

 

 

 

 

 

 

 

 

60.5 80802 880.85 0.0.0 8.03032 8000 A: :0 00.000 8 Sun— I...
00.5 00:3 08. 00.5 8002.2 8803.5 0.80 0000.82 88.: 809 .1
828.02 023.050 222.0 0.3 82.0 .2
0880 .2 8N8
.3 00220.03 08.. 0o 2. N 3 N <2 22.2.28 .02 2% .20 0880 .2 .3 Nam
88% .2 8N8
.3 3.2.23 82. 3 2. N 3 . <2 282.28 82 2% .20 0820 .2 .3 78
0an8 .2 5-2
.3 0820-28 82:5 0o 2. N 3 N <2 22.23% .oz 2% .20 .832 .2 .3 N-Nm
0N8? .2 5-2
.3 02.20-28 0825 .o 2. N 3 . <2 22.2.28 .02 2% .20 88mm .2 .3 .8
80va .2 AN. -.
.3 02.20-23 002. 3 2. N 3 N <2 22.23% .oz 2% .20 0083. .2 .3 N8
08va .2 AN.-.
.3 38.0-23 B... 3 2. N 3 _ <2 282...? .oz 2% .80 883. .2 .3 .8
02250 0 .3 t. .23 _ 2.5 028 2:0. 0 .30 0 0
EL. 00.02 00.3. :5 ._. 0058030...— .00.=t< 0.83 :02 > Em

 

 

 

 

 

 

 

 

3.2.08 0 030,—.

 

 

 

1

sampm

cm

CC}
3V

In

0f

33‘:

36

samples for INAA were then removed from the newly-exposed area. This was done in
order to avoid collecting samples from areas which might have been contaminated by
exposure to the soil in the archaeological context, or by handling during the curation of the
artifact.

In order to further avoid contaminating the samples during collection, powder-free
polyvinyl chloride gloves were worn during the sampling procedure. The gloves were
discarded after each sample was collected. Samples were collected on weighing paper and
transferred to N algene high density polyethylene (HDPE) bottles (4 mL size with screw-top
caps) which had been previously washed with dilute nitric acid, rinsed three times with
distilled water, and allowed to air-dry at room temperature.

The sampling technique utiliwd in this study permitted the collection of samples
which were relatively free of soil and other potential contaminants. It also allowed the
collection of relatively thin sections of the sherd, and affected an area of only 1-2 square
centimeters. However, although it permitted large pieces of temper to be avoided, complete
avoidance of temper was not possible, since in most cases the temper was thoroughly
mixed with the clay. An exception to this was found in Vessel 52, which, as described
above, contained apparently untempered clay on the surface of the vessel. The extent of the
untempered clay was not determined, but it extended to at least three mm below the inner
surface of the vessel.

The presence of temper throughout the vessel matrix was apparent in two ways.
First, the sherds were significantly harder and resistant to grinding than were the clay
samples. Second, the ground pottery samples (except for Sample 32) contained both a
powdery material (clay) and granular material of varying sizes (temper). The larger pieces
of this temper were removed from the pottery samples, but grains less than 1-2 mm in
diameter could not easily be removed. Therefore, except for Sample 32 the pottery samples
are all considered to be mixtures of clay and srnall-grained temper. The effect of this
admixture is discussed under Results (Chapter 6).

37

Two styles of dental bits were used for the sample collection. One was a "cone"
shaped bit (size 023, lot number 555828), and the other a "bud" shaped bit (size 023, lot
numbers 442988 and 625828). The best results were obtained from the "bud" shaped bits,
and consequently this style was used for the majority of the samples. Initially each bit was
used for the collection of two samples, with the bit being cleaned with distilled water and
powder-free tissues between samples. However, this practice was soon abandoned, since
the wear on the bits made the collection of the second samples difficult. Although the bits
were made of a hardened steel, they became heavily abraded during use. Presumably the
grit temper (consisting primarily of crushed granite [Lovis 1973]) had the effect of grinding
ofl’ the sharp edges of the bits during the collection of the sample. Because of this
contamination of the samples with the metal from the dental hits, the bits themselves were
sampled for analysis. Two representative bits from each lot were cleaned, after which the
tips of the bits were broken off with clean pliers. The tips were then weighed and analyzed
in the same manner as the pottery and raw material samples.

During the sampling, differences in the composition and hardness of the sampled
sherds were noted. These are summarized in Table 3, which also provides a listing of the
dental bit lot numbers used for each sample.

Sample

(Q oven.

Vials prep:
Inc., Bufo
mmparab
0f sample
amount 01

‘0 PTBCluc'

METHOD

Sample Preparation

All ofthepotteryandraw material samples were dried overnightin a90°Celcius
(C) oven. Afterdryingthe sampleswereweighedonaMettlerAE 2(X)balanceintoglass
vials prepared from high pmity quartz tubing (Supersil T/21, lot 22942, Herasil Amersil,
Inc., Buford, GA). As much as possible, sample weights and standard weights were kept
comparable (approximately 130-175 mg). In some cases, however, insufficient amounts
of sample material forced the sample weights to be as low as 70-100 mg. In one case, the
amount of sample material was so little (Sample 49, which consisted of a total of 46 mg) as
to preclude it from being run by INAA. A listing of the weight of each of the samples and
standards in each of the batches can be found in Table 4.

Prior to weighing the samples, the one-meter long tubing sections were cut into
sections, and one end of each section was sealed with a flame torch. The resulting vials
were then cleaned by soaking them in a solution of Pierce RBS-pf cleaning agent (Pierce
Chemical Co., Rockford, IL) and rinsing them five times with distilled water. All manual
operations following the soaking of the vials was performed using clean sm'gical gloves.
The tubes were subsequently dried ovemight in a 90° C oven and marked with a sample
number using a glass-scribing tool. After filling the vials with the samples, the open ends
of the vials were sealed using a flame torch and set aside for irradiation. All subsequent
operations involving the samples were performed by Michigan Memorial Phoenix Project /
Ford Nuclear Reactor personnel due to safety considerations in the handling of radioactive
materials.

The pottery and raw materials samples were prepared and analymd in three batches,
each of which was run with a blank (empty) vial, and three vials containing standards
whose elemental concentrations were subsequently averaged and used as the batch

standard. Additionally, a vial containing a "check standard" was run. This "check
38

 

 

 

 

 

 

 

 

 

 

 

 

. A 7.1 22:22 2222:2222
2222222222222 22..

 

39

Table 4 - INAA Batch Information

Vessel Ware Type * INAA
Number

INAA

   

Sample
Number

      

 

 

 

 
  

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l
2 NBS Standard 1633a -
3 Sample 4 Skegemog 164.0
4 Sample 6 SkegemL 164.5
5 Sample 8 Mac / Skegi‘“ 111.4
6 Sample 15 Skegemog 70.0
7 Sample 28 Skegemog 142.4
8 Sample 32 Skefl Prob *** 150.1
9 Sample 51 Skegemog 1 13.3
10 Sample 8 N1 57 Skegemog 151.2
11 Sample 8 N2 57 Skegefl 132.9
12 Sample 9 57 Skegemog 133.5
13 Sample 10 57 Skegemog 153.0
14 Sample 1 1 16 Mackinac 164.6
15 NBS Standard 1633a NIA NIA N/A 156.6
16 Ck Std (NBS 1633a) NIA N/A NIA 157.3
17 Sample 12 21 Mackinac 158.9
18 Sample 13 17 Mac / Skeg ** 96.9
19 Sample 14 53 Mac / Skeg ** 151.6
20 Sample 15 69 Mackinac 93.5
21 Sample 16 26 Mackinac 163.8
22 Sample 17 44 Traverse 158.6
23 Sample 18 36 Traverse 161.8
NIA NIA 19 50 Traverse not run
NIA N/A 20 50 Traverse not run
24 Sample 21 54 Traverse 155.4
25 Sample 22 77 Traverse 157.6
26 Sample 23 85 Traverse 150.8
27 Sample 24 85 Traverse 101.1
28 Sample 25 85 Traverse 154.2
29 Sample 26 3 Traverse 160.4
30 NBS Standard 1633a N/A N/A N/A 151.8

 

 

 

 

 

 

 

22L F2L

 

 

 

INAA
Number I

Unit

 

 

 

 

222222222222

 

 

Table 4 (cont'd)

 

 

 

 

   

 

Sample Vessel Ware Type * INAA
Number Number Spl Wt
"n... . . ......................................... (mg) ‘

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 N/A 0.0
2 NBS Standard 1633a NIA NIA NIA 156.0
3 Sample 27 62 Traverse 172.0
4 Sample 28 79 Traverse 145.0
5 Sample 29 1 1 Traverse 149.7
6 Sample 30 11 Traverse 130.2
(temper)
7 Sample 31 2 Traverse 153.8
8 Sample 32 52 Traverse 151.7
(clay)
9 Sample 33 Clay O'Neil Site 163.0
10 Sample 34 Clay O'Neil Site 167.0
11 Sample 35 Clay O'Neil Site 154.6
12 Sample 36 . Sand O'Neil Site 173.3
13 Sample 37 Clay O'Neil Site 162.6
14 Sample 38 Clay O'Neil Site 163.7
15 NBS Standard 1633a NIA NIA NIA 153.0
16 Ck Std (NBS 1633a) NIA NIA NIA 164.0
17 Sample 39 Clay O'Neil Site 155.5
18 Sample 40 Clay O'Neil Site 158.1
19 Sample 41 Clay Skegemog Site 168.0
20 Sample 42 Clay Skegemog Site 154.4
21 Sample 43 Clay Skegemog Site 167.1
22 Sample 44 Temper Skegemog Site 154.5
23 Sample 45 Clay Skegemog Site 152.3
24 Sample 46 Clay Skegemog Site 156.1
25 Sample 47 Clay Skegemog Site 166.1
26 Sample 48 10 Juntunen 164.5
NIA N/A 49 30 Juntunen not run
27 Sample 50 87 Juntunen 1 13.2
28 NBS Standard 1633a NIA N/A N/A 166.2

 

 

 

0.1.
1

 
    

11234567009

     

01.123456
1111111

 

123456789

     

‘0

41
Table 4 (cont'd)

 

  

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

____ males if“; if: ’ffzis._?_.i"'i':':‘j'f q'iii‘ifgliér
1 Blank
2 NBS Standard 1633a
3 Sample
4 Sample
5 Sample 53 N1 90 Juntunen 152.8
6 Sample 53 N2 90 Juntunen 164.3
7 Sample 54 90 J untunen 151.9
8 Sample 55 31 Juntunen 167.8
9 NBS Standard 1633a NIA NIA N/A 162.1
10 Ck Std (NBS 1633a) NIA N/A N/A 156.8
1 1 Sample 56 37 Juntunen 156.2
12 Sample 57 5 Juntunen 152.0
13 Sample 58 81 Juntunen 160.2
14 Sample 59 84 Juntunen 158.6
15 Sample 60 Clay Skegemog Site 158.3
16 NBS Standard 1633a NIA N/A NIA 159.8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 N/A 0.0
2 NBS Standard 1633a NIA N/A 224.4
3 Sample Bl-l (for spls 1-12) 50.6
4 Sample B1-2 (for spls 1-12) 48.6
5 Sample BZ-l (for spls 13-17) 20.2
6 NBS Standard 1633a NIA N/A NIA 258.2
7 Ck Std (NBS 1633a) N/A N/A NIA 261.9
8 Sample B2-2 bit (for spls 13-17) 23.2
9 Sample B3-l bit (for spls 18-60) 48.4
10 Sample B3-2 bit (for spls 18.60) 50.8
11 NBS Standard 1633a NIA NIA NIA 252.6

 

‘ New bit used (otherwise cleaned bit from previous sample used)

" ' Mackinac per Lovis (1973); Skegemog per Hambacker (pers. comm. 1991)

' ‘ ' Skegemog pottery type per Lovis (1973); problematic
per Hambacher (personal communication 1991)

 

 

between
standard
1633a, 'l
oven for
contain:

00ncent

PRpara
in placr
the use
Vial, 11‘
Mar
Race

of the
Samp
of flu
clay

POM

h ”h!

Q

The

De

3'9";

42

standard" consisted of a stande run as an unknown sample to check for within- and
between-batch reproducibility. The standard used for all the batch standards and check
standards was the National Btueau of Standards's (NBS) Standard Reference Material
1633a, Trace Elements in Coal Fly Ash, lot 682306, which previously had been dried in an
oven for one week by Phoenix Project personnel. A lisdng of the samples and standards
contained in each batch can be found in Table 4, while Table 5 lists the known
concentration of each element in the standard.

Thebatchofdental bittips(Batch4)waspreparedinamannersimilartothe
preparation of the first three batches, with the exception that plastic sample vials were used
in place of glass vials since the lower irradiation time for this batch (see below) allowed for
the use of plastic vials. As with the first three batches, this batch was run with one blank
vial, three standards vials, and one check standard vial. The vials for this batch were
prepared by personnel from the Michigan Memorial Phoenix Project / Ford Nuclear
Reactor, who also weighed the samples and standards.

During the weighing of the samples differences in the consistencies and densities 3
of the samples were noted. The clay samples from the O'Neil Site as well as two clay
samples from the Skegemog Point Site (Samples 41 and 42) were very sandy. The density
of this clay was similar to that of most of the pottery samples. However, the remaining
clay samples from the Skegemog Point Site (Samples 43, 45, 46 and 47) were very
powdery and were approximately half the density of the previous clay samples.

For the most part the Skegemog Ware samples (including Samples 1-10) were very
light and finely divided, and seemed slightly less dense than the following pottery samples.
The Mackinac Ware samples (except for Samples 12 and 13, but including samples 11, 14,
and 15) were more granular (possibly due to more grit temper in the samples) and less
powdery than the Skegemog Ware samples. Samples 12 and 13 (Mackinac Ware)
appeared more like the Skegemog Ware samples (Samples 1-10) than like the other
Mackinac Ware samples.

)Gv‘l‘flh h: nIN-fliltN" . Ill]! 4 I

amnc- tea—052m. V572 5 3:05.030 .3 5.2.2.200209 t m. .0355

43

cm H Omm

A: H 8.0.
o H o.

H “an

06 H Eva
06 H Wu
Nd H QN

on H 600

cd H 0.3
m H 00

N H 2mm

0 H R—

86 H 36

08% H ooofo

3.0 H 26
md H 02.2.
0 H on
ad H hm
m H 00

o H 03

he H m6.
5 H ms.—
06 H 06

EsmzoohN _..
ofiN

8S. ...
80:35. 0
83.8.5.
HEP—oh. a.
833:5. 0.
808.000
8080.00
Bum—800$ ...
8.5033—
.0322
5802
:9:
80:05 a.
8350m— 0

835—0006 0

8.38.0 ..
880 ..

8.3.25
83000 a.

83.00 0.
0008280.

 

 

A803

00300800000

8.8.0

 

0.80 .c 9.8.5 m 80.. 8.080 288.0

2.0:: 2
.. . .08 808.33. 0.2 e. 8:82 a 82....
880800 0.0 A... 00 .3 000008. 0030> 000800-002

.an> 00.00 02030.2. 00 300200000
<03. mmz .«0 2 030p. 80.a 20 0020> 009.000

 

 

N... a 3 83.2.3 ..
.0 a N... 8.85
0.. a 2.. 8880 _.
o. 0 2. 8.30.82 s
m a 00 8823.32 ..
0.... u N... 82.2.... s
0 a a. 60858.. _.
0.... a 8.. 8.3.80
n... a «N 8.880 ...
8. a 0N... .880 ..
0003
808.8080 .88....—

 

2800 .c 825 . 80< 8.080 388.0

88. 888.0 002 2 28.8.0 3 02.8.8080 . m 030...

44

In contrast, the Traverse Ware samples tended to be very granular and contained a
great deal of grit temper. An exception to this was Sample 32 which was believed to have
consisted almost entirely of clay. This sample was very powdery and was half as dense as
the pottery samples; essentially it behaved like the second group of Skegemog Point Site
clay samples described above. The Juntunen Ware samples were grainy and very finely
divided, but exhibited some variation in the consistency of the samples. Sample 50 had the
consistency of the clay samples, but had the density of the other pottery samples. Likewise
Samples 56-57 had a consistency resembling that of the lighter clay samples and, like these
clay samples, had densities one half to two thirds that of the other pottery samples.

Finally, Sample 58 was somewhat gritty and less finely divided, resembling the Traverse
Ware samples more than the other Juntunen samples.

Analytical Procedure

Theconeentrationsoftheelementsin thepotterymatrixandtherawmaterials were
determined using Instrumental Neutron Activation Analysis (INAA) which permits the fast
and simultaneous detection of a variety of elements ranging in amounts from the percent
(%) to the part-per-million (ppm) levels.

With INAA the sample to be analynd is plawd in a radioactive flux where it is
bombarded with neutrons. While situated in this flux the elements in the sample absorb
one or more neutrons, transforming these elements into radioactive isotopes of the original
elements . These isotopes then undergo radioactive decay, emitting energy in the form of
gamma rays and other forms of radiation (Nuclear Reactor Laboratory 1988). Each
radioactive isotope has a characteristic half-life and emission spectrum (Erdtman 1976,
1979; Brown, 1986) which is used to identify the element originally present in the sample.
Quantitative information is obtained by a direct comparison of the intensity of the emission
spectrum of the standards with those of the sample, since the quantity of a given element is

direct]:
1988).

thetw

Univ:

3 $3:
with
the s

detc‘
COD
N u

sut

We
inf
fiv
fiv

45

directly proportional to the intensity of a given spectral line (Nuclear Reactor Laboratory
1988).

The pottery, clay and temper samples (Batches 1-3) were irradiated for six hours in
the two mega-Watt Ford Nuclear Reactor of the Michigan Memorial Phoenix Project,
University of Michigan, Ann Arbor, Michigan. The samples were irradiated in-core in a
nominal flux of 1 x 1013 neutrons/cmzlsecond (Nuclear Reactor Laboratory 1988) within
a sample holder which spins on its axis in order to ensure equal irradiation of the samples
within each batch (Edward Birdsall, personal communication 1991). Following irradiation
the samples were allowed to decay for approximately one week, and the gamma radiation
emitted from the samples was counted on a Lithium-drifted Germanium (GeLi) gamma ray
detector equipped with an automatic sample changer. Each sample and standard was
counted for 40(1) seconds (live time), and the results were electronically transferred to a
Nuclear Data 6700 computer system, which automatically background-corrects the data by
subtracting the values obtained for the "blank" from the sample values. The samples and
standards were allowed to decay for another four weeks (for a total of five weeks) and
were counted in the same manner as the week-one counts. The week-one counts provide
information on elements with half-lives of between one and twelve days, while the week-
five counts provide information on elements with half-lives of between fifteen days and
five years (Nuclear Reactor Laboratory n.d.)

Due to the high metal content of the dental bits, these samples were irradiated
separately for only one hour in batch number four. This change was necessary since
samples with high metal concentrations activate very easily, and when they are irradiated
for long periods of time produce is0topes with activity levels beyond the safe limits of this
reactor (Edward Birdsall, personal communication 1991). The subsequent decay and
counting of batch 4 was identical to the parameters utilized in batch numbers 1-3.

Following irradiation and counting, data from the "peaks" (spectral emission lines

as seen by the gamma counter) were collected through the computer interface. The energy

of these p
qualiadv
intensity
quarries
ifiycfiu
3ch an
intens'r‘
Ymes s
indica

analy

infor
Slam
of A

encr

45

of these peaks, which is measured in thousands of electron volts, or Kev, provides
qualitative information about the identity of the element producing the emission. The
intensity, (1' activity level of the peaks (measured in counts per second, or cpm) provides
quantitative information about the amount of each element in the sample. The initial listing
of peaks for the week one and week five counts was checked against tables of neutron
activation (Erdtman 1976; Brown 1986) in order to select the peaks with the highest
intensity and the least interference from other peaks for each element in the standard. The
lines selected are listed in Table 6 by element. As a quick glance of the calculated results
indicates(seeAppendixA),someofthelines selectedprovednottobeusefulinthefinal
analysis of the data, since they were not found in all of the standards for each batch.

Subsequent to the selection of the lines to be used in the analysis, the following
information was entered into the computer for each batch: mass (mg) of each sample and
standard, concentration (ppm) of each element in the standard (as reported on the Certificate
of Analysis for the NBS standard), the half-er of each element in the standard, and the
energy emission line(s) attributed to each element in the standard. This information,
together with the peak intensity data (corrected for background) were run through the
Nuclear Data "Gamma Spectroscopy" and "NAA" software packages. Using the afore-
mentioned data, this software calculated the concentration of each of the specified elements
in every sample 4. Subsequently, a report of the results obtained during the week-one and
week-five counts was generated for each batch.

These reports contained calculated concentration values for 31 elements in each of
the samples analyzed, as well as "standard deviation" values (error estimates) associated
with each concentration value. A digital version (on floppy-disk) of these reports is on file
at the Michigan Memorial Phoenix Project / Ford Nuclear Reactor under file number R481,
and a hard copy version is on file at the Michigan State University Museum, Anthropology
Division. The concentration and standard deviation data from these reports are summarized
in Appendix A.

 

47

Table 6 - Energy Lines and Half -Lives used in Calculations

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Line 1 (KeV) Line 2 (KeV) Line 3 (KeV)
Barium 11.7 days 123.7 373.2 496.3
Bromine 35.4 hour 554.3 776.5
Lanthanum 40.23 hour 815.8
Lutetium 6.71 days 208.3
Molybdenum 66.02 hour 140.5 739.7
Neodynium 10.98 days 91.1 531
Samarium 46.5 hour 103.2
Uranium 2.35 days 99.5 106 277.6
Ytterbium 4.19 days 282.6

 

 

 

 

 

 

 

  

Element

-'-‘. . C.
.......................................

  

. ‘.‘.‘.‘.‘ r . t'. . . . '. ‘.’.'.‘¢'.:‘:-;.;. -:.;.’.:. -‘~ . .‘. .

........
........

 

Half-Life

Ener

       

'.‘.'.‘
._.,
' ~ '2-I-I‘I‘Z'f-I-EPN'I'}I'I-Z'Z'i‘?I'I‘I'I'I‘W'I f I 3‘: f‘Z-I‘Z‘C'Z-Ii-i-N

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Enegg
Line 1 (KeV) Line 2 (KeV) Line 3 (KeV)
Antimony 60.2 days 1691.04
Cerium 32.38 days 145.45
Cesium 2.062 years 604.7 795.76
Chromium 27.70 days 320.01
Cobalt 5.27 years 1332.51
Europium 12.7 years 1085.8 1 112 1408.08
Gadolinium 241.6 days 103.2
Hafnium 42.5 days 482.16
Iron 45.1 days 1099.22 1291.6
Mercury 46.59 days 279.17
Nickel 70.78 days 810.75
Rubidium 18.6 days 1076.63
Scandium 83.85 days 889.26
Selenium 120.4 days 136.00
Strontium 64.73 days 513.99
Tantalum 115.0 days 1 189.00 1221.28
Terbium 72.1 days 879.37 1178.00
Thorium 27.4 days 311.9
Thulium 128.6 days 84.26
Tin 115.1 days 391.71
Zinc 243.8 days 115.52
Zirconium 64.4 days 756.72

 

 

 

 

 

 

 

 

than one ener;

Standards (Set
Clem“, hox
Barium N60
each batch w
COHCentmu'QI

Check Stands

RESULTS

Data Evaluation

AsseeninAppendixA,thepereentstandarddeviation associatedwiththe
measm'ement of each element varies widely by element and even varies somewhat between
samples for the same element. As will be shown below, these standard deviation values
helped to determine which elements would be used in the final analysis of the INAA data.

For some of the elements present in the NBS standard it was possible to use more
than one energy line to calculate the concentrations of the elements in the samples and
standards (see for example Ytterbium, Emopium and Terbium in Appendix A). For other
elements, however, multiple lines gave different concentration values (see for example
Barium, Neodymium and Cesium in Appendix A). In such cases, the check standard for
each batch was examined to determine which line(s) provided the calculawd value of
concentration corresponding most closely to the known concentration of that element in the
check standard. Using this method the following multiple-line elements were eliminawd
from further analysis: I

Barium ( 373.2 KeV line)
Uranium (99.5 KeV line)
Gadolinium (103.3 KeV line)

An additional review of the concentration of elements in the check standards
revealed that the values for certain elements did not closely approximate the known
concentrations of these elements in the standards. In addition, some of the elements could
not be found in the batch standards by INAA, and therefore no accurate concentration
values could be calculawd for the check standards. Finally, in a few cases the
concentrations of elements in the check standards could only be reported as "less than"

somevaluesincethebaselinelevelinthisareaofthespectrumwastoohightoyieldamore

48

accmattme

further data

 

 

Simil
Values in exc

fouowmg ad

49

accurate measurement. For these reasons, the following 10 elements were eliminated from
fin'ther data analysis:
Bromine ( 554.4 and 776.6 KeV lines)
Molybdenum (793.7 KeV line)
Mercury (297.2 Kev line)
Nickel (810.8 Kev line)
Selenium (136.5 Kev line)
Strontium, 514.0 Kev line)
Terbium (879.4 and 1178.0 KeV lines)
Thulium (84.3 Kev line)
Tin (391.7 KeV line)
Zirconium (756.7 KeV line).

Similarly, elements whose concentration values had percent standard deviation
values in excess of 15% were not used in the analysis. This latter procedure eliminated the
following additional elements from consideration :

Molybdenum (140.5 KeV line)

Neodymium (91.2 and 531.2 KeV lines)

Uranium (106.1 and 277.7 Kev lines)

Ytterbium (282.5 KeV line)

Antimony (1690.5 Kev line)

Europium (1085.6 and 1112.2 KeV lines)

Tantalum (1189.1 and 1221.5 KeV line)
In addition, the 604.4 and 795.5 KeV lines for Cesium had relatively low standard
deviation values for most of the samples except for the O'Neil clay samples, which had
standard deviation values on the order of 10-30%. However the Cesium concentration
results for these clay samples were between one and two orders of magnitude smaller than

the Cesium concentration values for the remaining samples. Therefore, even with a high

contan
ftpres,
PrCSur
his 211

Such l

COne

50

percent standard deviation, the Cesium concentration values in the O'Neil clay samples was
significantly difi'erent from that in the other samples. For this reason, this element was
included in the final analysis of the INAA data. The 795.9 KeV line was chosen over the
604.7 KeV line because the concentration of Cesium in the check standards calculawd with
the former energy line was closer to the known value than was the concentration calculated
with the latter energy line.

Finally, several elements were not used because they were thought to be possible
contaminants from the drill bits used in collecting the samples. The elements which were
represented in significantly greater proportions in the dental bits than in the samples were
presumed to have been added to the in the pottery samples through contamination by the
bits and were eliminated from further consideration. The following elements represent
such potential contaminants:

Chromium (320.2 Kev line)
Cobalt (1332.5 KeV line)
Iron (1099.3 and 1291.6 KeV lines)

Since it was known that the drill hits were comprised in large part of Iron, the
concentration of this element in the bits and in the samples provided a useful measure of the
degree of contamination of the samples by the drill bits. As shown in Appendix A, Iron
comprises 82.5% to 89.3 % (825,000 ppm and 893,000 ppm, respectively) of the matrix
of the drill bits. In the pottery samples -- samples which were most contaminated by the
drill bits due to the hardness of the sherds -- Iron is found in levels ranging from 2.58% to
7.10% (25,800 ppm to 71,000 ppm, respectively). Hence, taking into account the purity
of the iron in the bits (as low as 82.5% pure iron) the greatest amount of contamination of
the samples by the bits is 8.6% 5. Even with this low level of potential contamination,
however, two other elements were eliminated from further data analysis:

Rubidium (1076.8 KeV line)

Althc
bits, 1
tone

pom:

rhcs

thcc

COM

38C

Ce]

Val

of

51

Zinc (115.6 Kev line).
Although the exact concentrations of these elements could not be determined in the drill
bits, their maximum concentration levels were as much as two to three times the
concentrations of these elements in the pottery samples, thereby raising the question of
potential contamination of the samples by the drill bits.

In one instance, the concentrations of elements in the drill bits relative to those in
the samples were used to select one calculated concentration value over another. In this
manner, the calculated concentration of Barium using the 123.7 KeV line was selected over
the concentration calculated using the 496.3 KeV line, since in the latter the maximum
concentration of Barium in the drill bits exceeded the concentration of barium in the pottery
samples, whereas with the 123.7 KeV line the concentrations of this element in the pottery
samples and in the drill bits were approximately equal.

The elimination of the foregoing elements from subsequent analysis yielded a list of
eleven elements which could be used to further examine the pottery and clay samples. For
each of these eleven elements - Barium, Lanthanum, Lutetium, Samarium, Ytterbium,
Cerium, Cesium, Europium, Hafnium, Scandium, and Thorium --- the concentrations in
each sample, along with the associated percent standard deviation of each concentration
value, are listed in Appendix 2. These eleven elements represent the elements whose
calculated concentrations in the check standards closely matched the known concentrations
of these standards as reported on the Certificate of Analysis for the NBS Standard. They
also represent only those elements whose concentrations were reported with relatively low
percent standard deviation values. Finally, the patential contamination of these elements
fiom the bits is considered to be negligible, since the concentration of these elements in the
bits was, for the most part, lower than the concentrations in the pottery samples. For these
reasons, the reported concentrations of these eleven elements in the samples are believed to

accruately represent the concentration of these elements in the pottery and clay samples.

52

One exception to this are the reported concentrations for Sample 46, a Skegemog
Point Site Clay sample which was collected from the same artifact box as Sample 47. As is
shown in Appendix A, many of the percent standard deviation values for the elements in
this sample exceeded 15% and several exceed 25%. When only the above-liswd eleven
elements are considered, the percent standard deviations for this sample are higher than for
any other single sample. It is presumed, therefore, that this sample represents a measuring
outlier, and that the results reported are not accurate representations of the concentrations of
these elements in the sample. No other such outliers were noted in the samples analymd.

Data Analysis

Following the selection of elements, the concentration of elements in each sample
was analyzed. Due to the multivariate nature ofthe data, the logarithms ofthe
concentrations were calculated and used in the analysis, since the log of concentration
standardizes the data and "corrects” for differences in magnitudes between elements whose
concentrations range from the percent level to the part-per-million level (Sayre 1977;
Bishop and Neff 1989).

A cluster diagram using centroid linkage of Euclidian distances was generated for
all of the pottery, clay, and temper samples using the Systat 5.0 Statistical Package
(Wilkinson 1989). As shown in Figure 4, this cluster diagram separated the O'Neil Site
clays and the Skegemog Point Site clays from each other and from the majority of the
pottery samples, but did not produce the expected clusters of "local" clay and pottery
versus "non-local" clay and pottery.

In order to determine whether only one or a few of the eleven elements used in the
analysis were responsible for producing the cluster outcome, the samples were sorted by
increasing order of concentration for each of the eleven elements. The resulting

compilation of ranked concentration values, listed by element in Appendix C, indicates that

 

53

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W.W-WmnanHHHHHHHHHHHHHHHHHHHHHHHHW W W-W-W-W . ....... . a. m
. .-mmmwnmnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnunnnnunnnnm ....... W W-an W m n .
. . W W-W W-W W- ”-unuunnnunnnuuunnun...”n"...“n.--- . W WNW .--um
. . . W W W -HW--HW--nnnunnuW----H--W W . W
W . W . W-W--HW . -W W . w- ---W
_.._ W. .. .....-.W..W_W_

' .
............n...._..._........a............a.....

mmumuuaau

 

Figure 4 - Centroid Linkage Dendrogram Using 11 Elements

athc
tide
ckwer

mtdu

codd
I989)
OHkr
numb
Opfior
50th
0ftm
ttoa

dive

joint”
Celtic]
@DUp

must:

54

each of the eleven elements alone more or less reproduces the clusta' result obtained with
all eleven elements. Similarly, a series of clusters diagrams produced using only ten of the
eleven elements -- sequentially deleting each of the eleven elements in turn -- also
produced results similar to the cluster diagram formed with all eleven elements (see
Appendix D). Hence, since there was no reason to believe that any one element produwd a
better clay or pottery "fingerprint" than any other element, or that any one element was
adding unnecessary "noise" to the cluster diagram, the original eleven-element cluster
diagram was used for subsequent interpretations.

To determine the optimal number of groups into which the multivariate samples
could be subdivided, a partitioning of the sample via K-means was performed (Wilkinson
1989). This method determines the best way in which to divide the collection of samples in
order to maximally separate the groups (Wilkinson 1989: 25). Without specifying the
number of groups into which the samples were to be divided, the K-means clustering
option yielded only two groups, one containing all of the O'Neil Site clay samples and
some of the Skegemog Point Site clay samples, and another group containing the remainder
of the samples . By specifying the number of desired groups (i.e., by "forcing" the results
into a given number of groups) the samples were subsequently divided into from three to
eleven groups, respectively (see Appendix E).

A further examination the eleven-element cluster diagram revealed that based on the
joining distances -- a measme of the relative distances from the center of one cluster to the
center of the next nearest cluster -- the cluster diagram could be subdivided into six major
groups whose components are listed below. A schematic drawing of the six groups

illustrating their relative joining distances is shown in Figure 5.

Group 1: All pottery samples (Samples 1-29, 31-32, 48-59)
+ Skegemog Point Site clay samples (Samples 41 and 42)

GROI

Ska

GROI
33C
Tta‘

GROI

0.0

0.

1

 

GROUP 1 - (0.0 - 0.124)

All pottery samples
including S32 (”pure"
clay, Traverse Ware)

+

S41 and 842 (Clay,
Skegemog Pt. Site)

 

55

I Relative Joinling Distlances I

032

0.3

 

 

GROUP 5 - (0.149)

S44 (Temper,
Skegemog Point Site)

GROUP 6 - (0.189)

 

 

830 (Temper.
Traverse Ware)

 

GROUP 3 - (0.106 - 0.113)

S43, S45, 847, 360
(Clay. Skegemog Pt. Site)

 

 

 

GROUP 2 - (0.102 - 0.110)

S33, 834, 835, 836,
S37. S38. 839. 840
(Clay, O'Neil Site)

0.4

0.5 0.6 0.7

 

 

Joining D'utances are
shown in parentheses ()
following group number

 

 

 

 

 

 

 

 

 

GROUP 4 - (0.361)
846 (Clay,

 

 

 

 

Skegemog Pt. Site)

Figure 5 - Schematic Dendrogram of Groups 1-6

0.8

The con
the six-1
groups]

in“) me.

SUbSeq.
Chalet ;
Lite W

by Sher,

Group3:

Group4:

Group 5:

Group 6:

56

: O'Neil Site Clay 8: Sand Samples (Samples 33-40)

Skegemog Point Site Clay Samples (Samples 43, 45, 47, 60)

Skegemog Point Site Clay Sample (Sample 46) - Probable
Outlier

Skegemog Point Site Temper Sample (Sample 44)

Temper Sample from Traverse vessel (S30)

The components of these six groups also correspond exactly to the group membership in

the six-group K-means cluster, suggesting that the division of the samples into six major

groups provides the best possible separation, while simultaneously dividing the samples

into meaningful groups.

Because Group 1 contains the majority of the samples analyzed, a schematic

diagram of this group, showing joining distances within it, was produced (see Figure 6).

Subsequent examination of the Group 1 cluster revealed that the pottery samples did not

cluster by pottery tradition (Northern or Southern tradition), by time period (early or late
Late Woodland time periods), by site provenience (Area A or B, or levels 1 through 4), or

by sherd type (rim or body sherd). In fact, with the exception of one small cluster of five

Juntunen samples representing three Juntunen Vessels (Samples 53Nl, 53N2, and 54, all

representing Vessel 90; Sample 56 representing Vessel 37; and Sample 57 representing

Vessel 5), few uruelawd samples of the same pottery type joined directly to another sample

of the same type. This lack of pairing, however, is not due to an inability of the Neutron

Activation Analysis to detect similar concentrations of elements in similar sherds. Nor is it

due to an inability of the clustering program to match similar samples. Instead, the lack of

W“.
_..n

2m. Wm WWW—Wm. m

0
5
8

WWW

0.8.0.

852

328

57

Relative Joining Distances
I I I I I I I . . I I I
0.00 ' 0.02 ' 0'04 ' obs ' 0.110 ' 0.10 0.12 0.14

"——353N1, 553m. s—54' . "sso. I
ss7 (0.027 - 0.037)

"_'so. sro—, 321, 351 In
(0.035 - 0.038)

829 (0.040) ’
S9, 812 (0.030) fl Joining Distances are
S8N 1, 88N2 (0.038) shown in parentheses ()

816 (0.042) following sample numbers
S50 (0.045)
S7 (0.046) 7

Si, 817, 818, 823, 824,
831. S48. 851. 855
(0.026 - 0.037)

859 (0.047)
82 (0.053) 1
827 (0.057)
815 (0.057)
822 (0.064)
813 (0.069)
S32 (0.073) 1
SS (0.072)

83', s' l'4',"s'25—, 826, S41, S42, sss * , .—
(0.042 -0.066) I'—

34 (0.099)
852 (0.113)
811 (0.116) to Group 5
828 (0.124)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6 - Schematic Dendrogram of Group 1

varia

com;
53, ll
duph'
B as

other

pair;

bting

SW]
and t
35 (c
d0 nt
Sam;
“Tire

elem

(Mid

58

pairingofsherdsofthesamepotterytypesseemstobeduetothelargewithin-type
variabilityinthe sherdmatrix, avariability which isgreaterthan thatbetweenpotterytypes.

The ability ofthe INAA and clustering methods to pair samples ofsimilar
composition is borne out by the results of the analysis of duplicate samples. Samples 8 and
53, representing a Skegemog and a Juntunen vessel, respectively, were analyzed in
duplicate by INAA, and the results of these analyses are summarized in Appendixes A and
B as Samples 8N1, 8N2, 53N1, and 53N2. When these results were clustered with the
other samples in the eleven-element cluster, both pairs of samples joined with its
corresponding duplicate sample (see Figlue 4). Furthermore, the duplicate samples in each
pair join at very low joining distances, indicating that these duplicate samples are seen as
being very similar in composition to one another, and that these pairs are closer in
composition to each other than to any other samples.

Although the method is capable of pairing duplicate samples of pottery, it is less
successful in pairing samples taken from lumps of clay collected from the same location
and curated together in the same storage bags or boxes. For example, Samples 33, 34, and
35 (collected from three lumps O'Neil Site clay from the same archaeological provenience)
do not join together directly, but instead are part of a larger cluster (Group 2) including
Samples 36, 37, and 38. In fact, none of the clay samples from the O'Neil Site which
were collected from the same archaeological context joined together directly in the eleven-
element cluster.

On the other hand, Sample 36, described as ”fired sand found beneath clay"
(Michigan State University Museum n.d.) appears to be compositionally very similar to the
O'Neil Site clay samples. This, together with the sandy texture of the O'Neil Site clay
samples suggests either that these clays were nannally rich in sand, or that sand was added
to the clays during the preparation of the clays by prehistoric potters. Given the available
data, preference cannot be given to either of these possibilities.

 

ll
collected
container
of this pa
Samples
joining d
the clay I
6). The 1

A
the other
the densi
some "te

bomadc'

59

Within the Skegemog Point Site clay samples only two pairs of samples were
collected from lumps of clay with the same archaeological provenience. One of these pairs
contained Samples 46 and 47, but since Sample 46 appears to be an outlier, no comparison
of this pair can be accomplished. However, the other pair of Skegemog clay samples -
Samples 41 and 42 -- while they do not pair together, do appear in a small cluster with low
joining distances. However, this small cluster appears within Group 1, and, in addition to
the clay samples, also contains samples fiom every pottery type sampled (see Figures 4 and
6). The remaining Skegemog Point site clay samples are clustered in Group 2.

As noted in the previous chapter, Samples 41 and 42 had a sandier consistency than
the other Skegemog Point Site clay samples. Also, unlike any of the other clay samples,
the densities of these samples were similar to the densities of the pottery samples. Since
some "temper" was noted in the clay lumps during sampling, it suggests that temper had
been added to this clay by the prehistoric potter who prepared the clay. However, it is also
possible that these lumps of clay contained nannal inclusions which mimic temper in
composition and cause these samples to fall in the pottery group.

The joining pattern of the raw material samples indicates that the clay samples fiom
the two sites are clearly distinguishable from each other. It also indicates that, with the
exception of Samples 41 and 42 (Skegemog Point Site clay), the O'Neil Site clay samples
and the Skegemog Site clay samples are more similar to themselves than to any of the
pottery samples. What is not clear, however, is what is responsible for the differences
between the clay samples and the pottery samples. If these differences are due only to the
elements in the clays, one could then conclude that the the clays sampled are not the same
clays used to manufacture the pottery from the O'Neil Site. However, because most of the

Pottery samples analyzed contain temper, the contribution of this material to the elemental
composition of the samples must also be considered.
In order to address this question of added temper, a mathematical test of mixing
Was conducwd according to the method described by Vogel, er al. (1989). Using this

nthoh the

elements a,

 

rigorous ten
dd. which i
denotninntt
do not yielt
TCSponsiblt
Onhttotx
Since the :
samples f.
Skegcmo
Ware ves
VCSStl re
necessar
itscga
elemem

Citroen
l .

Lil/Sm

”006s:
Pitsm
MN
the it:
Subsn

“Ugh

60

method, the mixing of two components can be evaluated by plotting the ratios of any four
elements a, b, c, and (1 found in the samples. According to Vogel, er al. (1989: 17,948) a
rigorous test is performed by plotting two pairs of ratios: the ratio of a/b versus the ratio
old, which should form a hyperbola, and the ratio a/b or c/d versus the ratio of the original
denominators (i.e. b/d or d/b) which should form a straight line. If the plots of these ratios
do not yield the expected hyperbola and straight line, then more than one process must be
responsible for the mixtures in the samples (Thomas Vogel, personal communication). In
order to perform this test, related samples containing both clay and temper were needed.
Since the relationship between the individual vessels was not known with certainty, only
samples from the same vessel were plotted together. Three vessels -- Vessel 57, a
Skegemog Ware pot represented by Samples 8N1, SN 2, 9 and 10; Vessel 90, a Juntunen
Ware vessel represented by Samples 53Nl, 53N2 and 54; and Vessel 85, a Traverse Ware
vessel represented by Samples 23 - 25 -- provided a sufficient number of samples
necessary to determine whether the plots of the ratios produced the expected curves. For
these samples, in order to achieve the greatest possible separation of plotted points, four
elements with high within-vessel variability were selected for analysis. A list of the
elements used as well as the numerical value of the ratios a/b. old and M appears in Table -
7 .

The plots of the ratios a/b versus c/d (i.e. La/Sm vs. Ce/Eu) and Nb versus d/b (i.e.
La/Sm vs. Eu/Sm) for vessels 57, 90 and 85 are shown in Figures 7-9. As shown in these
figures, the plots of these ratios do not provide the necessary hyperbolas and straight lines
necessary to confirm the simple mixing of two components. Therefore it must be
presumed that even within individual vessels more than one mixing process was
responsible for producing the resultant mixtme of clay and temper. One possibility is that
the temper added to the clay in the pottery-making process was not a homogeneous
substance, and therefore added elements in varying amounts to the clay substrate. The
crushed granite temper used in these samples (Lovis 1973), if not thoroughly pulverized

61

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80... ..00 00.0 00.... 00.0 :0... 00.0 2.0 .82... .080. 000
:0... 0.00 0.0. 0. .... 00.0 0.0... 00.0 2.0 82.0 00.0 000
0.0... 0.0 00.0 .0... 00.0 80... ...0 o..0l..oz.o .0... .00
00.... .0... .0. 0...... 0.... 0%.. 00.0 2.0 .82... 00.0 000
.00... ..0.. 000 3.0... 0.8. 00... 000 082.80 00 000
00.... 0.0.0l 00.0 0.... 3.0 000 0.00 082.80 00 000
0.0... 0.00 00.0 0... 0.0 00.0 0.00 08:8... 00 000
.00... .00 3.0 00.. 0.00 ..0.0 0.00 8888. 00 .00
80... .00 00.0 000... ....0 00... 0.00 8888. 00 02000
000... 0.00 00.0 08... 0.00 l8... 0.00 88.8.. 8 .2000
00.... 0.0.0 .00 0.... ....0 00.0 0.00 008008.0 00 ...0
0.0... 0.0 .00 00.. ....0 3.0 0.00 08.0000. 00 00
:0... 0.00 0.0 00.. 0.00 000 0.00 00808.0 00 0200
0.0... 0.00 ....0 00.. 0.00 00.0 0.00 880086 00 .200
W 900. 0...... _ $3. 0.... W 93. 0.... W 93. 0.0.
8030 :0:... 80:... =0 00 E0 0..
8:5— 0:58 22.00.200.50 0:0:—0:0 090,—. 0.55 0.008.» #2..an

 

00.0.5.0 0.3:...5 5......— 00008200 0000200 0.. 02.3— 0.... 00.00.508.30 . b 030,—.

 

62

 

 

7.00
6.80
6.60
6.40
6.20
6.00
$5.80 '
\5.60
3 5.40
5.20
5.00 I 1' : i : 4 =
50.00 55.00 60.00 65.00 70.00 75.00 80.00

Ce/ Eu

 

 

 

 

Figure 7 - Plots of Ratios for Samples 8-10

 

63

 

 

7.00
6.80
6.60
6.40
6.20
6.00
5 5.80
\ 5.60
5 5.40
5.20
5.00 . :

I I I I
T I I

50.00 55.00 60.00 65.00 70.00 75.00 80.00

6.90 --
6.70 ~-
6.50 --
6.30 --
W 6.10 --
\
5 5.90 ""
5.70 --

5.50 4.
0.18 0.19

 

Ce/ Eu

0.20 0.21 0.22 0.23
Eu / Sm

 

Figure 8 - Plots of Ratios for Samples 53-54)

 

 

 

7.00
6.80 .
6.60
6.40
6.20
6.00
w5.80
\5.60
5 5.40
5.20
5.00 . : : é : i Q.
40.00 45.00 50.00 55.00 60.00 65.00 70.00

Ce/ Eu

 

6.90 .-
6.7o .-
6.50 ~-
6.30 -- -

(,5, 6.10 .-

‘ 5.90 -- '
5.70 ~-
5.50 : ; : = .1

0.19 0.2 0.21 0.22 0.23 0.24
Eu / Sm

 

 

 

Figure 9 - Plots of Ratios for Samples 23-25

 

and 1
subs

mica

31H

Shem
from

00111

65

and mixed by the potter prior to addition, could be considered such a non-homogeneous
substance, since granite itself is made up of several minerals including quartz, feldspar,
mica and/or hornblende (Dorr and Eschman 1977: 36). In addition, the clay itself may not
be completely homogeneous. For example, when the test for mixing was performed on
three lumps of clay collected together from the O'Neil Site (Samples 33-36), the plots did
not yield the expected straight line and hyperbola for a simple mixing process (see Figme
10). In spite of the fact that these clay samples contain a great deal of sand, the mixing
process (either natural or as a result of human activity) was apparently not a simple mixing
process of sand and clay. Therefore, it appears that the components used to make pottery
are themselves heterogeneous mixtures of several materials.
Another possibility is that diagenesis within the site is not uniform, thereby causing
sherds found in different areas to have some elements differentially added to or extracted
from the individual sherds, and producing a reconstructed vessel whose component sherds
contain widely different concentrations of one or more elements. However, this appears
not to be the case for two of the three groups of samples taken from a single vessel. For
both the Skegemog Ware Vessel 57 (Samples 8-10) and the Juntunen Ware Vessel 90
(Samples 53-54), the multiple samples taken from each pot cluster closely together with
small joining distances, in spite of the fact that the sherds sampled from each vessels were
found in different archaeological contexts (see Table 8). Such is not the case, however, for
Traverse Ware Vessel 85 (Samples 23-25). Although the two sherds found in the same
unit (Samples 23 and 24) pair together with very small joining distances (see Figures 4 and
6), the third sample from this vessel (Sample 25), which was taken from a sherd collected
from Area B, has more similarity with other pottery and clay samples than with the two
other samples from the same vessel (see Figures 4 and 6).
These results preclude the complete dismissal of the role of diagenesis in the

analytical results obtained. However, two other factors could also account for this lack of

growing of the three samples taken from Vessel 85. The first is the previously-mentioned

 

 

17.00 --
15.00 -- '
13.00 --
E 11.00 “-
tn
\ 9.00 -~
7.00 --

5.00 l 1 i l l
50.00 60.00 70.00 80.00 90.00 100.00
Ce / Eu

II I 536 ("Sand")

 

 

15.50
14.50
13.50 I
12.50
11.50
E 10.50
‘2 9.50
8-50 -s36 ("Sand") -
7.50
6.50
5.50 . : i : : :
0.18 0.19 0.20 0.21 0.22 0.23

ll

 

 

Figure 10 - Plots of Ratios for Samples 33-36

 

67

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.. 250 .80 _ 003 0002 _ < _ .280. 0000 _ 00 _ .0 _ 02.280
0 _ 003 0002 _ m _ 0.80 _ 00 _ 00 _ 08.0.5.
€058 mm 5
.0820 .. 250 .80 003 0002 < 0.80 00 00 .5: .00. 022,80
.. 260 .80 003 0002 < 8.. 00 0. .5... 80. 6228.0
08 0.0 .800 250 .80 SB 0002 m 0000 00 00 608220
.. 250 .80 02 00>» 0002 < a... 00 ..0 022,0...
.. 250 .80 02 00>» 0002 < 8.. 00 00 ".0860...
0 003 0002 m a... 00 0. 00660020
0 003 0002 < 8.. 00 0 00882.0
0 00>» 0002 < 6.. 00 0 00682.0
_ _ _ .0008, :e _ 03:52 mi 03:52 _
.951— _ 0...: _ 09:... _ 0.2.8...— _ $08.» _ 2956 _ 25$

 

0.0000.» 055m 595 50.5. 8.0.55 2.5.32 . a 0.0.5.

 

 

hem

will

aSl

dun'
Gm
(Sal

C011

Vcs

Silt

311C
Par
abc
the

Cor

Cor.

68

heterogeneity within vessels, caused by the inclusion of varying proportions of clay and
temper in the analytical samples, or by the heterogeneity of the raw materials themselves
(i.e. clay and temper). The second possibility is that the classification of the vessels is not
completely accurate, and that more than one vessel is represented by Samples 23-25.

In addition to the groups of pottery and clay noted above, three other groups, each
with a single member, were found by the cluster analysis. Group 4 consists of Sample 46,
a Skegemog Site clay sample considered to be an outlier. Groups 5 and 6 are temper
sarrrples collected during the excavation of the Skegemog Point Site (Sample 44) and
timing the sampling of a Traverse Ware vessel (Sample 30, collected from Vessel 11).
Group 5 (Sample 44) join most closely to Group 1 (pottery samples), while Group 6
(Sample 30) joins most closely to Group 3 (Skegemog Point Site clays), indicating that the
composition of the temper collected at the Skegemog Point Site is most similar in
composition to the pottery samples from the O'Neil site, and the temper collected from
Vessel 11 has elemental concentrations most similar to the clay from the Skegemog Point
Site. The latter result is surprising, since the temper was taken from a mixture of clay and
temper (Sample 29, containing both clay and temper from Vessel 11) which, when
analyzed, fell within Group 1 and showed little similarity to Sample 30. This suggests that
both the clay and temper are contributing to the elements contained in the pottery samples,
and that each component by itself will not necessarily detemrine the group to which a
particular sample is assigned. However, given the results of the mixing test described
above, it appears that not only is the temper itself adding to the elemental composition of
the pottery samples, but individual fractions of the temper and/or clay are probably
contributing elements differentially to the overall mixtme.

Although the elemental concentrations of the majority of clay samples support the
contention that both clay and temper are necessary for determining group membership,
three samples of clay contradict this premise. First, the two previously-mentioned clay
samples from the Skegemog Point Site (S41 and 42) are found within the pottery cluster

(Gt

85;

Wt

”0s;

69

(Group 1), and more specifically within a smaller cluster containing a Juntunen Ware
sample (Sample 58, from Vessel 81), two Traverse Ware samples (Sample 25, from Vessel
85; and Sample 26, from Vessel 3), a Skegemog Ware sample (Sample 3, from Vessel 53)
and a vessel attributed to both Mackinac and Skegemog Wares by different investigators
(Sample 14, from Vessel 53). These clay samples, then, are compositionally similar to
each of these wares. This is not surprising, since these clay specimens were observed
dtn'lng sampling to have contained pieces of temper (see Table 3).

Likewise, the "pure clay” sample collected from Vessel 52 (Sample 32), when
evaluated by multivariate analysis, was also placed in Group 1. This sample's placement
near the small cluster containing Samples 41 and 42 (Skegemog Point Site clay samples)
suggests that each of these three clay samples are relatively similar to each other. Fmther,
the placement of these clay samples inside Group 1 indicates that these clay samples have
more similarities with the pottery (clay + temper) samples than with the other clay samples.
This suggests that in certain cases the composition of the clay alone may be sufficient in
determining the groups to which pottery samples belong. It also underscores the variability
in the composition of archaeologically-obtained "raw" clays, particularly if they have been
altered through the addition of temper.

Clearly, several factors are at work within the samples analyzed for this study. The
importance of temper, as well as the non-homogeneity of the temper and/or clay used is
suggested by the difierences in elemental concentrations between samples from the same
vessel. Differences in the composition of the clay samples from the same site are seen in
the large joining distances within the O'Neil Site and Skegemog Point Site clay groups, and
in the placement of a pair of Skegemog Point Site clay samples with the pottery samples.

In addition, the potential use of a variety of temper and clay raw materials by the prehistoric
potters may further complicate the analytical outcomes, resulting in a cluster diagram which
does not replicate the known pottery groups in the samples. Such is indeed the case for the

pottery samples found in Group 1 (see Figure 6).

70

Within Group 1, only six small sub-groups can be found, the remainder of the
group being comprised of single stringers attaching sequentially to these sub-groups. Of
the six sub—groups, only one is comprised exclusively of one ware type. This sub- group
contains five Juntunen Ware samples, including three samples (Samples 53Nl, 53N2 and
54) from one vessel, and two samples (Samples 56 and 57) from two additional vessels.
Given the many sources of variability described above, this sub-group stands out in its
relative homogeneity, which is underscored by the small relative joining distance of 0.037
within this sub-group. The significance of the joining distances of this sub- group is
apparent when compared to the relatively small joining distance of 0.038 for the duplicate
samples SBNl and 88N2 (Mackinac Ware), which comprise a second sub-group within
Group 1.

Three other sub-groups of Group 1 also have relatively small joining distances, but
none are comprised solely of samples atu'ibuted to the same ware, pottery tradition, time
period, or provenience within the O'Neil Site. One sub- group consists of Sample 6
(Skegemog/'Problematic" Ware), Sample 10 (Skegemog Ware), Sample 21 (Traverse
Ware), and Sample 51 (Juntunen Ware), . Another consists of Samples 9 and 10
(Mackinac Ware). A third sub-group is comprised of Sample 1 (Skegemog Ware),
Samples 17, 18, 23, 24 and 31 (Traverse Ware), and Samples 48, 51 and 55 (Juntunen
Ware). Finally, the last sub- group within Group 1 has larger joining distances than the
previously-described sub-groups (from 0.042 to 0.066), and additionally contains both
pottery and clay samples. This set consists of Sample 3 (Skegemog Ware), Sample 14
(Mackinac/Skegemog Ware), Sample 25 and 26 (Traverse Ware), Sample 58 (Juntunen
Ware), and Samples 41 and 42 (Skegemog Point Site clay).

The dearth of clusters within Group 1 representing single ware types suggests three
possible explanations. The first is that the INAA and multivariate analytical methods are
incapable of separating the samples into their proper groupings. This possibility is negated
both by the ability of the methods to group duplicate samples, and by their ability to form a

close
is ill:
with
com

mug

H01

clus

(tag

Site

Pm

the

C01]

71

closely-related cluster of Juntunen Ware samples within Group 1. The second possibility
isthat thevarietyofrawmaterialsused by thepotters, aswell as the inherentvariability
within each raw material, makes this type of composition analysis difficult for within-site
comparisons of vessels. However, the between-sample variability of the pottery sherds is
roughly the same as the variability within the clay samples from each site (see Figure 5),
and is less than the difference between the pottery samples and the clay or temper samples.
Ifthe raw materials sampled are analogous to those actually used to manufacture Late
Woodland pottery vessels, this implies that the mixing of the ingredients obliterated the
unique chemical compositions inherent in the clay and temper samples and produced
pottery which closely resembles all of the other pottery from the site. Alternatively, the raw
materials used to make the pottery may have been unlike those sampled for this study.
However, the fact that two of the Skegemog Point Site clay samples (Samples 41 and 42)
cluster with the pottery samples suggests that the sampling procedure was capable of
capuningatleastsomeofthematerialsusedtomanufacmredteponeryfiomtheO'Neil
Site.

The third possible explanation of the lack of clustering by pottery type is that the
pottery types are not generally congruent to behaviors which would cause the vessels to
group together by type. That is, vessels of different types could be assembled by different
groups of potters from local materials of similar composition. Conversely, vessels
produced by the same group or individual could be manufactured from different materials
procured at the site or elsewhere. In either case the style of the vessel, which determines
the ware category to which it is assigned, would not necessarily be parallel to the chemical

composition of the vessel's paste.

Su

72

Summary of Data

Since the Late Woodland ware categories are based largely on stylistic
considerations (primarily vessel form and design) it should not be entirely surprising that
the composition of these vessels does not conform to the same categories as the stylistic
classification. In the upper Great Lakes region, archaeological sites frequently contain
pottery samples representing a variety of styles and time periods. Occasionally, as in the
case of the O'Neil Site, pottery vessels from separate pottery-making traditions within the
sametimeperiods arefound. Thepresenceofthesedifl‘erentpotterytypes attire same site
has frequently been interpreted as evidence for the occupation of the site by culturally
distinct groups, each with unique pottery-making traditions. What the results of this study
suggest, however, is that the technological and stylistic realms may function independently
of one another, producing stylistic categories which do not match clusters based on
composition analysis.

However, it must also be recalled that the analysis indicated a high degree of
variability in the composition of the raw materials. This variability undoubtedly accounts
for a large proportion of the within-ware variability. One conspicuous exception is the
cluster comprised of three Juntunen Ware vessels (Vessels 5, 37 and 90) which cluster
very closely together, in spite of consisting of sherds found throughout the site at various
levels. The similarity in the composition of these three vessel suggests that they were
manufactured from identical or nearly-identical raw materials, possibly by the same person
or group of people. The fact that sherds from these vessels were found across the site is
somewhat surprising, but may simply be an indication of the amount of post-depositional
churning which occurred at the site. Alternatively, it may represent different areas utilized
by one individual or a single group at the site. If the latter is true, then further studies of
paste composition may reveal that areas in a site used by a single potter or family may be
discernible from the elemental composition of sherds found at the site.

of
tha

(Q

73

Based on the results of the clustering algorithm, several conclusions regarding the
nature ofthe pottery and raw material samples can be drawn. The first is that the O'Neil
clay samples are unlike any of the pottery samples and unlike the clay samples fiom the
Skegemog Point Site. Similarly, with the exception of Samples 41 and 42, the Skegemog
Point Site clay samples are unlike any of the pottery samples. In order to test for the
homogeneity of clay samples collected from the sarrre deposit at the site, multiple nodules
of clay from the same storage box or bag were sampled and analyzed The results indicate
that although these pairs frequently join together, their joining distances are large,
indicating a relatively large degree of difference between these samples. These differences
could be due to inherent differences in the distribution pattern of elements within geologic
clays, especially sedimentary clays. It may also reflect a different pattern of diagenesis
between adjacent nodules of clay. Fm'ther, due to the unfired nature of ”raw" clays, the
absorption of elements from the soil, or conversely the leaching of elements from the
individual clay nodules, may result in larger differences between nodules than between
sherds of fired pottery. Finally, some of the differences between the clay sampl ~-
especially within the O'Neil Site clay samples -- may simply be due to the effects of
statistical counting errors, which are larger for smaller concentrations of elements such as
were found in these samples, or to errors associated with the measurement of
concentrations far smaller that those found in the standards.

The two temper samples analyzed -- Sample 44 which consists of material
collected at the Skegemog Site and identified as "temper," and Sample 30, which is
comprised of relatively large pieces of temper from Vessel 11 -- are not grouped with
either the pottery or the clay samples. However, the temper sample from the Skegemog
Point Site (Sample 44) is more closely related to the pottery samples than it is to either the
temper from Vessel 11 (Sample 30), or to the clay samples. This suggests that material
similar to that collected as temper from the Skegemog Site was used as temper in at least

74

scmeofthepotteryanalyzed, sincethecompositionofthismaterialisnotunliketlratof
some of the vessels found at the O'Neil Site.

However, the fact tlrattlre tempertakerr from Vessel 11 clusters moreclosely tothe
Skegemog Point Site clay group (Group 3) than to either the pottery samples (Group 1) or
the Skegemog Point Site temper (Group 5) is intriguing. It is possible that the collection
method precluded the possibility of obtaining a representative sample of temper from this
vessel. Since only the larger pieces of temper were collected, this may have skewed the
results in favor of elements incluad only in the larger-grained temper fragments.
Alternatively, the larger pieces of temper in this vessel may represent only one of several
tempering materials used in the vessel, one which adds a lower concentration of elements to
thetctalpotterysampletlrandotlrectherpotential tempersin thepottery.

Since the elemental concentrations of both temper samples lie between the higher
pottery concentration values and the lower clay values, it appears that the temper has the
effect of emiching the clay with most of the elements examined. However, this assumes
that the clay samples analyzed were the same clays which were used to make the pottery
vessels from the O'Neil Site. Such an assumption at this time may be unwarranted,
possibly leading to incorrect conclusions regarding the use of clay at the O'Neil Site.

An alternative view is that the temper collected from Vessel 11 (Sample 30)
accurately represents the temper fiom this vessel. If this is the case, it would appear that
the temper has the effect of "diluting" the pottery sample, since the elemental concentrations
for the temper are lower than those of the pottery (clay + temper). If this is the case, then
the concentrations of elements in the clay raw material should be greater than those found in
the temper. This clearly is not the case for the clay samples collected fiom the O'Neil Site
and the Skegemog Point Site. Two options are therefore possible. First, the assumption
that the temper is diluting the clay may not be correct. The second possibility is that the
assumption is correct, but that the clays analyzed (i.e. the O'Neil Site and Skegemog Point

Shtcia

meow

Sand:
the por
b6 Cllht

chfiv

nfixuu
day 53
Sample
tkunr
mum.
(Skegr
knpor
Cathc
“turn
mesk
Samp}
event
afthaf
SUgge

Of C13:

75

Site clays) are not the clays which were actually used to manufacture the pottery found at
the O'Neil Site.

The incorporation of Samples 41 and 42 (Skegemog Site clay samples) as well as
Sample 32 ("pure" clay from Vessel 52, a Traverse Ware pot) in the group containing all of
thepottery samples (Group 1) isalsoofinterest. Given thattlre temperinthepottery may
be either enriching or diluting the clay used to make the pottery, the inclusion of these raw
clays with samples of pottery suggests that these clay samples may have contained small
grains of temper. Alternatively the clays themselves may be sufficiently similar to the
mixtureofclayandtemperastomakethemmore sirrrilartothepotterysamplestlrantothe
clay samples from either site. Unfortunately, neither tempering material nor a pottery
sample with temper and clay was collected from Vessel 52, making impossible a
determination of the effect of known tenrper inclusion on the chemical composition of the
paste. In any case, the difference in the chemical composition between Samples 41 and 42
(Skegemog Point Site clay samples) and the other Skegemog Point clay samples is
important tonote, sinceit suggests that the claysfiom this site arequite dissimilarfiom
each other. (It also suggests that Sample 44, another Skegemog Point Site clay sample,
may not be an outlier, but my represent the other extreme in the range of variation within
the Skegemog Point Site clays.) Whether this dissimilarity is due to added temper in some
samples, or to differences in the elemental content of various nodules of geologic clays, or
even to the presence at one archaeological site of clays transported from other
archaeological sites, the large difference in the cherrrical composition of these clays
suggests that single clay samples fiom single sites do not adequately represent the variety
of clay composition at any one site.

Other conclusions can be drawn from the pattern of clustering within Group 1, the
group consisting of all of the pottery samples analyzed. Unlike the small cluster of five
Juntunen Ware samples noted above, no other cluster of this size is comprised of a single
pottery type. Nor are any ccrnposed of samples from a single pottery tradition, time

period, or

 

Skegemog.
Skcgcmo
Figures 4

samples (

76

period, or provenience on the site. Furthermore, except for Samples 8N1 and 8N2, no
Skegemog Ware samples pair with each other. That is, at least compositionally, the
Skegemog Ware samples appear more like other wares than like the Skegemog Ware (see
Figures 4 and 6). For example, Skegemog Ware samples pair with Mackinac Ware
samples (e.g. Sample 9 and Sample 12) and with Juntrrnen Ware samples (e.g. Sample 1
and Sample 55). Other Skegemog Ware samples are grouped with Juntunen-Traverse pairs
(e.g. Sample 3 with Samples 58 and 26, and Samples 10 and 6 with Samples 21 and 51).
In even larger clusters within Group 1 Skegemog Ware falls into clusters with all three of
the other ware types.

The pattern of grouping for the Mackinac Ware is not dissimilar to that of the
Skegemog Ware. No Mackinac Ware samples pair with each other. This includes all of
the samples designated "Mackinac" and "Mackinac/Skegemog." In one instance a
Mackinac Ware vessel is paired with a Skegerrrog Ware vessel (Sample 12 with Sample 9).
Also one Mackinac Ware sample is clustered with a Traverse Ware pot and a sample of clay
from the Skegerrrog Point Site (Sample 14 with Samples 25 and 42). All other Mackinac
Ware samples are grouped with larger clusters containing the three other pottery types.

Traverse Ware samples, however, do pair with each other, as shown by Samples
17 and 18, and Samples 23 and 24. They also pair with Juntunen Ware samples, as
exhibited by the pairing of Sample 26 with Sample 58, and the pairing of Sample 21 with
Sample 51. Traverse Ware pottery also forms a small cluster of five Traverse Ware and
one Juntunen Ware samples (i.e., Samples 17, 18, 23, 24, 31, and 48). This small cluster
has joining distances of 0.026 - 0.031 -- the smallest joining distances for any cluster. In
another instance Traverse Ware is found in a cluster of two Traverse Ware, two Skegemog
Ware and one Juntunen Ware samples (Samples 6, 10, 21, 29, and 51). As noted above,
the sample of "pure clay" from a Traverse Ware pot (Sample 32) lies within Group 1. It
falls in a cluster containing samples from all four pottery types. The remaining Traverse

Ware samples (Samples 22 and 27) group with large clusters containing all pottery types.

Ex
itself. H0}

 

SS3N2, Sf
within this
Juntunen c
Juntunen V
In larger cl I
POW WP

In

 

pairs and s
manufac
related p0
However, i
for analys'
Within the r
In t
is a great (it
Vm&mWi
in the pone
Several P05

1.

77

Except for Samples 53N1 and 53N2, the Juntunen Ware pottery does not pair with
itself. However, as noted above, it does form one distinct cluster of five samples (SSBNI,
853N2, SS7, 856, and SS4) representing three separate vessels. The joining distances
within this cluster (0.027 - 0.037) is very small; only the above-mentioned Traverse /
Juntunen cluster has smaller joining distances than this five-sample Juntunen cluster.
Juntunen Ware samples also group with Traverse Ware samples, as previously described.
In larger clusters, the Juntunen Ware samples are found in clusters containing all other
pom WP“

Interestingly, the more recent Late Woodland pottery in this study tends to form
pairsandsmallclusterswitlrthesamepotterytypes. Thismayindicatethatmorevessels
manufactured with the same raw materials (perhaps at roughly the same time by the same or
relawd potters) are more likely to be found at the O'Neil site in the later time period.
However, it may also simply be a consequence of the larger number of samples available
for analysis from the late Late Woodland period, resulting in a higher probability in pairing
within the ware type.

In general, the clustering pattern of the clay and pottery samples indicates that there
is a great deal of variability in the clay and pottery samples. The fact that the within-pottery
variability is similar to the within-clay variability is puzzling, since the inclusion of temper
in the pottery samples is presumed to contribute additional variability to these samples.
Several possible explanations for this phenomenon are possible.

1. The pottery was made from a more uniform clay (either a different

O'Neil Site clay or clay fiom another location).

2. The clay fiom the O'Neil and Skegemog Sites is "prepared clay" with
more or less material (temper) added to it. Because of the presence of
this added material, the variability within the clays is as great as that
within the pottery group.

3. The clay samples analyzed are "rejects" which were considered by the
prehistoric potters to have been unsuitable for pottery-making. The

78

fact that these pieces of clay were apparently discarded may support
this explanation.

4. The mixing of clay and temper results in a fortuitous mixture which is
relatively similar for all pottery.

The last possibility seems highly unlikely, since it would imply that each potter throughout
time and space had similar mental "recipes" for pottery which resulted in similar paste
composition regardless of the composition of the raw materials. This possibility also
seems to be negated by the results reported by Trigger, et al. (1980) and Clark (1991)
which indicate that variations in pottery composition occur over large regional areas.
Therefore, it appears that the clay collected from the O'Neil Site and analyzed by INAA
was not the clay used to manufactrue the pottery at the O'Neil site, or, if it was, the clay
analyzed was so modified as to render a definitive conclusion impossible. In the case of
the Skegemog Point Site clay it is tempting to suggest that the clay nodules which clustered
with the pottery samples (Samples 41 and 42) represent the clay used to manufacture these
vessels. However, the variability within the Skegerrrog Site clay samples is greater than
that of the O'Neil Site clay samples, and therefore any conclusions regarding the nature of
individual clay samples from the Skegemog Point Site are at best tenuous.

With very few exceptions, the pottery samples grouped more fiequently with other
pottery types (including types of a dissimilar pottery tradition or time period) than they did
with the same pottery types. The duplicate samples (Samples 8N1 and 8N2, and Samples
53N1 and 53N2), however, join at the first level of joining. Therefore for a given portion
of a given sherd, the results obtained by INAA are reproducible. For multiple samples
taken from a Skegemog Ware vessel (Samples 8-10, representing Vessel 57) the analytical
results were very similar to one another. Likewise, the samples taken fiom the same
Juntunen Ware vessel (Samples 53-54, representing Vessel 90) gave very similar
composition profiles. However, for the multiple samples collected from a Traverse Ware
vessel (Samples 23-24, representing Vessel 85), only two of the three samples show close

 

similarity
the possit
vessel T
prehiston
AltcmzuhI
diagcncsi I

53316 VCSS

Ofptoblcr

Validit)
Gi
with Egan

 

O'Neil Sit
Obtained .

79

similarity in composition. The relative dissimilarity of the third sample (Sample 25) raises
the possibility that one or more factors may be operating differentially within a single
vessel. The results obtained may be due to the incomplete mixing ofclay and temper by the
prehistoric potter, leading to differential sampling of clay and terrrper from the three sherds.
Alternatively, the variability within these three samples may be due to differential
diagenesis across the site, or even to the incorrect assignment of these three sherds to the
same vessel. The analytical results from this single vessel, therefore, illustrates the range

of problems associated with each of the vessels analyzed.

Validity of Hypotheses

Given the results obtained in this study, only tentative conclusions can be drawn
with regardtothe behaviorofthepotters manufacturingtheLate Woodland pottery ofthe
O'Neil Site. In fact, most of the hypotheses put forth are not supported by the results
obtained. The first hypothesis states that:

1: There are no significant differences between the chemical

compositions of the pottery from each of the residential occupations
(i.e. the Mackinac, Traverse and J untunen pottery).

Although it appears that each pottery vessel from the residential occupations is similar to
every other vessel from residential occupations, the significance of this result is
questionable, since all of the pottery samples «- both those from residential and those from
logistic occupations - are chemically similar to each other. No clear differentiation
between the presumably locally-made pottery from the residential occupations and the
purported non-locally—made pottery associated with the logistic occupations was possible.
This negates the validity of the fourth hypothesis which states that:

 

GroUp ‘.

FBI [3

reSu]

80

4: There is a significant difference between the chemical compositions
of the pottery from the logistic occupations (i.e. the Skegemog Ware
pottery) and the pottery from the residential occupations (i.e. the
Mackinac, Traverse and Juntrmen pottery)

As previously shown, all of the pottery samples analyzed are compositionally similar to one
another, and the samples from residential occupations are intermixed with samples from
logistic occupations in the cluster of Group 1.
Likewise, the second hypothesis, which states that:
2: There are no significant differences between the chemical
compositions of the local (O'Neil Site) clay and the pottery from the

residential occupations (i.e. the Mackinac, Traverse and Juntunen
pond?)

is not supported, since the pottery from the residential occupations is grouped separately (in
Group 1) from the O'Neil Site clay samples (Group 2).

The final hypothesis to be considered (hypothesis 3) states that:

3: There is a significant difference between the chemical compositions

of the non-local (Skegemog Point Site) clay and the pottery from the
residential occupations (i.e. the Maclo'nac, Traverse and Juntunen

POW)-

For most of the Skegemog Point clay samples the hypothesis is supported. However, for
two Skegemog Site clay samples (Samples 41 and 42) which are found in Group 1, the
results do not support the hypothesis, since these non-local clay samples are chemically
similar to the pottery from residential (as well as logistic) occupations.

The inability to support the majority of the hypotheses indicates that the underlying
assumptions about the nature of the clay and pottery samples may be invalid. That is, the
homogeneity of the clay samples, as well as the relationship of the clay samples to the
pottery samples may not be as clear-cut as expected The results also suggests that the
assumptions about the regularity of past behaviors may not be correct. The pattern of clay

procutm

the come. .

assumpti

 

81

procurement and manufacture into pottery containers may be highly variable, even within
the context of a residential community occupying a site with clay resorn'ces. These

assurrrptions will be dealt with in the following summary chapter.

Errors

h
being an
material:
were batc
assumpt
the tesul
35$umpt
extent 01
We:
Wat f0
material
Pithisto

SUMMARY

Errors in Original Assumptions

Inundertakingthisprojecaseveralassumptions aboutthenatmeofthematerials
being analyzed, as well as about the behaviors which produced the archaeological
materials were necessary. Some of these assumptions were stated from the outset, and
were based on the results obtained by previous investigators. Other underlying
assumptions, however, were not obvious until after the analytical data were reviewed and
the results were found not to be in accord with the expected results. Some of these
assumptions are practical in nature and involve simply the choice of sampling tools or the
extentofsampling witlrina site. Others aremoretheoreticalinnature andrequirea
modification of the research design in order to control for them in future investigations.
What follows is a list of the underlying assumptions regarding the nature of the raw
materials and finished pottery, as well as assumptions relating to the behavior of the
prehistoric potters and analytical procedtne.

The local clays are compositionally difl'erent from the non-local clays. This
assumption held true for the O'Neil Site clays versus the Skegemog Site clays. However,
the O'Neil Site clays were extremely sandy and crumbly, while the Skegemog Site clays
were more compact and firm. This raises the question of whether the O'Neil Site clays
contain added sand (possibly as a tempering material) which would have the effect of
diluting the clay substrate. The fact that the "fired sand found beneath [the] clay" at the
O'Neil Site (Sample 36) falls in the middle of the O'Neil Site clay cluster supports the
premise that the O'Neil Site clay samples may be showing more similarity to its sand
components than to its clay components. Of course, it is also possible that the O'Neil Site
clays are by nature very sandy, and the elemental composition obtained by INAA is
representative of the composition of the geologic clays found in this area. In order to

definitively resolve this issue, a systematic sampling and analytical program is required in
82

83

order to determine the elemental composition of geologic and archaeological clay samples
within a small region such as the so-called Traverse Corridor.

The clays from one site are similar to themselves (i .e. these clays are fairly
homogeneous). Surprisingly, what this study revealed is a great deal of variability in the
composition of the clays collected from these archaeological sites. Whether this
heterogeneity in the clays from a particular site is due to the variability of the parent clay
formation, to the addition of temper to the "raw" clays, or to the importation of clays from
other sites cannot as yet be determined. Again, a systematic analysis of geologic and
archaeological clays from a variety of sites would be helpful in ascertaining the cause of
this variability.

Unworlced clays were not transported long distances, but instead were mined near
the area where the vessels were made. One of the major premises of this study was that
locally-made pottery could be distinguished fiom non-locally-made pottery on the basis of
the composition of the ceramic paste. In addition to assuming that local clays were similar
to themselves and different fiom non-local clays, it was also assumed that the vessels made
from non-local clays were not manufactured at the O'Neil Site, but instead were made at
another site near the source of the clay raw material. While this may indwd be the case for
the majority of vessels, it may not be true in all cases. In particular the large variability of
the composition of the Skegemog Point Site clays suggests that some of these clay raw
materials could have been brought to the Skegemog Point Site from another location.
Whether this clay was then used to manufacture pottery is still an open question, but it
suggests that clay (and temper) procurement behavior was probably more variable than
once believed.

The clays collected from the archaeological context are representative of clays used
to manrfacture prehistoric pottery vessels. This assumption is related to the above two
premises that clays found locally are in fact local clays and are homogeneous in

composition. However, it also includes the assumption that potters were exploiting the

84

same sources of geologic clays throughout the prehistoric period, and that a sample of
unused clay formd in an archaeological context is representative of all of the clays used to
manufactrne pottery vessels at that site. At this point it is not clear whether the potters who
occupied the O'Neil Site used the clay sources within their procurement area in the same
marrnerandtothe sameextentthroughoutthelateWoodlandperiod. Norisitkrrown
whether the clay som'ces available to Late Woodland potters remained the same throughout
this period, or whether some clay deposits became exhausted during this time. Analyses of
presently-available geologic clays as well as archaeologically—derived clays could shed
some light on this question, provided that the archaeological clays have not been
extensively modified by the addition of temper.

The clays from the ONeil Site (and/or the Skegemog Point Site) are compositional-
ly similar to some of the pottery samples. None of the pottery analyzed proved to be
similar in composition to the clay from the O'Neil Site. Likewise, rrrost of the clay samples
from the Skegemog Point Site were unlike any of the pottery samples. One simple
explanation for this may be that the pottery was not manufactured fiom the O'Neil Site
clays or from most of the clay found at the Skegemog Point Site. However, the
complexities of the clusters discussed previously suggests that a such a facile explanation is
not adequate. What is apparent, though, is that behaviors involved in the production of
prehistoric pottery are far more complex than originally assumed, and that hypotheses
dealing with the chemical composition of pottery sherds must be more carefully
constructed.

Temper added to the raw clays does not significantly alter the chemical composition
of the resultant pottery. This assumption appears to be incorrect, although a definitive
answer is not possible. The fact that the temper samples from the Skegemog Point Site lies
closer to the pottery samples than do all of the clays from the O'Neil Site and most of the
clays fiom the Skegemog Point Site suggests that the elerrrents contained in the temper may

be adding constituent elements to the pottery samples. However, the inclusion of the

85

"pure" clay sample from Vessel 52 (Samples 32) as well as the inclusion of the Skegemog
Point Site clay samples (Samples 41 and 42) with the pottery samples in Group 1, suggests
thatinsomecasestempa'may notbenecessary fordeterminingthegroup membershipof
the sample.

Temper is evenly dispersed throughout the vessel, its composition is homogeneous.
and it is constant between vessels. All of these assumptions are negated by the
observations of the pottery paste made during the sample collection step. Both the size of
the tempergranulesandtlre amountoftempercontained in the sherds was foundtovary
between vessels, although there was a greater degree of similarity in the physical
composition of the sherds within the same pottery types. Also noted visually were the
individual light- and dark-colored grains which comprised the crushed- granite temper. The
lack of temper homogeneity suggested by the light- and dark-colored fragments is also
supported by the results of the mathematical test of mixing performed on samples of
pottery. These results indicate that the pottery matrix (composed of clay and temper) is the
result of more than one mixing process, that is, either more than two components were
mixed together to produce the pottery, or the clay and temper themselves were made up of
several components. Finally, as discussed previously, the inclusion of varying amounts of
temper in the samples analyzed by INAA is suspected to have contributed, at least in part,
to the large variability in the composition all four pottery wares. One extreme end of the
spectrum of temper inclusion was seen in Sample 32 which contained no visible temper.
Other samples (for example Sample 29) contained such large amounts of temper that the
larger pieces in the original sample were removed prior to placing the material in the sample
bottles. In both cases the remaining pottery sherds from these vessel were not examined to
determine if the amount of temper in the samples were representative of the amount in the
vessels, but it is suspected that the amount of temper does vary within each pottery vessel.
That the relative amount of temper varies between vessels is undeniable. Therefore, if the

inclusion of temper has either an additive or diluting effect on the final composition of the

86

sherd, then varying amounts of temper in the sample will yield varying results. Although
Mommsen and others (Mommsen 1981; Mommsen, et al. 1988) have established means of
mathematically sorting out temper from clay in INAA results, this method requires analyses
of known temper and clay raw nurterials. Certainly, additional studies of both temper and
clay composition would assist in eliminating this problem for future research.

Temper can be avoided in the sampling process. Even with the use of a fine-tipped
dental bit it was found that temper could not be completely avoided. By using these bits
only the larger pieces of temper could be avoided and/or removed from the sample before
placing the remainder in the collection bottle. But because even the small grains of temper
are larger than the individual clay particles, it may be possible, in future investigations, to
separate the clay from the temper with soil sieves, with a mechanical shaker, or by means
of flotation using ultra-pure water. Although such procedures would add significantly to
the sample collection time, the resulting clay and temper samples obtained could more
easily be used to interpret past behaviors relating to the procurement of raw materials and
their manufacture into pottery.

The addition of water by potters during the preparation of the clay does not alter the
chemical composition of the finished pottery vessel. This assumption was not tested in this
study but reportedly (Thomas Vogel, personal communication 1991) the addition of flesh
water would not add significant amounts of trace elements to the pottery matrix. This may
be a factor, however, for pottery made at locations where sea water might have been used
in the manufacturing process.

The sampling procedure will not contaminate the samples. In an attempt to avoid
temper as much as possible in the collection of the samples, hardened steel (Tungsten-
Vanadium) dental bits were used to collect the samples. However, the bit material proved
to be softer than the temper in the pottery sherds, and the bits became heavily abraded by
the sherds. Therefore, although care was taken to avoid contarrrinating the samples with
other materials, the samples became contaminawd with metal filings from the dental bits.

87

Forumately the extent of the corrtarrrination could be determined by analyzing representative
bits, but the contamination ofthe samples precluded the use in this analysis ofcertain
elements (notably Iron) which might have been useful in separating the samples into more
meaningfulclusters. Onesolutiontothisproblemmightbetouseaharderdrillbit, suchas
a carbide bit, but the brittleness of such materials may also lead to contamination of the
samples. Anotheralternativeistouseanagatemortarandpestletobreakapartthepottery
sherds. However, this method would not pernrit the removal of the surface material from
the sherd prior to sampling -- material which could include soil and other potential
contaminantsnotdesiredinthefinalsample. Also,carewou1dneedtobeexercisedin
ordertoavoidbreakingupthegrainsoftemperto suchan extentastomake tlreirseparation
from the clay particles impossible. All in all, where temper inclusion can affect the final
analyticalresults,itappears thatthe bestaltemativeistouseadrill bitmadefromtlre
hardest material available, to incorporate a method or removing small-grained temper fiom
the clay, and to analyze the drill bit(s) used in order to determine the extent of possible
contamination of the samples with the drill bit(s).

The elements analyzed for and used in the clustering program are important in
difl’erentiating between difl’erent types of pottery and between clays fi'om difi'erent sites.
This assumption was found to be at least partially correct, since the analytical procedure
was able to differentiate between the clays from the O'Neil Site and those fiom the
Skegemog Point Site. However, the same analysis clustered all of the pottery into a single
group, with only small sub-groups within it. Whether this was due to characteristics
inherent in the pond? matrix, or to the choice of elerrrents used for the clustering program
cannot be determined. Other investigators have used a variety of elements to specify
groups within the sample population (see Appendix F for a listing of the elements used by
these investigators). At present there is no consensus as to which elements provide the best
differentiation between pottery types or clay sources, although for the Michigan clays the

concentrations of Iron, Magnesium, Potassium, and Sodium are thought to be effective

88

markers for differentiating clays from different sources ( Randy Schaetzl, personal
communication 1991). In this study Magnesium and Potassium were not determined, since
their half-lifes were too short for the method used. Sodium is not detectable using the
standardsattheMichiganMemorial PhoenixProject. Finally,Ironwas notusedin thedata
analysis due to the contamination of the samples with rrretal filings fiom the drill bits.

Although the concentration of Iron may be useful in differentiating one clay source
from another, the addition of crushed granite temper (Lovis 1973) would significantly alter
the concentration of Iron in the pottery since granite and other rocks of igneous origin are
rich in Iron (Randy Schaetzl, personal communication 1991). The inability of Iron to
difi‘erentiate the pottery from the O'Neil Site is borne out in Table 9, which lists the mean
Iron concentration of each sample in increasing order. As this table illustrates, the
concentration of Iron in the pottery samples cannot distinguish between pottery types, and
is not helpful in determining whether or not the O'Neil Site clay or the Skegemog Point Site
clay was used to manufactrne the pottery from the O'Neil Site. The usefulness of
Magnesium, Potassium, or Sodium in differentiating clays from nearby sites or in
distinguishing one group of pottery from another has yet to be determined.

The pottery sherds analyzed belong to discrete ware types discernable on stylistic
grounds. The validity of this last assumption is perhaps the most difficult to determine,
since the analytical results of the pottery samples were inconclusive. Although it has been
argued that much of the variability seen in the paste composition of the forn' pottery types is
due in large part to methodological problems and variable behavior on the part of the
potters, it could also be argued that some of the variability is due to the classification of the
pottery itself. In many region in Michigan, the prehistoric pottery frequently exhibits
structural and decorative attributes sirrrilar to those of adjacent areas (see for example
Brashler 1981). Similarly, the evolution of pottery design from one style to another over
time rarely proceeds with sharp breaks between pottery types. Instead, pottery styles, like

other aspects of material culture, are often fluid and ever-changing. Because of this, there

89

Table 9 - Mean Iron Concentrations, in Increasing Order

Spl Sample Site or Mee- Fe Grp Spl Sample Site or Mean Fe Grp
Type Ware Type Cone. # Type Ware Type Cone.

O'Neil
O'Neil
O'Neil
O'Neil
O'Neil
O'Neil
Pt

50 Juntunen
22 Juntunen
6 Prob
27 Traverse
51 Juntunen
15 Mackinac
4

BN2
1 l
7
22
56
9
32
2

333333..3
IdHHuhilflO-IMHHHt-It‘tdI-l—I—IHHI-IHHH

CkStd
CkStd
CkStd
CkStd

>>>>

Drill Bit
Drill Bit
Drill Bit
Drill Bit
Drill Bit
Drill Bit

>>>>>>

2
2
2
2
2
2
4
2
3
6
3
3
3
l
l
2
l
l
l
l
l
l
l
1
l
l
l
l
1
1
l
1
l
l
l
l

 

90

is the risk of lumping "transitional" pottery types into earlier or later pottery styles, simply
because the classificatory scheme is not sufficiently detailed to accommodate it.

Likewise pottery vessels with design elements borrowed from adjacent areas have the
potential of being misidentified and assigned to the wrong cultural group. Finally, and
perhaps because of these problems, there is the potential of having classificatory schemes
which are interpreted differently by different investigators. Some evidence of this was seen
with this collection, as exemplified by the conflicting pottery types assigned to Vessel 17
(Sample 13), Vessel 28 (Sample 5), Vessel 32 (Sample 6), and Vessel 53 (Sample 14).
The extent of these potential classificatory problems and their effect on matrix studies of
this type is as yet undetermined. However, future studies may reveal that ceramic
categorization by paste composition is not precise enough to separate the pottery discarded
at a single site. Conversely, once the sources of variation in paste composition analysis are

sufficiently well understood, this analytical method may prove useful in the classification of
POW types.

Significance of Results

This study attempted to discern from the chemical composition of the O'Neil Site
cerarrric assemblage the residency of the prehistoric potters in the hopes of shedding light
on the question of resource utilization among the Late Woodland hunter-gatherers of
northern Michigan. Instead, what was found was a ceramic collection which could not be
easily differentiated into sub- groups on the basis of chemical composition. In addition, a
high degree of variability was found in the samples of archaeologically-derived clays from
the O'Neil and Skegemog Point Sites. Ftu'ther, little correspondence between these clays
and the pottery samples from the O'Neil Site was apparan Although the expected results
of this study were not forthcoming, some tentative conclusions about the behavior of the

prehistoric potters can be suggested.

91

Although the O'Neil Site was re-occupied several times, it was never intensively
occupied, and the re-occupations took place over a period of several centuries (Lovis 197 3,
1991). The site is known to have been occupied throughout the Late Woodland period by
groups with different pottery u'aditions. In addition to having their own pottery making
u'aditions, these groups may also have had different strategies for obtaining raw materials
for pottery unnufacturing. In this light, the lack of large distinct clusters corresponding to
one pottery type or even to one pottery u'adition or time period is not entirely unexpected.
given the mobility of the groups in question and the access to a potentially large number of
clay and temper resources.

The lack of such clusters implies that the people making the O'Neil Site pottery
were obtaining their raw materials from a variety of locations, most likely including the
Skegemog Point Site, and possibly including the O'Neil Site itself. In addition, the
occupants of the O'Neil Site may have brought finished vessels to the site, but the evidence
for this is still inconclusive. What is evident, however, is that the potters manufacttuing
the vessels found at the O'Neil Site were not limited to discreet somces of clay and temper
which would manifest themselves in clear clusters of vessels and a tighter grouping of
unfired clays. Instead, one sees pottery types whose ranges of variation are as great as the
range of variation of all the pottery samples combined. This implies not only the use of a
variety of raw materials, but potentially a mixing of pottery styles and populations,
resulting in pottery whose stylistic attributes are not congruent to the elemental composition
of its paste.

The unexpectedly high variability in the clay samples, particularly those from the
Skegemog Point Site, may simply be a reflection of the variability of the geologic clays
available in the immediate vicinity. Alternatively, it may indicate that raw clays were
transported to the site from other locations within the seasonally-traversed territory of the
groups inhabiting the site. Two reasons for u'ansporting raw clays fiom one occupation to

another can be envisioned. The first is that although clay is ubiquitous in Michigan, those

92

clays desired for pottery-making may not have been abundant in all areas. The second
reason is that the difficulty in transporting clay may have been offset by the greater
difiiculty in transporting finished pottery from one location to another. Fired pottery,
particularly the pottery of the Upper Great Lakes which is fired at low temperatures, is
relatively fragile. Raw clay, on the other hand, is also relatively fragile in its dried state,
but with the addition ofwater it can easily be re-molded into a lump ofclay, or, with the
Ma addition of temper and subsequent firing, it can be fashioned into a pottery vessel.
Although it is not suggested that all pottery vessels were made from lumps of clay which
were carried from one site to another, it is probable that some vessels were made from such
imported raw material. A final possibility is simply that the clay samples obtained from the
O'Neil Site and the Skegemog Point Site were representative of various stages of clay
preparation. Thus samples 41 and 42, which were noted as having some temper in them,
may represent the stage of clay preparation just prior to its being shaped into a vessel,
whereas other clay samples might have been discarded or lost by the potter prior to the
addition of temper. Further investigations into the range of variation of the natm'ally-
occm'ring geologic clays are necessary before these premises can be considered more fully.
Given the large variability in the analytical results of both the clay samples and the
pottery samples, the small tightly-clustered group of three Juntunen Ware vessels,
(represented by Samples 53Nl, 53N2, 54, 56 and S7) is quite unexpected. This small
group of vessels represents the only cluster of pottery whose members are comprised of the
same pottery type. Yet the similarity between the members of this group is greater than that
between some lumps of clay collected together at the same site. This suggests that these
vessels were manufactrned from identical or nearly-identical raw materials, and possibly by
the same potter or group of potters during a single work episode. The question of where
these vessels were manufactrned still cannot be not resolved, but a close temporal and
spatial relationship between these vessels is suggested. A novel application of this method
of analysis thus emerges from these results. Although the analysis of ceramics for gross

93

similarities between pottery types has proven to be difficult due to the potential variability
of the clay and temper, this technique may be useful in determining whether any vessels at
a given site were made with the same raw materials, and presumably by the same potter(s)
atroughly the same time period. Finally, the near absence of clusters of the same p0ttery
typemaybeafrn'therindicationofthe wide availability ofraw materials forpottery
making, as well as a potential marker of the relatively mobility of the groups manufacturing

thepottery.

Future Directions

Previous investigations of paste composition have shown that INAA is useful in
helping to identify somces of clays used for pottery making among settled populations. It
has also been used with some success on a regional level for more mobile hunter-gatherer
populations. However, more work is needed to determine whether this method is useful in
the intra-site analysis of hunter-gatherer behavior.

Of primary importance for futru'e work on paste composition in the Great Lakes
area is a determination of the name] range of variation in the chemical composition of
geologic clays collected from particular regions. This would require an intensive and
systematic sampling program of geologic clays tluoughout Michigan and other areas.
Additionally, it would be useful to include samples of archaeologicallysderived clays in
order to determine whether these clays are representative of nann'al clays from the region.
The analysis of archaeological clays alongside natm'al clays could also help establish
whether the archaeological clays had been modified in prehistory by the addition of other
materials. It may also help determine whether the clays found in archaeological contexts
are native to the region of study.

In addition to analyzing the clays fiom various regions, it would also be useful to
analyze Late Woodland pottery collections fiom nearby sites in order to establish the

94

relationship between the pottery from the O'Neil Site and similar pottery from neighboring
sites. This would help determine whether or not the pottery collected from within small
geographic areas in Michigan can be differentiated on the basis of paste composition. It
would also help establish the general relationship between raw clay samples and pottery
samples recovered from various archaeological sites.

Finally, any ftu'ther work on the analysis of pottery composition within small
regions should take into account the effect of temper on the final concentration of elements
in the pottery samples. If temper cannot be avoi®d in the sampling process, care should
be taken to analyze enough temper samples to enable the effects of this material to be
"subtracted" from the communion values of the whole pottery samples. Alternatively,
methods for separating ground temper from powdered clay could be explored. If such
separations are possible, this would enable the individual components of pottery to be
analyzed separately, and would permit the comparison of pottery clays to geologic clays.

The use of Instrumental Neutron Activation Analysis for the determination of
pottery paste composition is relatively rare in the Upper Great Lakes. Research has
indicated that this method can be useful in large-scale regional studies relating the
composition of pottery paste with past human behavior. This study has shown that
archaeologically-derived clays from closely-spaced sites can be differentiated by INAA.
However, the determination of the relationship between these clays and archaeologically-
related pottery sherds is difficult to establish, presumably due to the effect of temper which
is incorporated into the pottery samples, and possibly also because of the natural variation
between clay samples collected from the same site. It is hoped that future work in the area
of paste composition analysis can resolve some of the methodological problems identified
in this study.

NOTES

The MSU Museum, Anthropology Division catalogue number consists of a unique .
accession number which identifies the site, as well as a series of decimal suffixes whrch
identify the provenience of the artifact within the site.

The Minimum Vessel Sheets (Lovis n.d.) are internal worldng documents originally
used to assist in the analysis of the O'Neil Site pottery. The vessel numbers assigned to
each vessel are recorded only on these Minimum Vessel Sheets and in the storage boxes
where each individual vessel is stored. The original Minimum Vessel Sheets do not
indicate the type/variety designation of each vessel. In fact, no listing correlating vessel
numberstopotterytypescouldbeobtained.

Densities were not measured, but were noted visually as gross difi‘erences in volume
(height of sample in the sample tube) relative to the weight of the samples in the tubes.

This process is performed in two steps. First the K, or constant (in cps/mg/ppm) for
each element in each standard is calculated according to the equation:

[(Peak area std " BaCkgrOund area std) / live counting time std] e (0-693)(l / T)
K std =

 

(Mass std) (Concenuation std)
where t= decay time in seconds for each element
(relative to the time of irradiation, t o)

and T = half-life in seconds for each element

In addition, the counting error associated with each peak in the standard is calculated
according to the formula:

95

96
l
ErrorofPeakstd = [2(backgroundareastd)-t-areaofpeakstd]I2

This error value for the peak is then plugged into the formula for the constant K std in
order to yield an error associated with K std. or Error ofK std. An additional error
associated with the concentration of the standard (i.e. the "error" of the concenu'ation
value) is not used in this analysis because the current software package is incapable of
handling this variable.

Other potential sources of error which affect the calculated values of K include variations
in the neutron flux received by each sample due to the sample's position in the reactor,
errors in weighing the samples and standards, errors related to the determination of the
concentration of elements in the standards (leading to later calculation errors), and errors
due to non-homogeneous standards (Meyers and Denies 1972: 21). Due to the many
sources of error, Birdsall (personnel communication 1991) estimates that the final
concentration values obtained from INAA at this facility are within t 10% of the true
values for these concentrations.

OncetheK stdvalues and associatedenorvalues foreach elementin the standardare
calculated, the mean K std ( or K std "bar") is calculated using the equation:

Estd = (E: Kstdi) / n

i=1
where n= numberofstandardsin each batch

and the error of K std (error K std "bar" ) is calculated according to the equation:

Errorof I-(std = {(5 [ErrorosttdiPN/z} /n

1:!
(Edward Birdsall, personal communication 1991).

The second part of the computer calculation involves the calculation of the concenuations
of elements in each sample. This is done according to the following equation:

[(Peak area spl - Background area spl) / live counting time SP1] e (0693th T)
(Mass spl)

 

Cone 3p] =

 

Kstd

97

where K std , t and T are as defined above.
(Edward Birdsall, personal communication 1991).

Likewise, the calculation of the counting error associated with each peak in the sample is
analogous to the calculation performed for the standards. Thus

1
ErrorofPeakspl = [2(backgroundareasp1)+areaofpeaksp1] ,2

The concentration represented by this enor estimate is obtained by plugging this value
into the formula for determining the concenu'ation of each element in the sample, or
Cone 3131- Once the concentration errors related to each peak in the standards and
samples are established, a "standard deviation" (or error estimate for each concentration
measm'ement) can be determined for each concentration determination according to the

equation:

Std. Deviation = {(21,1 [Error of {($412 + [Error of sample]2 )1/2} / n
1

The concentration values for each of the elements, as well as the associated standard
deviation values for each sample are reported in Appendix A.

5 Ifthe greatest amount oflron in the samples is 7.1%, and the least amount of Iron in the
bits is 82.5%, then the greatest level of contamination of the samples by the bits is

[ (7.1% Fe in samples) / (82.5% Fe in bits) ] x 100% = 8.6%

APPENDICES

APPENDIX A

98
APPENDIX A

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Barium Barium Barium Barium Barium Barium
7 7 496.3

Cone % Std Dev Cone ‘5 Std Dev Cone % Std Dev
1

 

99
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Barium Barium Barium Barium Barium Barium
7 7 496.3 496.3
Cone % Std Dev Cone it Std Dev Cone % Std Dev

not not
not not
not not
not not
not not
not not

§§§§§§

 

100
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Bromine Bromine Bromine Bromine Lanthanum Lanthanum
.4 4 15 15

Cone ‘5: Std Dev Cone ‘5 Std Dev Cone % Std Dev
< not

not not

BO! 00!

 

101
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Bromine Bromine Bromine Bromine Lanthanum Lanthanum
.4 .4 15 15
Cone 5 Std Dev Cone % Std Dev Cone % Std Dev

§§§§§§§§§§§§§§§§

not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not

§§§§§§§§§§§§§

3 5 § 5 5 5
E 5 5 § § §

5 5 § § 5

 

102
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY

Lutetium Lutetium
.7 .7
Cone ‘1» Std Dev Cone ‘5 Std Dev Cone % Std Dev

5

AAAAA

<
<
<
<
<
<
<
<
<

AAA

 

103
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Lutetium Lutetium
140.5 40.5 .7 .7
Cone ‘5 Std Dev Cone ‘5 Std Dev Cone ‘5 Std Dev
< not not stds not
not not
< not not
not not
not not
not not
not not
not

not
not

not
not
not
not
not

not

not
not
not
not
not
not
not
not
not

5

555555
3

i

 

104
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Samarium Samarium
1.2 1.2 1.2 03.2 03.2
‘5 Std Dev Cone ‘5 Std Dev Cone “lb Std Dev

 

105
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Samarium Samarium
1.2 1.2 1.2 03.2 03.2

it Std Dev Cone % Std Dev Cone % Std Dev
1 1
not

not not not
not not not
not not not
not not not
not not not
not not not

 

106
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Urani- Uranium Uranium Uranium Uranium Uranium
.5 .5 06.1 06.1 .7 7

Cone % Std Dev Cone ‘5 Std Dev Cone ‘5 Std Dev
1

5.

not
not
not
not
not
not
not
not
not
not
not
not

 

107
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Uranium Uranium Uranium Uranium Uranium Uranium
SPL Q .5 .5 06.1 06.1 .7 .7
Cone ‘5 Std Dev Cone % Std Dev Cone ‘5 Std Dev
1 not
not
not
not
not

not not
not not
not not
not not
not not
not not

 

108
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY
Ytterbium Ytterbium Ytterbium Ytterbium

Cone 9b Std Dev Cone % Std Dev
1

 

109
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 1 WEEK OF DECAY

Ytterbium Ytterbium Ytterbium Ytterbium
SPL #

Cone it Std Dev Cone ‘5 Std Dev

not
not
not
not
not
not

353533

 

l 10
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Cerium Cerium Cesium Cesium
690.7 45.5 45.5
% Std Dev Com: ‘5 Std Dev Cone

l

 

ll 1
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Cerium Cerium Cesium Cesium
690.7 45.5 45.5
‘5 Std Dev Cone ‘5 Std Dev Cone

not not
not not
not not
not not
not not
not not

 

1 12
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Cesium Cesium Chromium Chromium Cobalt
.7 .7 1332.5
Cone ‘5 Std Dev Cone '5 Std Dev Cone

 

1 13
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Cesium Cesium Chromium Chromium Cobalt Cobalt
.7 7

Cone ‘5 Std Dev Cone ‘5 Std Dev Cone % Std Dev
1

not
not
not
not
not
not

 

l 14
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY

085.6 112.2 1112.2 1408.1
‘5 Std Dev Cone ‘5 Std Dev Cone

 

1 15
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY

SPL # 1085.6 1085.6 1112.2 1112.2 1408.1
Couc 96 Std Dev Cone ‘5 Std Dev Cone
1 1
not
not

not

not not
not not
not not
not 1101.
not not
110! not

 

1 16
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Gadolinium Gadolinium Hafuium Hafnium Iron
SPL 6 103.3 103.3 482.3 482.3 099.3

Cone it Std Dev Cone ‘5 Std Dev Cone
not

 

l 17
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Gadolinium Gadolinium Hafnium Hafnium Iron
SPL # 03.3 03.3 482.3 482.3 099.3

Cone ‘5 Std Dev Cone % Std Dev Cone
< 1
<1
<

l.

§§§§§§§§

not
not
not
not
not
not

 

l 18
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Iron Nickel

1.6 .2 .2 10.8

‘5 Std Dev Cone ‘5 Std Dev Cone

not
not
not
110!
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not

iii

 

119
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY

Iron Nickel
1291.6 .2 .2 10.8
5 Std Dev Corn: 5 Std Dev Cone

not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not
not

5 § § 5 5 §§§§§§§§§§§§§ §§§§§§§§§§§§§§§§

 

120
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Rubidium Rubidium Scandium Scandium Selenium Selenium
SPL # 076.8 076.8 .3 .3
Cone 5 Std Dev Cone 5 Std Dev Cone 5 Std Dev

 

121
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Rubidium Rubidium Scandium Scandium Selenium Selenium
SPL 0 1076.8 1076.8 .3 .3 136.5

Cone 5 Std Dev Cone 5 Std Dev Cone 5 Std Dev
not

not not
not not
not 1'10!
not not
not not
not not

 

122
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Strontium Strontium Tantalum Tantalum Tantalum Tantalum
SPL I 14.0 14.0 1189.1 189.1 1221.5 1221.5
Couc 5 Std Dev Cone 5 Std Dev Conc 5 Std Dev

5 3333

not
not
not
not
not
not
not
not

 

123
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Strontium Tantalum Tantalum Tantalum Tantalum
14.0 189.1 189.1 1.5 1.5
5 Std Dev Conc 5 Std Dev Cone 5 Std Dev

1 not not
not not
not not
not
not
not

not
not
not
not
not
not
not
not
not

555555

5555555
555555

555555

 

124
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Terbium Terbium Terbium Terbium Thorium Thorium
.4 .4 178.0 178.0 12.0 12.0
Cone 5 Std Dev Cone 5 Std Dev Cone 5 Std Dev

not

 

125
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Terbium Terbium Terbium Terbium Thorium Thorium
.4 .4 1178.0 178.0 12.0 12.0
Conc 5 Std Dev Cone 5 Std Dev Com 5 Std Dev

not

not not
not not
not not
not not
not not
not not

 

126
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Thulium Tin Tin Zinc Zinc
3 1.7 1.7 15.6 115.6

5 Std Dev Cone 5 Std Dev Conc 5 Std Dev
not not not

ll)!

5555555555555555 a

5555555555555555555 5 5

not
1'10!
not
not
not
not
1101
not
not
not
1101
not
not
not
not
not
not
not
not
not

5555555555555555555 5

55

555555
5555555555
555555

 

127
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

COUNTED AFTER 5 WEEKS OF DECAY
Thulium Tin Tin Zinc
.3 1.7 1.7 15.6

5 Std Dev Conc 5 Std Dev Conc
<

5555555555
55555555555

5
9.

5

5

not
not
not
not
not

555

5555

555555 55
5 5 5 5 5 5555
5 5 5 5 5 5

5 5 5 5 5 5
5 5 5 5 5

 

128
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

UNTED AFTER 5 WEEKS OF DECAY
Zirconium Zirconium
7 7
Cone 5 Std Dev

 

129
APPENDIX A (cont'd)

Concentration and Standard Deviation Results (All Elements)

UNTED AFTER 5 WEEKS OF DECAY
Zirconium Zirconium
7 7

Conc 5 Std Dev
not

not
not
not
not
1101
not

 

APPENDIX B

130
APPENDIX B

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION (PPM)
Barium Barium Cerium Cerium Cesium Cesium
SPL # 123.7 123.7 145.5 145.5 .9 .9
Cone 5 Std Dev Conc 5 Std Dev Cone 5 Std Dev

 

131

APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION
Barium Barium Cerium Cerium Cesium Cesium
SPL # 7 7 45.5 45.5
Cone 5 Std Dev Cone 5 Std Dev Conc

not not
not not
not not
not not
not not
not not

 

132
APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION
Hafnium Hafnium Lanthanum Lanthanum
408.1 482.3 482.3 15 15
5 Std Dev Cone 5 Std Dev Conc 5 Std Dev

 

133
APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION
Hafnium Hafnium Lanthanum Lanthanum
1 15 15
5 Std Dev Cone 5 Std Dev Conc 5 Std Dev

not not
1101 not
not not
not not
not not
not not

 

134
APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION
Lutetium Lutetium Samarium Samarium Scandium
SPL # 03.2 03.2 .3
Cone 5 Std Dev Cone 5 Std Dev Conc

 

135
APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION (PPM)
Lutetium Lutetium Samarium Samarium Scandium Scandium
SPL S 03.2 03.2 .3 .3
Cone 5 Std Dev Cone 5 Std Dev Cone 5 Std Dev

1101 not
not not
not 110i
110! 1101
not not
not not

555555

 

136
APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION
Thorium Thorium Ytterbium Ytterbium
SPL O 12.0 12.0
Cone 5 Std Dev Conc 5 Std Dev

 

137
APPENDIX B (cont'd)

Concentration and Standard Deviation Results (11 Elements)

SAMPLE ELEMENT CONCENTRATION
Thorium Thorium Ytterbium Ytterbium
SPL # 12.0 12.0

Cone 5 Std Dev Conc 5 Std Dev
1

not not
not not
not not
not not
not not
not not

 

APPENDIX C

138
APPENDIX C

Concentration Values for 11 Elements, in Increasing Order

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm mm
Sample Cone. Sample Cone. Sample Cone. Sample Cone.
Number (ppm) Number (ppm) Number (ppm) Number (ppm)

46 35.8 5 620 39 6.92 53N2 64.1
40 120 2 625 40 7.02 48 64.5
45 149 53N2 625 46 7.36 51 66.5
39 156 11 633 36 7.89 29 66.8
36 198 53N1 639 35 7.93 12 66.9
37 219 51 648 34 8.05 2 67.2
38 225 26 654 37 8.49 55 67.4
28 228 56 654 38 9.78 31 68.1
47 229 27 660 33 10.6 21 70.2
35 248 1 665 60 25.3 9 71.4
43 249 23 678 45 28.3 59 73.4
60 267 30 679 30 31.8 6 73.6
33 326 57 705 43 35.8 1 73.8
34 402 12 715 47 40.9 57 75
13 415 59 726 3 41.7 56 76.7
42 451 9 727 25 44.5 28 78.1
14 464 54 730 14 46.9 54 78.2
41 470 44 736 26 47.3 27 79.3
3 528 24 741 41 51 16 81.6
48 528 6 748 42 51.5 8N2 82.9
25 542 29 750 58 54.1 50 84.8
15 548 55 751 11 55.1 7 85.9
18 570 22 758 18 55.1 5 87.9
16 571 32 770 17 58.1 8N1 92.6
17 575 10 833 13 58.5 32 97.7
21 580 52 836 22 58.6 4 102
31 588 7 861 10 60.1 15 106
58 595 8N2 881 23 61.7 44 106
50 603 SN] 913 53N1 62.7 52 123
4 619 24 64.1

 

 

 

 

 

 

 

 

 

 

 

 

 

139
APPENDIX C (cont'd)

Concentration Values for 11 Elements, in Increasing Order

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Caum i795 Kev; ‘ ‘ Europr'um Hm Kev;
Sample Cone. Sample Cone. Sample Cone. Sample Cone.
Number (ppm) Number (ppm Number (ppm) Number (ppm)

40 0.125 2 3.95 46 0.0501 29 1.01
35 0.127 55 4.15 33 0.108 48 1.02
33 0.183 50 4.16 35 0.115 59 1.02
36 0.191 12 4.26 39 0.121 10 1.03
34 0.193 7 4.32 36 0.125 17 1.03
37 0.242 5 4.47 34 0.131 21 1.06
38 0.256 52 4.65 40 0.134 56 1.06
39 0.348 13 4.81 38 0.147 6 1.07
46 1.08 21 4.89 37 0.156 24 1.07
60 1.8 9 4.96 47 0.182 18 1.09
11 1.83 8N2 4.97 60 0.219 15 1.1
45 2.08 3 4.98 43 0.286 50 1.1
47 2.24 10 5.06 45 0.333 31 1.11
43 2.28 16 5.1 30 0.491 55 1.11
4 2.37 58 5.18 3 0.605 52 1.14
25 2.45 54 5.42 58 0.757 1 1.15
41 2.65 51 5.43 11 0.767 23 1.15
42 2.74 26 5.5 41 0.775 51 1.17
31 2.91 8N1 5.73 26 0.799 7 1.19
27 3.02 53N2 5.76 13 0.826 4 1.2
48 3.13 57 5.8 14 0.863 32 1.2
24 3.26 6 5.98 2 0.898 27 1.22
30 3.44 53N1 6.03 22 0.902 8N2 1.23
18 3.45 28 6.16 44 0.925 16 1.26
1 3.47 29 6.84 28 0.933 54 1.26
59 3.61 56 6.89 25 0.944 5 1.33
15 3.71 32 7.02 53N2 0.97 8N1 1.33
23 3.74 22 7.7 57 0.971 12 1.33
14 3.75 44 13.4 53N1 0.992 9 1.38
17 3.77 42 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

140
APPENDIX C (cont'd)

Concentration Values for 11 Elements, in Increasing Order

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

um e Lanthanum iSIS KeWi

Sample Cone. Sample Cone. Sample Cone. Sample Cone.

Number (ppm) Number (ppm) Number (ppm) Number (ppm)
46 0.535 58 5.31 46 2.78 29 32.7
33 0.695 31 5.42 40 4.73 51 32.7
35 0.919 22 5.43 34 5.01 59 33.3
36 0.943 51 5.57 36 5.02 1 33.6
38 0.994 57 5.58 37 5.35 55 33.6
39 1.02 4 5.63 39 5.84 18 33.7
34 1.18 8N2 5.7 38 5.88 31 33.9
37 1.26 3 5.79 35 7.33 24 34.5
40 1.26 10 5.79 47 7.64 6 34.7
47 1.36 15 5.9 33 8.78 57 35.2
60 1.48 1 5.93 60 9.42 50 35.5
45 1.52 55 5.93 43 12.7 8N2 35.6
43 2 25 5.99 45 12.8 12 35.8
30 2.9 53111 6.08 30 17.2 32 36
41 3.53 32 6.13 3 19.1 48 36.3
14 4.25 7 6.27 41 22.5 7 36.4
18 4.51 8N1 6.27 14 23.3 56 36.5
44 4.58 54 6.35 26 23.6 54 36.7
50 4.62 56 6.4 25 25.2 5 36.8
42 4.65 5 6.46 58 25.3 9 36.8
23 4.68 9 6.53 42 26 23 36.9
29 4.7 16 6.64 11 26.8 8N1 37
24 4.73 11 6.65 2 30.2 16 37.6
6 4.77 12 6.77 10 30.9 21 37.6
17 4.89 59 6.81 17 31.9 27 38.7
26 4.91 53N2 7 22 31.9 15 44.3
48 4.95 27 7.52 13 32 4 44.9
13 5 2 7.58 53N1 32.2 44 45.9
21 5.21 28 7.98 53N2 32.2 52 61
52 5.29 28 32.7

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Concentration Values for 11 Elements, in Increasing Order

141

APPENDIX C (cont'd)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

utetrurn e

Sample Cone. Sample Cone.
Number (ppm) Number (ppm)
46 0.0324 48 0.308
39 0.0334 23 0.312
35 0.036 18 0.314
34 0.04 59 0.315
38 0.0462 50 0.318
0.0481 15 0.319

33 0.0696 57 0.324

60 0.078 51 0.33
37 0.083 1 0.331
47 0.0918 21 0.333
36 0.105 53N2 0.336
30 0.132 53N1 0.339
43 0.148 6 0.342
45 0.154 10 0.347
58 0.231 56 0.35
14 0.237 8N2 0.352
41 0.245 54 0.355
26 0.254 52 0.356
11 0.257 27 0.358
3 0.261 22 0.361
25 0.267 28 0.363
13 0.269 8N1 0.364
24 0.278 4 0.395
17 0.29 12 0.396
2 0.291 9 0.397
42 0.295 16 0.397
31 0.297 7 0.399
55 0.299 32 0.445
29 0.301 5 0.457

44 0.301

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

armrrurn e ,
Sample Cone. Sample Cone.
Number (ppm Number (ppm)
46 0.355 21 5.22
39 0.517 10 5.23
35 0.545 13 5.26
33 0.582 18 5.26
34 0.609 29 5.28
36 0.611 28 5.3
40 0.635 31 5.36
37 0.733 15 5.38
38 0.751 55 5.38
47 1.1 6 5.39
60 1.46 23 5.42
43 1.82 50 5.42
45 1.93 24 5.46
30 2.16 56 5.48
3 3.14 51 5.49
11 3.65 1 5.67
58 3.65 54 5.7
41 3.81 8N2 5.82
26 3.93 32 5.88
14 3.95 8N1 6.07
25 4.25 44 6.12
2 4.37 27 6.33
42 4.65 7 6.4
53N2 4.65 16 6.42
22 4.78 9 6.44
53N1 4.79 12 6.52
57 4.82 4 6.97
48 5.03 5 7.69
17 5.1 52 7.99
59 5.15

 

 

 

 

 

 

 

 

142
APPENDIX C (cont'd)

Concentration Values for 11 Elements, in Increasing Order

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sal-Iaium 1889 KEV) N onum e
Sample Cone. Sample Cone. Sample Conc. Sample Cone.
umber (PPIII) Number (ppm) Number (ppm) Number (ppm)
' 35 0.258 4 9.68 35 0.623 10 10.1

33 0.261 1 9.76 39 0.68 21 10.1
39 0.307 23 9.95 37 0.691 55 10.2
37 0.36 13 10.2 36 0.732 8N2 10.7
40 0.381 5 10.4 40 0.762 6 10.9
34 0.4% 21 10.5 34 0.78 29 1 1
36 0.4% 28 10.6 38 1.04 9 1 1.7
38 0.477 51 10.7 46 I.“ 50 12
46 0.914 50 10.9 33 1.08 8N1 12.4
47 3.04 52 10.9 45 2.83 27 12.9
45 3.22 27 11.1 47 3.38 12 13
60 3.6 2 11.2 60 3.95 22 13.1
43 4.34 10 1 1.3 43 4.94 57 13.2
30 4.59 15 1 1.7 30 5.32 2 13.3
25 6.74 16 11.8 41 6.35 1 1 13.3
41 6.83 8N2 11.9 42 6.7 13 13.5
42 7.49 7 12 14 7.16 7 13.6
14 7.78 6 12.1 26 7.63 59 13.9
58 7.96 12 12.3 48 7.75 32 14.2
31 8.56 56 12.7 25 7.98 44 14.4
26 8.59 22 13 31 8.02 16 14.5
18 8.67 8N1 13.2 58 8.13 53Nl 15.5
59 8.75 53N1 13.4 17 8.2 53N2 15.5
17 8.83 9 13.5 18 8.2 4 15.7
48 8.94 53N2 13.6 3 8.63 28 15.8
55 9.12 57 13.8 23 8.7 54 15.8
3 9.45 54 14 24 9.09 5 16.7
29 9.52 32 16.4 51 9.75 56 16.7
1 1 9.53 44 20.1 1 9.99 52 18.9
24 9.65 15 10

 

 

143

APPENDIX C (cont'd)

Concentration Values for 11 Elements, in Increasing Order

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

tte turn ,
Sample Cone. Sample Cone.
Number (ppm) Number (ppm)

46 0.163 6 2.36
39 0.247 10 2.37
35 0.271 59 2.37
40 0.282 1 2.38
34 0.31 48 2.39
38 0.34 55 2.4
37 0.42 8N1 2.41
47 0.48 21 2.41
33 0.499 22 2.41
60 0.738 53N2 2.41
43 0.757 56 2.42
36 0.766 51 2.44
45 0.822 29 2.46
30 0.984 50 2.46
41. 1.61 8N2 2.51
58 1.67 15 2.54
11 1.7 57 2.54
14 1.87 28 2.6
25 1.94 54 2.61
26 1.95 53N1 2.7
44 2.02 32 2.71
42 2.03 9 2.74
13 2.04 12 2.82
24 2.12 27 2.83
23 2.16 16 2.9
3 2.21 52 2.93
17 2.22 4 3
2 2.25 7 3.21
18 2.3 5 3.5
31 2.31

 

 

APPENDIX D

144
APPENDIX D

 

Barium (123.7 KeV) Deleted
I I
I I
I I
I
I
I
I
I
I
I
I
I
I
I
I
II
II
II
I
9-I
II
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
l
I
I
I
I
l
I
I
I
I
I
I
I
I
I
I
I
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.-
__ E

O O D, 1. 6 R ‘
mmumumunummumnmmuuumnnoummmmumu

811
I)

II.
ms
I‘M
I“
It!
823
a

m
81!
an
an
I!

mmmumumuttnasm
u. a.

Centroid Linkage Dendrograms Using 10 of 11 Elements

.1.
m
m
m
D
m
m

145

APPENDIX D (cont'd)

 

- -2 _ .................................................................................
.uwmeum-uta — . _

Cert. (145.5 KeV) Deleted
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1

_ ‘7 ’5 8843 3 6 I I 0.8 7.. 0 6
m.m.m.mm............m.am..mn.»...uums.mamauaamaaam.mmms.m..

Centroid Linkage Dendrograms Using 10 of 11 Elements

146

APPENDIX D (cont'd)

Cesium (795.9 KeV) Deleted

 

Centroid Linkage Dendrograms Using 10 of 11 Elements

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

d ..... _
_ __ ____-.__ __

‘ T 3 : ‘Oa‘ at, D. .1 ‘ '
gammamummum-mumumuuuunnuumunummmnummnnmmmmmmmmnmmu-mnmm

 

 

 

 

 

 

 

 

ummeramurmwunmm

147

APPENDIX D (cont'd)

Europlum (1408.1 KeV) Deleted

 

 

 

 

 

 

 

 

 

 

Centroid Linkage Dendrograms Using 10 of 11 Elements

 

 

 

 

mmmmmamnum

 

__ a»:

 

 

8 1
numuunun-m

 

 

 

 

 

 

 

 

 

 

1 G 3 41
uunmmaumummn

 

 

 

 

 

 

 

 

 

 

I
..

 

 

 

 

 

 

m
m
m
m
u
m
m

148

APPENDIX D (cont'd)

 

Hamlum (482.3 KeV) Deleted

I
0'- I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
II I
“II I
II I
9—II I
II I
I I
I I
I I
I I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

’ D .1 O 1 0 an t : Cal 21 s, 1‘
womanmumnumunnmmnmnnuuumumnumm—uuammmnmnnumuamuumuummmmmmmm

m
......... ...--.- M
m
m
n
m
m

Centroid Linkage Dendrograms Using 10 of 11 Elements

149
APPENDIX D (cont'd)

11111111

 

Lanthanum (815 KeV) Deleted
'T
I
I
i
I
i
E
i
I
I
I
I
I
I
i
i
l
I
I
I
I
I
:-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O .7 O: T o D a o 1
mnunmumunmmmoummamnunun—mummmummuommnnmmmu «magnum

Centroid Linkage Dendrograms Using 10 of 11 Elements

D!“ “C 18 marma- urea-n. me In!“ turn

ISII

APPENDIX D (cont‘d)

 

 

L:

kmuuhl
__.r_:

 

 

 

 

 

 

Ianeflunr(2085ilhwn Defined

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m

m

---_- -.--wawwwwwmmwwwwmwwumuflwna m
tit—t _ t—tetct _ .. m
m

m

s O . T o T O a ) r C 1 1a 3 i 1 O o
u.m...m..m”sameness..m...u.a..s..nmm.am..m.u.m..a...mmam.mm

Centroid Linkage Dendrograms Using 10 of 11 Elements

151

APPENDIX D (cont'd)

Samarium (103.2 KeV) Deleted

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..... _H-- _
no} _a_

 

 

 

 

 

s .‘T . .3 1. 0", 3.1.... ,,
mmuummmmmunnmmnmnuanceuuummuummmunumnommmununmuumm ::ummm

Centroid Linkage Dendrograms Using 10 of 11 Elements

arm ”16 I. mama- arm. me In!" turn

152

APPENDIX D (cont'd)

 

O

.- “WU
_ _ .3 _

 

 

 

 

 

 

 

 

 

 

Scandium (889.3 KeV) Deleted
I
I
I
I
I
I
I
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

., 3 O s 7. s, O s ‘11 ,3 7 ‘7
m....mmauu.smammasam..msma.mmaammnma.asmanauau.amumaammm

Centroid Linkage Dendrograms Using 10 of 11 Elements

 

 

 

mam-stunt: IIMDIM moms-nun

153

APPENDIX D (cont'd)

Centroid Linkage Dendrograms Using 10 of 11 Elements

Thorium (312.0 KeV) Deleted

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

euttttet
:i. __

 

 

 

 

 

 

 

s o 1 I. a ‘ 5’ .‘T 3 3
m...m.am.aum.mams...ua"an.a.m.aam.mmaammu

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a m m m m m m m m m m m m m

 

 

m
m
m
m
m
m
m

154
APPENDIX D (cont'd)

Centroid Linkage Dendrograms Using 10 of 11 Elements

Ytterbium (396.5 KeV) Deleted

.-
II
II
II
II
II
O-I
I
9..
i
Q-
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mumnmmmmttmasm
an an

 

II
m

111111111111111111111111111111111111111111111111111111111111111111111111111111111

 

’
m

 

 

 

 

 

 

 

 

1111111111

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cram-m mare ll mama: arm. we in!" m

APPENDIX E

155

 

 

00.0 00.0 00.0 00.01 00080 . 00.0 000

00.0 00.0 00.01 00.01 0080 . 00.0 000

00.0 00.0 00.0 00.0.. .380 . 00.0 000

00.0 00.01 00.01 00.01 :3 _ 00.0 000

00.0 00.0 00.01 00.01 0980 . 00.0 000

00.0 00.0 00.0 00.0 0080 . 00.0 000

00.0 00.01 00.01 00.01 0.80 . 00.0 000

00.0 00.0 00.01 00.01 $080 _ 00.0 000

00.0 00.01 00.01 00.01 380 . 00.0 000

00.0 00.0 00.0 00.0 (.080 _ 00.0 00.0

00.0 00.0 00.0 00.0 g _ 00.0 000
408.0. g; It! £000! 333 _ 8.2.0.009 g

800000000 0.03

0 “as: E0900

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 03000

. 00.0 03000

_ 00.0 000

52.8. .8555 2.886 .o .852 oz

 

_ 00.0 000

_ 00.0 000

_ 00.0 000

_ 00.0 000

_ 00.0 000

_ 00.0 000

. 00.0 000

_ 00.0 000

. 00.0 000.

. 00.0 000

_ 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

_ 00.0 000

_ 00.0 000

_ 00.0 000

. 00.0 000

_ 00.0 000

_ 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

. 00.0 000

_ 00.0 000

. 00.0 000

00.0 00.0 00.0 00.0 00080 _ 00.0 000

00.0 00.0 00.0 00.0 8000 . 00.0 00

00.0 00.0 00.0 00.0 ‘30 . 00.0 0300

00.0 00.0 00.01 00.01 308 . 00.0 0800

00.0 00.0 00.0 00.0 g _ 00.0 00

00.0 00.0 00.0 00.0 880 . 00.0 00

00.0 00.0 00.0 00.01 g . 00.0 00

00.0 00.0 00.0 00.0 0880 . 00.0 00

00.0 00.01 00.01 00.01 00080 _ 00.0 00

00.0 00.0 00.0 00.0 (.080 . 00.0 00

00.0 00.0 00.0 00.0 g _ 00.0 00
$8.00 gus- i 00000.03 30; . 092.0000 .05

0000000500 g

0 «gig
000.0 000.000 00 000.0 0 000.0 E
000.0 000.000 00 000.0 0 000.00 008.0
000.0 000.000 00 000.0 0 000.0 .3080
000.0 000.000 00 000.0 0 000.0 30
000.0 000.000 00 000.0 0 000.00 3
000.0 000.000 00 000.0 0 000.0 3
000.0 000.000 00 000.0 0 000.0 >80
000.0 000.000 00 000.0 0 000.0 3
000.0 000.000 00 000.0 0 000.0 0.080
000.0 000.000 00 000.0 0 000.0 108.0
000.0 000.00 00 000.0 0 000.0 g

800 00.0.5010 8 00 30000.0. 8 00 El 520;

go 0 I: 000.000.01.00 0050.50

2:255 00 ”En: 9.82:0 335.0—

m— N0nzmmm<

858% £336 .c .35.: oz

156

 

 

 

_ 00.0 02000
_ 00.0 02000
. 00.0 000
. 00.0 000
_ 00.0 000
. 00.0 000
_ 00.0 000
. 00.0 000
_ 00.0 000
_ 00.0 000
_ 00.0 000
. 00.0 000
_ 00.0 000
. 00.0 000
. 00.0 000
_ 00.0 000
_ 00.0 000
. 00.0 000
_ 00.0 000
. 00.0 000
00.0 00.0 00.0 00.0 00.080 _ . 00.0 000
00.0 00.0 00.0 00.0 0080 _ _ 00.0 000
00.0 00.0 00.0 00.0 kg _ . 00.0 000
00.0 00.01 00.01 00.01 E _ _ 00.0 000
00.0 00.0 00.0 00.0 03 . _ 00.0 000
00.0 00.0 00.0 00.0 g _ . 00.0 000
00.0 00.01 00.01 00.01 .080 _ 00.0 000 _ 00.0 000
00.0 00.0 00.0 00.0 £003 . 00.0 000 _ 00.0 000
00.0 00.01 00.01 00.01 380 . 00.0 000 00.0 00.0 00.0 00.0 5.80 _ 00.0 000
00.0 00.0 00.0 00.0 580 . 00.0 000 00.0 00.0 00.0 00.0 0080 . 00.0 00
00.0 00.0 00.0 00.0 g . 00.0 000 00.0 00.0 00.0 00.0 I80 . 00.0 0800
.E. g 2‘! 2300.00! 0.03; . 00.2.0000 0040 00.0 00.0 00.0 00.01 g . 00.0 0200
0000000500 015 00.0 00.0 00.0 00.0 0093 . 00.0 00
0 .35: 100030 00.0 00.0 00.0 00.0 g . 00.0 00
00.0 00.0 00.0 00.0 g _ 00.0 00
00.0 00.0 00.01 00.01 00080 . 00.0 00.0 00.0 00.0 0003 _ 00.0 00
00.0 00.01 00.01 00.01 0080 . 00.0 00.01 00.01 00.01 380 . 00.0 00
00.0 00.0 00.0... 00.01 .0080 . 00.0 000 00.0 00.0 00.0 00.0 580 . 00.0 00
00.0 00.01 00.01 00.01 0080 _ 00.0 000 00 0 00.0 00.0 00.0 g . 00.0 00
00.0 00.0 00.01 00.01 0080 _ 00 0 00 >8 .00 g; :8. 80000300- 30; . 00.2.0000 03
00.0 00.0 00.0 00.0 0080 _ 00 0 000 000.000.0400 as
00.0 00.01 00.01 00.01 g _ 00 0 000 0 50858 5.0.300
00.0 00.01 00.91 00.01 0.080 _ 00 0 000
00.0 00.01 00.01 00.01 380 _ 00 0 000 000.0 000.000 00 000.0 0 000.00 00.0.80
00.0 00.0 00.0 00.0 580 _ 00 0 000 000.0 000.000 00 000.0 0 000.00 00000
00.0 00.0 00 0 00.0 g _ 00 0 000 000.0 000.000 00 000.0 0 000.0 fig
58. g a 2030! 0.03; _ 00850000 840 000.0 000.000 00 000.0 0 000.0 g
000.000.00.00 003 000.0 000.000 00 000.0 0 000.00 03
0 .30.! 00.00030 000.0 000.000 00 000.0 0 000.0 g0
000.0 000.000 00 000.0 0 000.0 g
. 00.0 000 000.0 000.000 00 000.0 0 000.0 g
_ 00.0 000 000.0 000.000 00 000.0 0 000.0 380
_ 00.0 000 000.0 000.000 00 000.0 0 000.0 (.080
_ 00.0 000 000.0 000.00 00 000.0 0 000.0 ‘30
_ 00.0 000 50 000001.. 8 00 30000.0. .8 00 808' 53g
_ 00.0 000 3.00300 0 5.- 0000000500 #05050

Ace-8. 358.5 53.6 n .8555 :56 a

5:25..."— 00 055 9.83.0 2.35.0—

3.253 m— 00092009000

00000000000

 

O O I I O O I O I I I
OOOOOOOOOOO
.

 

157

I I I I O I O O I I I
00900000006
I

00.0 00.0
00.0: 00.0!
00.0: 00.
00.01 00.01
00.0 00.0
00.0 00.0
00.0: 00.0I
00.0: 00.0!
00.0: 00.0:
00 0 00.0
00 0 00.0
:510x48 :10!
0000000090
00.0 00.0
00.0 00.0
00.0 00 0
00.0: 00 0|
00.0 00 0
00.0 00.0
00.0: 00.01
00.0 00.0
00.01 00.0I
00.0 00.0
00.0 00.0
IDIHX¢8 :40!
0000000080
00.0 00.0!
00.01 00.0I
00.0 00.0
00.0: 00.0I
00.01 00.0I
00.0 00.0
00.0! 00.0!
00.01 00.0!
00.0- 00.00
00.0 00.0
00.0 00.0
20:00:! I‘ll
0000000090

 

 

_ 00.0 02000
_ 00.0 02000
. 00.0 ~00
_ 0o.o 000
. 0o.0 000
00000 . _ 00.0 000
00000 . _ 00.0 000
has . . 00.0 000
30000 _ _ 00.0 000
00000 . _ «0.0 000
00000 _ _ 0o.o 000
g . _ 3.0 000
$0000 . _ 00.0 000
0.009 _ _ 36 000
580 . _ 00.0 000
(anxu0 . oo.o 000 _ 00.0 000
a; . 09.50000 '00 _ 00.0 000
000010: _ 00.0 000
0 .000150 0000000 _ 0o.0 000
_ no.0 000
:59 _ _ 00.0 .00
0080 . . 00.0 000
hxuna _ _ 00.0 000
BBQ— _ . 36 000
00000 _ . 00.0 000
0080 . . 00.0 000
080 _ 00.0 000 _ 00.0 000
g _ 00.0 000 . 00.0 000
380 _ 00.0 000 00 o 00.0 00.0 on 0 =53 _ 00.0 000
503 . 36 000 3 o 30 00.0 0- o 0008 . 3.0 on
g _ 006 000 0o o 00.0 00.0 00.0 0.83 _ 00.0 «:00
0.0040; _ 025.000 g 00 o 00.0 00.0 00.0.. EQ— . 00.0 0200
g 00 o 00.0 00.o 00 o 0080 _ 00.0 00
0 .zunxsu 0000000 o0 o no.0 00.0 00 0 00000 _ 00.0 00
00 o 00.o 00.0 00.o HRH. _ 00.0 00
0009 . oo o 00.0 ~0.o on... g . 00.0 00
00000 . 0o o 00.on 00.on 00.0: 00200 _ 00.0 00
0:80 . 00 o 00.0 00.0 00.0 <83 _ 00.0 00
0g _ 00.0 own 00 o 00.0 00.0 00.0 g _ 00.0 00
00000 _ 00.0 000 .>un.00 20:00:: :00! 801000! 0000000, _ 00000000 0040
00000 _ 00.0 .00 0000000000 0:00:08
r000 _ 00.0 000 0 .xunzax 0000000
80000 _ 00.o 000
00000 . 00.o 000 ooo o 000.000 00 000.o 0 0o0.o0 00000
580 . no.0 000 ooo 0 000.000 00 000.0 0 000.00 0080
(g . 006 000 ooo 0 000.000 00 000.0 0 000.0 .089
g . 3000 g 80 0 000.000 00 000.0 0 000.0 0800
g ooo 0 000.000 00 000.0 0 000.00 00000
0 .nunzal 0000000 ooo o 000.o0~ 00 000.0 0 000.0 00000
ooo 0 000.000 00 000.0 0 000.0 g
. 0o.o 000 ooo o 000.000 00 000.o 0 000.0 00000
. «0.0 000 ooo 0 000.900 00 000.o 0 000.0 00000
_ 0o.o 000 ooo 0 000.000 00 000.0 0 000.0 13000
_ 0o.o 000 ooo o «00.00 00 ooo.o 0 000.0 ¢uuoa
_ 00.0 000 g 00.0.4.0 00 00 200.00. B 00 E 00040;
_ 00.0 000 go 0 a 000.000.0500 005030
$.28. 358.5 5.86 v 358.5 5.36 ..

3:25.”.— 00 05a: 9.82:0 2.35.00

3.0.88 0.— anzmmm<

no.“

no."

 

580 .

 

 

 

158

 

hon hon 00.N g _ 00.0 C.»
gut! a gnu! flag; _ 0025—.an g
nor—5H2: g
n “g Eng
no.0 no.0 no.0 =3 .
00.0.. v0.0l v0.0- 03 .
anon 0N0. 0N0.- § .
onét 0n.HI 0n.un as _
no.0 no.0 no.0 n08.— _
00.0 00.0 00.0 g _
00.0.. 00.0. 00.0.. g .
no.0: no.0. «v.0: luau _
00.? no.7 00.“: 380 _
00.0 00.0 vv.0 580 _
0.04 an; on." g _ 00.0 000
g; a gad- gg _ 8.2::- g
«0an32: nag
v "595! fig
230 00.0 nv.0 =3 _
00.0 00.0 ov.o 03 _
00.0 nN.0 nu.0 bg _
«not 0.0.0.. v0.01 :3 _
v0.0 0n.0 0N0 mg _
do; on; 00$ 88» .
«0.0.. n~.0I «no: 3 _ 00 0 00¢
nn.0 NN.o v0.0 .303 . n« o 03
«0.0! no.0! 3.7. 380 . z o ova
0N." 00..— 00.0 5000 _ 00 0 n3
n0.N vv.~ ban (and . on 0 on»
g; a Ea 53g _ gubnd 50
”Gauge g
n .5952 gag—U
no.0 3.? «Non F003 .
~n.o.. no.0. 00.0: 03 .
00.0 no.0 010: fig _
«0.0I 00.0: 00.0- 93 _ N10 09»
00.0- 00.01 00.0I 008.— . nu o Gnu
n04 no.0 v0.0 3 _ 00 0 on»
«“0: v0.00 no.0- r03 _ 00 0 0n“
N10. "N20! 0N0- g _ On 0 onn
00.0- out—I 00.": 380 _ an 0 an»
3.0 00.0 00.0 58.— . no 0 on»
00.~ vn.~ 00.N g . nn 0 nnn
a: li gun! as . game '40
swank‘h. ”as
N ”SR! Ema
. no.0 on.
_ ~10 own
. 00.0 0mm
_ 00.0 own
_ v0.0 nan

3.2.30 90592—0 flail—U m

35.5—m 3 win: 9.32:0 2:35.0—

23233 H Nmnzmmm<

no.“ "N
a o n a o o a o o o o
N~°°°~°°°~~

. o I I u I a o a I o
Nd°°°fl°°°flfl

no nhn =2:

 

«00.no—
000.0nn
oo~.00u
ovo.~hu
Ono..."
«on.nhn
oom.n0u
ono.noa
«o—.ovn
ooo.vnn
0Nh.hv

Onh‘ljh

_ 00.0 v...»

_ 00.0 «Inn»

_ 00.0 nannn

_ n10 Nun

_ 00.0 3n

. v0.0 own

_ 00.0 ova

_ n70 «v»

. 00.0 no.»

_ «To «nn

_ 00.0 Ann

. 00.0 «Nu

_ 070 0N0

_ 00.0 0N0

. N10 can

_ nu.0 nun

_ 00.0 03

. no.0 nun

. 00.0 «n»

_ no.0 «an

. no.0 can

_ 00.0 bun

_ 00.0 wan

. 00.0 on»

. n70 Gun

_ 00.0 nun

_ 00.0 Nun

. 010 «an

E _ no.0 own

880 . 00.0 an

.80 _ 00.0 «Son

2 _ 00.0 «800

g _ 00.0 ha

893 . 00.0 on

g _ .30 an

.88— _ n10 on

380 _ 010 nu

580 _ 00.0 N»

g . 00.0 «a
a; _ £50 5

a;

a .g 5
95.0.0 v 0:.0u lg
anvo v 2.0.nu g
vnno C 000.0 g
non.o v 000.0 g
000." v «00.8" g
«3.0 v v9.0 g
0010 v «010 g
«on.0 v 03.0 g
03.0 v can.» 380
nevo v 00n.n 58—
000.0 v 025 3

an rumba. 8 an E a;

358.5 9.336 m

 

159

00.0 on; On.“ on." E _
00.0 on.” on." on.“ g _
00.0 3.0 00.0 00.0 g _
00.0 no.0- no.0: no.0- g _
00.0 nu." nu.“ nu." g .
00.0 no.“ no.~ no.« 30 .
00.0 «no «no «no 3 .
00.0 00.0 00.0 00.0 g .
00.0 «0.0. «0.0- No.0- 3 .

6.38. 858%. 9.336 a

8:25:— 2 "in: 283.0 2.35.0—

8.238 m— N—szmm<

 

' C Q D O Q 0 . I C .
OOOOOQOOOOO
C

 

160

 

I U I I C O C O I D .
OOOOOOOOOOO
0

no N no N
talnxgl 3":
nunhnnhtfiu
no.0 no.0
v0.0! 00.0!
0N.ol hN.0I
0n.«l 0n.~l
no.0 no.0
00.0 00.0
00.0I 00.0.
no.0! no.0:
00.”: 00.nl
v0.0 vv.o
oo.n no.“
lalnxtl I‘m:
QUMHlotfin
00.0 no.0
v0.0 no.0
0n.0 0N.0
00.0: «0.00
on.o Nn.0
no.~ 0o.—
oo.oI Cu.0l
0N.0 no.0
~0.0| no.0I
nu.a N0."
nv.N vn.N
Insult! I402
nu~an~9¢fiu
no.0 un.0l
Nn.ol ov.0I
0~.0 «0.0
«0.00 00.0I
00.0! 00.0:
no.“ No.0
No.0: vv.01
No.0: uN.0|
00.0- 0N.nu
00.0 00.0
00.N vn.N
talHX¢l 8(fll
lunhu~h¢fin

 

 

509 . _ 00.0 com
4009 . 00.0 :0 . 00.0 nanon
53g _ 2930 g . 00.0 «anon
nag _ 3.0 «on
o annals: uflnnaau . 00.0 «on
. v0.0 0on
19000 . . 00.0 000
00000 . . no.0 «on
has . _ 2.0 so
00000 _ . No.0 “no
00000 . _ 00.0 "no
00000 . _ 00.0 own
3 . _ 3.0 can
.309 _ _ 00.0 an
303 _ _ No.0 own
509 . _ no.0 o~n
(003 . 00.0 03 . 00.0 03
a; . gamma .5 _ oo.0 93
g _ no.0 NNn
v naunzsx auhnaau _ no.0 «Nn
_ 00.0 00o
30.09 _ _ 00.0 Sn
g _ _ 00.0 on»
0:03 _ _ 00.0 ooo
00000 . _ no.0 3»
303 . . 00.0 nun
0008 . _ 0.0.0 «on
g _ _ 3.0 .3»
8003 _ so 0 00» No.0 a.“ 3.0 00.0 2.080 _ o0.0 So
380 . on 0 03 no.0 3.— No.0 no.0 0003 _ 00.0 on
509 . 0“ 0 ova 0.0.0 00.0 230 oo.0 .083 _ 00.0 «zoo
‘0000 . 00 0 nvn 00.0 00.0 «0.0 nu.0u :HRHH . 00.0 "Inn
53; . gunn— ln‘u 3.0 00.0 3.0 0N6 808 . no.0 an
on: 00.0 o0.N .0." no." 0008 _ 00.0 on
n “g 53.8 00.0 vo.0 3.0 3.0 g _ 2.0 on
00.0 00.0 «0.0 3.0 ‘03 . no.0 vn
:83 . no.0 vn.0n 3.0. 3.0- 303 _ 3.0 no
0.0.03 . 00.0 2.." No.0 0N4 580 _ 00.0 No
0:03 _ 00 0 on.“ 00.N on u (003 _ 00.0 on
089 . No .0 03 $8 .90 gui- 3 gun! in; _ 250 020
nuuoa _ no.0 on» nonhnuhtfin quanta:
0000a . 00.0 non a .aunxox xuhnoao
r000 . 00.0 0n»
26000 . 90.0 on» 000.0 ooo...“ no «np.0 o ova on 09000
00000 . «0.0 onn 000 0 n0n.~«v no oov.0 o 000.ou un0Q0
503 _ 00.0 vnn 000 0 0.0.0.: no 00n.0 o 3N6 0:09
(0000 _ no.0 nnn 000 0 00o.v0~ no "no.0 o ooh.0 00000
305; _ 20.30 g 000 0 gonna no 30 o o 000.2 309
930:! 000 0 «00.02 no So 0 o 09.0 309
a ”g Ema 000 0 03.03 no onv 0 o Noni g
000 0 «no.00N no con 0 o 0ov.> xnuoa
. o0 0 non 000 0 ooo.0oo no «no 0 o «on.o 00000
_ NH 0 con 000 0 0342 no a: 0 o ovn.o 580
. 00 0 won 000 0 000.0o no no. 0 o oun.n ¢0000
. no 0 won 810 80.570 8 no .8299 .8 on g 53;
_ v0.0 oon nag 0 I: nonhnnikn tin
€2.80 30.2.0 5.36 o 358...». p.336 a

8:255 2 ”5.5 282:0 2:65.0—

33008 H 02973730.

 

161

 

 

00.0 nn.0 220 no.0 F83 .
00.0 3.0 3.0 00.0 808 _
00.0 3.0 3.0 00.0 has _
00.0 nn.0- nn.0: nn.0: g .
00.0 vo.0 3:0 0o.0 88o _
00.0 0o.u ooz— 0o.o g _
00.0 3.0: «0.0- «0.0- .69— .
00.0 nn.0 nn.0 nn.0 g _
00.0 00.0: 00.0: 00.0: 308 _
00.0 on.“ 004 on.” <58 .
00.0 n0.N no.0 n0.~ g _ 00.0 onn
$8.: 2333 25! £25- gut) . gamma ”3
mansions: mg
0 “55: ago
00.0 on; on; 0H." 5.03 _
00.0 on.» on." on." g .
00.0 3.0 3.0 00.0 g .
00.0 n0.0n 36: no.0- 303 .
00.0 no:— noé no." 88o _
00.0 n0.~ n0.~ no.0 608 _
00.0 nn.0 3.0 3.0 .63 _
00.0 00.0 00.0 2.0 .88.— _
00.0 no.0.- No.01 no.0: .38.— .
$2.8. 358..» 9.386 a

2:25:— 2 ”Eu: 232:0 2.35.:

3.308 M Nuazmmm<

 

 

vai—

 

 

 

162

 

00.N g . 00.0 000
00.0-Hg a ‘53:! Hug; . 2.6000 "‘0
”garage a;
0 «g E0040
00.0 00." 00.0 than .
00.0 00.." 00.0 03 .
00.0 00.0 00.0 E _
00.0.. 00.0: 00.0! g _
00.0 00.0 an." 0980 _
00.“ 00.0 00.N g .
«0.0 nn.0 «0.0 g .
06.0 06.0 06.0 00080 .
«0.0: «0.0I «0.0: 380 _
00.0 00..— 004 580 .
60.N 60.N 60.N g . 00.0 000
g; a 0325:! a; . 2.6000 g
000.606.0560 0%
0 ”a; E030
00.0 00.0 00.0 .6609 _
00.0.. 00.0! 00.0.. 93 _
6~.0I 6N0: 6~.0.. fig _
00.7 00.7 00.0: 0g .
00.0 00.0 00.0 0080 _
60.0 60.0 60.0 880 .
06.0.. 06.0I 06.0: g .
00.0: 00.01 00.0I g .
00.0.. 00.7 00.0: 38.— .
00.0 00.0 00.0 580 .
00.0 00.0 00.0 380 . 00.0 000
g; It! 0500:“! flag; _ 2.6000 g
30.60:.“ 96g
0 “g E098
00.0 60.0 00.0 .6800 _
00.0 00.0 00.0 03 _
00.0 00.0 00.0 .0180 _
00.0.. «0.? 060' pg _
00.0 «0.0 0.0.0 mug _
«0.0 00.~ 004 Hug .
00.0: 0a.? «0.0! g _
00.0 04.0 00.0 20.80 _ 600 000
«0.0! 00.0: 3.0.. 300.— . 000 600
an." «0.0 00.0 580 . 000 000
00.0 on." 60.0 (g _ 000 000
g; a glau- agg . 00.2.6000 g
000.600.61.00 g
0 .g E0940
00.0 «0.? nn.0l cg _
«0.0: 00.0: 00.0- 03 _
00.0 «0.0 0.6.00 .03 .
«0.0I 00.? 60.0: as . «0.0 000
00.00 06.? 00.00 0800 _ nn.0 000

€28. 358% 5.88 6

 

 

00.0 004 «0.0 g _ 00.0 06.0
3.0 2.0- 3.0- g _ 00.0 63
00.0 2.0- 3.0: g _ 3.0 an»
60.0 00.0.. 06.7 363 _ 3.0 03
00.0 00.0 66.0 503 _ 00.0 on»
3.0 00.~ 0.0.0 . g _ 00.0 000
.E. gus- § glut an; _ 80250 0040
000.600.6560 nag
N “ES... 5.6030
_ 00.0 000
_ v0.0 600
_ 00.0 000
_ 00.0 000
_ 00.0 Zn
_ 00.0 «.33
. 00.0 2.0...»
_ 2.0 «an
_ 00.0 3»
_ 00.0 0...»
_ 00.0 «00
. 00.0 000
_ 00.0 0N0
. 60.0 6N0
_ 00.0 «N»
_ 00.0 «N»
_ 00.0 :0
_ 00.0 000
_ 00.0 2»
00.0 00." S.” 00.0 .688 _ 00.0 a:
00.0 3." 00.0 3.0 803 . 00.0 03
00.0 00.0 06.0 00.0 ~83 _ 00.0 0»
00.0 3.0 3.0 00.0.. 303 _ 00.0 «:00
3.0 00.0 00.0 60.0 800.— _ 00.0 300
00.0 00.~ 00." 66.0 803 _ 00.0 60
00.0 :0 3.0 «0.0 g _ 00.0 on
00.0 00.0 06.0 3.0 g _ 00.0 00
00.0 00.0: 00.0: 60.0: 380 . «To 0.0
00.0 06.” 00.0 00..— 5§ _ 60.0 «0
3.0 00.0 «0.0 00.0 g _ 00.0 00
58:60 58¢! a it": an; . game ”3
33056560 nag
a «£85: 3340
000.0 63.03 30.0 0 63.00 F83
000.0 060.000 02.0 0 600.: 089
000.0 «00.62 000.0 0 03.0 .389
000.0 660.066 000.0 0 000.6 :83
000.0 03.0: 066.0 0 03.3 803
000.0 36.000 600.0 0 600.0 880
000.0 000.0: «00.0 0 000.0 3
000.0 000.600 02.0 0 000.6 30000
000.0 000.6: 000.0 0 "60.0 380
000.0 060.63 000.0 0 03.0 509
000.0 000.: 03.0 0 000 n g
810 056:..- 00 Inch: 8 0.0 SHE! 33;

35:35 2 050: 9.30:5 0:35.0—

33003 H N—QZHmm<

3.6300 6 6.0 80.60:.(90 6.20.50

00592.0 B335 6

APPENDIX E (cont’d)

K-means Clusters Using 11 Elements

7 Clusters Speclfled (cont'd)

0—00—000
no‘nn'.”
e. "0 °..

 

7

CLUS YER HWIEI:

163

$8§§§$§QSQSS

.—
n

5;; gastazs:

-N-°.°.—..°'-—

53933328822;

33333288888

164

 

 

 

 

 

 

00.0 00.0: 00.0: 00.0: 380 _ 00.0 00.0: 06.0: 00.0: g . 00.0 000
00.0 00.0 00.0 00..— 5§ _ 00.0 00.0 00.0 00.0 g _ 00.0 000
00.0 00.0 00.0 00.0 (g _ 00.0 000 00.0 00.0: 00.0: 00.0: g . 00.0 600
4.68.0.0 g a gin: Egg _ 2.6000 g 00 .0 00 . 0: 00. 0: 00 .0: g . 00 . 0 000
000.000.0560 ”.3 60.0 00.0: 00.0: 00.0: 380 . 00.0 000
0 ”595: E0000 00.0 00.0 66.0 60.0 580 _ 00.0 000
00.0 00.0 00.0 00.0 g . 00.0 000
00.0 00.0 00..— 00.0 mg . .5300 0520; It! gluc- gg _ "002.0an g
00.0 00.0 00.0 00.0 g _ 0:00-<90 g
00.0 00.0 00.0 00.0 .38— _ 0 $30.58 ago
00.0 00.0: 00.0: 00.0: 93 _
00.0 00.0 00.0 00:— g . . 00.0 000
00.0 00.0 00.0 00.0 0080 _ . 00.0 600
00.0 00.0 «0.0 00.0 g _ _ 00.0 000
00.0 06.0 06.0 06.0 0.080 . . 00.0 000
00.0 00.0: 00.0: 00.0: 380 . _ 00.0 000
00.0 00.0 00..— 004 580 _ _ 00.0 08000
00.0 60.0 60.0 60.0 (g _ 00.0 000 _ 00.0 «8000
(68.0.0 0.0-08‘- ; 830080! a; _ gamma "6 _ 00 . 0 000
000.000.6490 g _ 00.0 000
0 .5952 00.09.30 . 00.0 000
_ 00.0 000
00.0 00.0 00.0 00.0 2.680 . . 00.0 000
00.0 00.0: 00.0: 00.0: 0008 _ _ 60.0 600
00.0 60.0: 60.0: 60.0: 0180 _ _ 00.0 000
00.0 00.0: 00.0: 00.0: 0800 _ . 00.0 000
00.0 00.0 00.0 00.0 008.— _ _ 00.0 000
00.0 60.0 60.0 60.0 0980 . _ 00.0 000
00.0 06.0: 06.0: 06.0: g _ 00.0 00.0 2.0 00.0 05.80 _ 00.0 000
00.0 00.0: 00.0: 00.0: g _ 00.0 00.0 60.0 00.0 880 _ 00.0 000
00.0 00.0: 00.0: 00.0: 380 _ 00.0 00.0 06.0 00.0 E . 00.0 00
00.0 00.0 00.0 00.0 580 . 00.0 00.0 00.0 00.0: g _ 00.0 N300
00.0 00.0 00..— 00.0 g . 00.0 000 00.0 00.0 00.0 60.0 g . 00.0 «.00
58.0.0 2: l4! gun! a: . gamma 3 00.0 00.0 00.0 66.0 g . 00.0 60
000.600.0900 05093. 00.0 00.0 00.0 00.0 g _ 00.0 00
0 ”g E0340 00.0 00.0 06.0 00.0 g . 00.0 00
00.0 00.0: 00.0: 00.0: 303 . 00.0 00
00.0 00.0 60.0 00.0 mg _ 00.0 06..— 00.0 00.0 580 . 00.0 an
00.0 00.0 00.0 00.0 008» . 00.0 00.0 00.0 06.0 g . 00.0 «0
00 . 0 00. 0 00. 0 00 . 0 .3800 _ .58.: g: a glut gag . 00250000 50
00 . 0 00.0: 00 .0: 06 . 0: pg . 000.600.0500 0;
00.0 00.0 00.0 00.0 nua— _ 0 “is. 5.6300
00.0 00.0 00.0 00.0 g _
00.0 00.0: 00.0: 00.0: g . 000.0 000.000 00 000.0 6 000.00 E
00.0 00.0 00.0 00.0 g _ 60.0 000 000.0 600.000 00 000.0 6 000.00 0080
00.0 00.0: 00.0: 00.0: 380 _ 00.0 600 000.0 000.000 00 000.0 6 000.0 g
00.0 00.0 00.0 00.0 g _ 00.0 000 000.0 000.600 «0 000.0 6 000.6 g
00.0 00.0 00.0 60.0 (g . 00.0 000 000.0 000.000 00 606.0 6 000.00 g
.E.‘ :2 xi gnu. a; _ 2.6000 "40 000.0 006.000 00 060.0 6 000.0 g
0000.05.60 «as 000.0 000.000 00 «00.0 6 000.0 g
0 .33. 309.00 000.0 606.000 00 000.0 6 600.6 088—
000.0 600.000 00 000.0 6 000.0 380
00.0 00.0 00.0: 00.0: 00.80 . 000.0 006.000 00 000.0 6 000.0 580
00.0 80.0: 00.0: 00.0: 0080 . 000.0 000.00 00 000.0 6 000.0 g
00.0 00.0 «0.0 00.0: 0.3 . 8x0 008::0 .8 0.0 3006:! B 00 En sag
00.0 «0.0: 00.0: 60.0: g . 00.0 000 go 0 60 000.600.0560 in

.9380 0.5800 5.2.6 0

00:25.0— 3 050: 282.5 0535.0—

quoov H N—szmm<

358.; £386 a

165

00.0 00.0
00.0 00.0
00.0 00.0
00.0 06.0
00.0 00.0
00.0 00.0
00.0 06.0
00.0 00.0
00.0 00.0
00.0 00.0
>nn.00 IDZHXII

O
N

833§388383

g;...........
E

§§§§§§§§§§§§

 

00.0 00.0
00.0 00.0
00.0 06.0
00.0 00.0
00.0 00.0
00.0 06.0
00.0 00.0
00.0 60.0
60.0 60.0
>uu.hh Saluxil

00000
HUG...

33333233222

304006400000

£5
3333333
04006400000
2

00 0 000
00249000 undo
axilxul

0 .xllipl udhnaau

. 00.0 000
_ 00.0 000
. 00.0 000
. 00.0 000
_ 00.0 000
. 00.0 000
_ 60.0 000
_ 60.0 000
. 60.0 000
. 00.0 000
_ 00.0 600
. 00.0 000
. 00.0 000
_ 00.0 000
_ 00.0 00

0‘5 . nuacfinnn 0010

 

00.0 06 0
00.0 00.0
00.0 00 0
00.0 00 0
00.0 00 0
00.0 00 0
00.0 00 0
00.0 00 0

8:255 2 055 9.83.0 2.35.0—

G..§3 m x3200:

3313333333.?

"own".

0 o o a a o o I

00.400000
I

§5§§§§§§ §§§§§§§§§§§§

5
§
'3
E
3:
B
8
E
L?
0

166

 

 

 

 

 

 

000 .00.. :0: 0.0.. 303 _ 30 0.0- 2.0.. 000: g _ :0 ova
000 vs; 3 a cut" 503 . 000 30 «a0 :0 g _ 20 an»
000 :0 no ~ no.“ ‘83 _ 000 on» 20 nn.0: :0... ~00: g . .00 on»
$80.» in! a gnu! an; _ 30.03 g 000 20: 30: nn.0: g _ 30 on»
003.0355 g :0 00.7. «0.7 00.7 303 _ 30 man
0 can! and 000 30 2.0 2.00 E . 000 on»
:0 00.... an." .00 g . 30 and
000 on." 0”." 0a..— ag . $80.0 gut: i gun: a; _ 00.2030 g
00 . a on. a on. n on. a 000.3 _ 00:03.55 a;
000 000 000 000 has _ N .085: «0.0250
000 30.. 30: 30.. 0003 .
000 3.“ nu." an.” 0009 _ _ 000 on»
30 no.0. 30 no.“ 089 _ . 30 ran
80 30 «n0 :0 >08 _ _ 000 0rd
000 2.0 2.0 2.0 8009 _ _ 000 or.»
000 «n0: «n0. 30: 308 . . 00 can
000 000 00:— 00.“ 500.— . _ 000 «29.0
000 2.0.0. :0 :0 489 _ 000 :0 _ 000 an...»
58 . .8 g; I; gun! a; _ gas 50 _ Nu . c «an
82.3.2.5 g . no . c «an
n .g «H.030 _ 30 on»
. 30 an»
000 30 30 30 3.53 . _ 000 24.0
000 v00- 30: 00: 0000.— _ _ 30 run
000 :0: :0: :00 fig _ . 30 «an
000 out? 3.7 onél 0009 _ _ 30 «an
000 30 30 30 0003 _ . 30 0:
000 :0 :0 2.. o g _ _ .00 30
000 2. on 2.0. 2. 0.. >03 . .00 ca.” :0 000 E . 30 «.3
000 no on 30. av on 8003 _ 000 nu.” no." 30 88.— _ 000 can
80 3 u: 3.? av 7 303 . 000 :0 2.0 000 fig . 30 an
000 3 o :0 3 o 508 . 30 =0 r00 :00: 0303 . 30 «8.0
000 an a an." an H g . 000 000 «"0 :0 000 2.00 0008 _ 000 :30
580.0 g a gau- gg _ 2.33 g 000 000 .0:— 24 g _ 30 2.0
00:03.5 g no.0 30 «v0 nn.0 g . 30 00
0 .3! E230 000 0.0 2.0 30 g . 000 an
30 30: 30: 3.0. 308 _ 20 3
2.0 000 S0 :0 E _ 000 2.0 0nd 3." 503 _ 000 «n
000 30 nn.0 :0 0003 _ 000 000 3.“ 2.0 g . 000 an
000 on... 0&0 20 .509 . 58.80 g It! gun: a; _ 83450 g
3 . 0 3. on no . on 2. . on 0009 _ 00:395.» «ii
30 000 «n.0 0«0 0000.— _ a .2 go
.00 3.— 3." 004 0003 _
000 000.. :0: «n0: 3 _ 000 0 30.4.: on 030 0 30.3 5.03
30 as... :0 30 g _ 30 own 0000 non.va on 2.30 o 000.: 0009
20 «00: 0.0: 3.7 309 . on o :0 0000 00~03 on 3N0 0 ~30 fig
30 S n so.” :0 509 _ on o :0 0000 «:03 on N30 0 3.... pg
30 3 u :0 S a g . 30 no: 0000 3903 on 02. o 0 mad— 3
450.: g; a gun: in; . gas .40 0000 02.02 on 2h 0 0 000.0 g
00:052.; a; 0000 30.non or. mom 0 o 03.0 >03
n .535: g 0000 3.0: or. :00 0 an»... I000.—
0000 «202 on 030 o 030 303
30 30 30: 30.. 5.09 . 0000 30.02 or. 230 0 .00.... 509
«00 «n0: :0. :0: 0009 . . 0000 03.0... on can... u «a n g
30 30 «00 :0- 0:09 . g 02.57.. no 00 gas .8 an as mag;
30 «00.. :0: :0: 0g . go a E 002.035.» 2.5050
$2.8. 858.5 5.85 a 353.5 9.336 a

35520 n— 05.5 222:0 2.35.0—

3.288 H N—QZm—mm<

APPENDIX E (cont'd)

K-means Clusters Using 11 Elements

9 Cinders Specified (cont'd)

9 Clusters Specified (cont'd)

O 3’
-0.1Q

.
0H

eé

-0.39
-0 14

§§

00000000

28333382

00000000

 

167

~FQ'.“.H....
000000140006

...........
OOOOOOOOOOO

H NdOOOe-OOOOOO
2 l I

33333332228
...........

§§§§§§§§§§§§

”NOB

“FFPCWOU'.

rte-COOSOOOOOCOOOO

BOOOOOOOOOOOOOOO

«Ea

an CCCCCC Oreo-0N

g 2338883322833

'3
a.

 

; 3323223888
N

SIAIISTICS

.HvNe-OOOHSHC
HUICFOCFO ON
a

...........
NC" COHO?°F¢M

§§§§§§§§§§§§

 

 

 

99.9 99.9.. 99.9.. 09.9:

 

 

 

168

 

 

 

. 9.989 . 803 _
99 9 99.9 99 9 4903 _ 99.9 99.9 99.9 99.9 9093 _ 99.9 999
99 9 99.9 99 9 g . 99.9 999 99.9 99.9- 99.9: 99.9.. .63 _ 99.9 999
9999995- : 9:999:92 a; . 2.9999 g 99 .9 99 . 9.. 99. on 99 .9- g _ 9o . 9 999
9099.99.92.99 99989! 99.9 99.9: 99.9- 99.9- 303 _ 99.9 999
9 "59.52 9.8.99.5 99.9 99.9 99.9 99.9 (993 . 99.9 999
99.9 99.9 99.9 99.9 938 _ 99.9 999
99.9 99.9 99.9 .9989 _ $8.99 9.399999. at! 99999999. 992996, _ 90249999 9940
99.9 99.9 99.9 0903 . 899999.599 9.93
99.9 99.9 99.9 99809 _ 9 5805: 5.9390
99.9: 99.9: 99.9: :83 _
99.9 99.9 99.9 9809 . . 99.9 999
99.9 99.9 99.9 889 . . 99.9 999
99.9 99.9 99.9 53 _ _ 99.9 999
99.9 99.9 99.9 .9903 . _ 99.9 999
99.9... 99.9.. 99.9- 393 _ . 99.9 999
99.9 99.9 99.9 9.989 . _ 99.9 98999
99 9 99.9 99 9 4g _ 99.9 999 . 99.9 93999
99.99995- 33: 99.99929: Q9293 . 8259.999 9940 . 99 . o 999
909.999.94.99 nag _ 99.9 999
9 “59.5: «8.99.5 . 99.9 999
. 99.9 999
99.9 99.9 99.9 999.93 _ _ 99.9 999
99.9.. 99.9: 99.9.. 0903 _ _ 9o 9 999
99.9- 99.9: 99.9.. .3989 . . 99.9 999
99.9- 99.9: 99.9: 9989 _ _ 99.9 999
99.9 99.9 99.9 9009 _ _ 99.9 999
99.9 99.9 99.9 803 _ . 99.9 999
99.9: 99.9: 99.9: g . 99.9 99.9 99.9 99.9 .958 _ 99.9 999
99.9: 99.9: 99.9: 8989 . 99.9 99.9 99.9 99.9 0993 _ 99.9 999
99.9- 99.9- 99.9- 9.989 _ 99.9 99.9 99.9 99.9 .9809 . 99.9 99
99.9 99.9 99.9 9.9909 . 99.9 99.9 99.9 99.9- 9993 . 99.9 9199
99.9 99.9 99.9 (g . 99.9 999 99.9 99.9 99.9 99.9 9089 _ 99.9 9999
959995- ; 9599998. 1929959 _ 90.2.9999 9940 99.9 99.9 99.9 99.9 80.3 _ 99.9 99
909.999.92.99 99993: 99.9 99.9 99.9 99.9 g _ 99.9 99
9 .5959 99999.5 99.9 99.9 99.9 99.9 “as _ 99.9 99
99.9 99.9: 99.9.. 99.9- 9.53 _ 99.9 99
99.9 99.9 99.9 99.9 .9989 _ 99.9 99.9 99.9 99.9 593 _ 99.9 99
99.9 99.9 99.9 99.9 0989 _ 99.9 99.9 99.9 99.9 g _ 99.9 99
99.9 99.9 99.9 99.9 .9983 . $8.99 999993 I! 99.999899. g9; _ 8.2.9.990 9940
99 . 9 99. on 99 . 9: 99 . 9.. pg . 909.999.9599 9;
99.9 99.9 99.9 99.9 9003 . 9 .9859 «9.9.3.90
99.9 99.9 99.9 99.9 889 _
99.9 99.9.. 99.9: 99.9- .609 . 999 9 999.999 99 999 9 9 999.99 99.83
99.9 99 9 99.9 99 9 .9903 . 99.9 999 999 9 999.999 99 999 9 9 999.99 0989
99.9 99 9.. 99.9. 99 9.. 389 . 99 o 999 999 9 999.999 99 999 o 9 999.9 9.809
99.9 99 9 99.9 99 9 509 . 99 9 999 999 9 999.999 99 999 9 9 999.9 9809
99.9 99 9 99.9 99 9 (g . 99 o 999 999.9 999.99 99 999 9 9 999.99 9989
58. 99.999959 :1! 9999939! sag . 905.9999 9940 999.9 999.999 99 999 o 9 999.9 9089
909.999.94.99 9999:! 999.9 999.999 99 999 9 9 999.9 g
9 .9995. 99999.90 999.9 999.999 99 999 9 9 999.9 8989
999.9 999.999 99 999 9 9 999.9 389
99.9 99.9: 99.9- .9989 . 999.9 999.999 99 999 o 9 999.9 589
99.9: 99.9.. 99.9.. 0909 . 999.9 999.99 99 999 9 9 999.9 9809
99.9 99.9 99.9» .389 _ :9 09.9.1919 B 99 39.99.98 .8 9.9 2 9.995;
99.9- 99.9: 99.9.. .9809 _ 3.99990 99 69 9099999599 S
95.8. 358% £336 2 3592.9 5.36 3

92.0an 99 959: 9.335 9535.99

3.288 m— N—GZm—mA—t.

169

99.9

09.9

90.9:
9«.9:
90.9:
99.9:
09.«

90.9:
v«.9:
o«.«:

«0.«

g It!
8«h0«.«.<hn

 

 

§§§§§§§§

 

9« .505: Emu—AU

 

v«.9:
00.9:

3.83 859...... £3.39 2

3.253 m “Calm—.30.

09.9: 09.9: 09.9: kg _
90.9: 90.9: 90.9: g _
«9.9: «9.9: «9.9: g _
90.9 90.9 90.9 g _
««.9: ««.9: ««.9: g .
««.9: ««.9: ««.9: g .
90.9: 90.9: 90.9: g .
99.9 99.9 99.9 g _
90.« 90.« 90.« g .
g; a. g 3; _
BEE—.590
“g go
9«.« 9«.« 9«.« E _
09.« 09.« 09.« g _
90.9 8.9 90.9 E _
09.9: 09.9: 09.9: g .
09.9 09.9 09.9 g _
09.« 09.« 09.« g .
«9.9 «9.9 «9.9 g _
«9.9 «9.9 «9.9 g _
99.9: 3.9: 99.9: 38.— .
«0.« «0.« «0.« (Ag _
90.« 90.« 90.« g _
gas- 3 g a; _
8:02.55
.0852 E5
_
_
_
.
0«.« «0.9 99.9 g _
«9.« 00.9 09.9 g .
«9.9 99.9 00.9 g .
99.9 99.9: ««.9: g _
99.9 90.9 9«.9 g .
09.« 09.« «9.« g _
90.9 «0.9 ««.9 g _
99.9 09.9 90.9 g .
90.9: 90.9: 99.9: a _
90.« 09.« 9«.« g _
99.« v9.« «0.« g .
83:; a Elfi- gflug _
8559590 fig
9 ”g as
09.9 09.9 09.9 g _
99.9 09.9 99.9 g _
99.9 99.9 99.9 g _
«0.9: «0.9: «0.9: «as _
90.9 90.9 90.9 083 _
90.« 90.« 90.« g _
«9.9: «9.9: «9.9: g _
00.9 00.9 00.9 g .

EcoEoE «« ”5.5 233.0 2305.9—

€....§ .8583 p.336 2

 

 

 

 

 

170

 

 

 

90.9 90.9 90.9 983 _ 99.9 99.9 9«.9: «9.9: «983 .
90.« 90.« 90.« 3 _ 09.9 90.9: 99.9: 00.9: 3 _
«9 9: «9.9: «9 9: 3 . 09.9 9«.9 99.9 99.9: .983 _
00 9 00.9 00.9 953 _ 09.9 «9.9: 90.9: 90.9: 3 _
99 9: 99.9: 99.9: 393 . 9«.9 99.9: 99.9: 90.9: 803 .
99 « 99.« 99 « 3 _ 09.9 00.9 90.9 99.9 3 _ 9«.9 999
09 9 09.9 09.9 3 . 99.9 909 99.9 90.9: 90.9: «9.9: 3 _ ««.9 009
g a «99.9908 9.0953 . 9099993 ‘90 99 . 9 9« . 9: 99 . 9: 09 .9: «l3 _ 99 . 9 909
93999299 9999‘. 0« . 9 99 . «: 90 . «: 99 . «: 33 _ 9« . 9 909
9 "in! 9.999.909 99.9 99.9 09.9 99.9 («3 _ 9«.9 009
9«.9 99.9 90.9 99.9 3 _ 09.9 909
9«.« 9«.« 9«.« =3 _ $3.99 «99.«; a 99.9-«99: as; . 90.29999 9949
90. « 90 . « 90 . « U993 . 9909909999 999'!
99.9 99.9 99.9 3 . 9 .9905: 99999.5
09.9: 09.9: 09.9: :3 .
0«.« 0«.« 0«.« 9003 _ . 99.9 009
09.9 09.9 09.9 3 . . 99.9 909
«0.9 «0.9 «0.9 3 _ . 99.9 909
09.9 09.9 09.9 3 . . 99.9 009
90.9: 90.9: 90.9: 9.83 . . 09.9 909
99.« 99.« 99.« 3 _ . 99.9 92009
99.9 99.9 99.9 3 _ 99.9 999 _ 09.9 «9009
9990094: .2! 99.9-«3 9.2g _ 9999993 g . 99 . 9 «09
93992299 9983. _ 09.9 909
0 .39! 99999.6 . 99.9 909
. 99.9 099
09 9 09.9 09 9 «093 _ _ 99.9 999
99 9: 99.9: 99 9: 03 . . 09.9 «99
99 9: 99.9: 99 9: 3 . 99.9 99.« 09.« 00.9 5.93 _ 09.9 93
90 «: 90.«: 90 «: 3 _ 99.9 «9.« 99.« 90.9 3 . 99.9 0«9
09 9 09.9 09 9 3 _ 09.9 99.9 99.9 99.9 3 . 09.9 9«9
99 9 99.9 99 9 3 _ 09.9 9«.9 09.9 09.9: 9993 . 09.9 39
09 9: 09.9: 09 9: 3 _ 09.9 09.9 «9.9 90.9 9903 _ 09.9 09
09 9: 09.9: 09 9: :93 . 99.9 09.9 99.« 99.« 3 _ 99.9 9999
09 «: 09.«: 09 «: 9.83 . 99.9 «0.9 99.9 00.9 .83 _ 99.9 «999
99 9 99.9 99 9 3 _ 09.9 «9.9 99.9 99.9 .893 _ 99.9 99
00 « 00.« 00 « 3 _ 99.9 999 99.9 00 9: 99.9: 90.9: 303 _ 99.9 99
«99.«; .28.. 999.993 3; . 90.2993 993 99.9 09 « 90.« 99.« («93 . 99.9 99
9399599 g 99.9 90 9 09.9 99.9 3 . 99.9 «9
9 .5991 99999.8 53.99 g i 99:99: a; . 90.2993 9949
999992.999 g
09.9 09.9 90.9 09.9 9993 . « «g 99999.5
99.9 99.9 00.9 99.9 0093 _
99.9 90.9 99.9 0«.9 .383 . 999.9 990.00« 99 9«0.9 9« 900.9« :93
9«.9 99.9: «9.9: 99.9: 93 . 999.9 999.090 99 999.9 9« 009.9« 0993
99.9 90.9 90.9 99.9 9003 _ 999.9 9«9.99« 99 99«.9 9« 990.0 0.83
99.9 «9.« 90.« 99.« 3 . 999.9 00«.«9« 99 999.9 9« 999.9 9993
09.9 09.9: 9«.9: 90.9: 3 . 999.9 990.«0 99 909.9 9« 909.0« 903
9«.9 09.9 0«.9 99.9 3 _ 99.9 999 999.9 «09.90« 99 9«9.9 9« 990.9 3
0«.9 «9.9: 00.9: ««.«: 3 . 9«.9 999 999.9 «9«.«99 99 «9«.9 9« 909.9 3
9«.9 ««.« 99.« 99.9 3 _ 9«.9 099 999.9 ««9.99« 99 999.9 9« 999.9 8993
9«.9 09 9 90.9 9« 9 3 . 99.9 099 999.9 009.99« 99 009.9 9« «99.0 393
$3.99 gas. a gun! ag . 9039993 349 999.9 999.0«« 99 909.9 9« 900.0 53
93939499 9999.3. 999.9 999.99 99 090.9 9« «09.0 3
0 .39! 99999.5 39 3949:.— 3 99 99393 B 99 99.9393 a;
999939 «« 6.9 93939599 995959
«383 35.2.0 5.36 = 85.2.... :35 ««

3:25.9— «« 950: 23020 0:90:79—

33998 n.— N—DZm—mma.

171

 

 

 

 

 

 

99.9 9«.9: 9«.9: 9«.9: :93 _
99.9 00.9: 00.9: 00.9: 003 _
99.9 09.9: 09.9: 09.9: 3 _
99.9 90.9: 90.9: 90.9: 993 _
99.9 99.9.. 99.9: 99.9: 903 _
99.9 90.9 90.9 90.9 903 .
99.9 9«.9: 9«.9: 9«.9: 83 _
99.9 «9.9: «9.9: «9.9: 3 _
99.9 90.9 90.9: 90.9: 33 _
99.9 99.9 99.9 99.9 <83 _
99.9 90.9 90.9 90.9 3 . 99.9 900
$3.90 33 99! 95395. 993; _ 8.2908 908
03939599 9.9909!
0 ”.9859 8909.8
99.9 99.« 99.« 99.« :93 .
99.9 09.« 09.« 09.« 003 .
99.9 90.9 90.9 90.9 3 .
99.9 09.9: 09.9: 09.9: 93 _
99.9 09.9 09.9 09.9 003 _
99.9 09.« 99.« 09.« 3 .
99.9 «9.9 «9.9 «9.9 .63 _
99.9 99.9 99.9 99.9 3 _
99.9 99.9: 99.9: 99.9: 9.83 _
99.9 «0.« «0.« «0.« 53 _
99.9 90.9 90.9 90.9 3 _ 99 9 999
99.9 99.« 99.« ««.« :3 . 93.90 99.999959 9:! 299: 32g _ 8.2.9090 908
99.9 09.« 99.« 00.9 093 . 0099999390 g
99.9 99.9 09.9 99.9 3 _ 9 2.3959 99.930
99.9 9«.9 09.9 99.9 93 .
9«.9 99.9 90.9 90.9 93 _ . 99.9 900
99.9 09.9 90.« 90.« 93 _ _ 99.9 990
09.9 90.9 99.9 09.9 3 _ _ 99.9 990
99.9 90.9 99.9 99.9 9.03 _ 99 9 909 _ 09.9 «99
09.9 90.9: «9.9: 09.9: 9.83 _ 99 9 999 09.9 0«.« 99.9 99.9 :93 _ 99.9 «09
09.9 09.« 09.« 90.« 4.83 . 09 9 09 99.9 «9 « 00.9 09.9 003 _ 99.9 999
09.9 90.9 09.9 09.9 3 . 99 9 90 99.9 99 9 99.9 00.9 .983 . 99.9 090
$3.90 8.99995. 94! gnu. 9.095; _ 8.2903 9040 99.9 99 9 99.9: 99.9: 3 _ 99.9 990
039999599 09999! 0«.9 99.9 90.9 99.9 003 _ 99.9 099
«« ”.0995.- 99909.8 99.9 09 « 09 « 99.« 3 . 09.9 9«9
09.9 90.9 «0.9 «9.9 83 . 99.9 39
99.9 09.9 09.9 09.9 :93 _ 99.9 99.9 09.9 90.9 3 . 99.9 9«9
99.9 90.9: 90.9: 90.9: 003 _ 99.9 90.9: 90.9: 99.9: 9.83 _ 09.9 0«0
99.9 9«.9: 9«.9: 9«.9: 3 . 09.9 90.« 09.« 99.« 3 _ 9«.9 ««9
99.9 90.9: 90.9: 90.9: 983 _ 99.9 99.9 99.9 99.9 83 _ ««.9 09
99 . 9 99 . 9: 99 . 9: 99 . 9: 03 _ 3.99 «99395. It! .3590! 3; _ 9095939 998
99 . 9 09 . « 09 . « 09 . « 93 _ 9099099599 03
99.9 90.9: 90.9: 90.9: 3 _ 9 ”g 8.99.8
99.9 99.9: 99.9: 99.9: :03 .
99.9 9«.«: 9«.«: 9«.«: 9.83 .
99.9 90.9 90.9 90.9 9.83 _
99.9 «0.9 «0.9 «0.9 83 . 99.9 000
$99.99 999395- a 396. 333 . 99.2998 9040 99.9 09.9 09.9 09.9 :93 _
9099999599 9999.! 99.9 99.9 99.9 99.9 3 _
9« “gs—“90.98 99.9 99.9 99.9 99.9 3 _
99.9 «0.9: «0.9: «0.9: 3 _
«5.83 358...... 5.36 = 6.2.8. 25890 2.386 =

3552.9 «« 9503 232:0 0.329.9—

29.9998 m— EDZHQA<

APPENDIX F

172
APPENDIX F

Elements Detected by INAA and Used in Pottery Paste Analyses

Michigan Memorial Phoenix Project / Ford Neuclear Reactor (Dunn)
l. W (1 minute irradiation)
W
a. Count after 13 minutes of decay (half-lives up to 30 min):

Aluminum, Bromine, Chlorine, Copper, Iodine, Magnesium, Niobium, Palladium,
Titanium, Vanadium

b. Count after 2 hours ofdecay (half-lives up to 1 112 days):

Arsenic, Dyaprosium, Erbium, Gallium, Germanium, Manganese, Nickel, Potassium,
Praseodynium, Sodium, Tungsten

 

2. WW for 6 hours):
a. Counted after 1 week of decay (half-lives 1-10 days):

Barium, Bromine, Cadmium, Lanthanum, Luthenium, Molybdenum, Neodymium,
Samarium, Uranium, Ytterbium

Excluded: Bromine, Cadmium, Molybdenum, Neodymium, Uranium
11m: Barium, Lanthanum, Luthenium, Samarium, Ytterbium
W: none

b. Counted after 5 weeks of decay (half-lives 15 days to 5 years)

Antimony, Cerium, Cesium, Chromium, Cobalt, Europium, Gadolinium, Hafnium,

Indium, Iron, Mercury , Nickel, Rubidium, Scandium, Selerium, Strontium, Tantalum,

Terbium, Thorium, Thulium, Tin, Zinc, Zirconium _

Excluded: Antimony, Chromium, Cobalt, Gadolinium, Indium, Iron, Mercury , Nickel,
Rubidium, Selerium, Strontium, Tantalum, Terbium, Thulium, Tin, Zinc,
Zirconium

11m: Cerium, Cesium, Europium, Hafnium, Scandium, Thorium

WW: none

 

173
APPENDIX F (cont'd)

 

Elements Detected by INAA and Used in Pottery Paste Analyses

Clark (1991):
l. W (Irradiated for 5 seconds)
a. Counted after25 minutes ofdecay:

Aluminum, Barium, Calcium, Dyaprosium, Potassium, Manganese, Sodium, Titanium, '
Vanadium

W:Aluminum, Calcium, Potassium, Sodium
Used: Barium, Dyaprosium, Potassium, Manganese, Sodium, Titanium, Vanadium
121mm: rare earth elements

 

2. WWM for 24 hours):
a. Countedafter l weekofdecay:

Arsenic, Lanthanum, Luthenium, Neodymium, Samarium, Uranium, Ytterbium
Excludedfirsenic, Neodymium, Uranium

Used: Lanthanum, Luthenium, Samarium, Ytterbium

W: rare earth elements

b. CounmdafterSweeksofdecay:

Antimony, Cerium, Cesium, Chromium, Cobalt, Europium, Hafnium, Iron, Nickel,
Rubidium, Scandium, Strontium, Tantalum, Terbium, Thorium, Zinc, Zirconium

Excluded: Cesium, Chromium, Cobalt, Hafnium, Nickel, Strontium, Tantalum, Zinc,
Zirconium

Used: Antimony, Cerium, Europium, Iron, Rubidium, Scandium, Terbium, Thorium,
W: rare earth elements + Iron and Scandium

 

 

174
APPENDIX F (cont'd)

Elements Detected by INAA and Used in Pottery Paste Analyses

Gunneweg and Mommsen (1990):
Methcdnmmscrihed (used inmedim and long-lived isotopes»
Arsenic, Barium, Cerium, Cesium, Chromium, Cobalt, Europium, Hafnium, Iron, .
Lanthanum, Lutetium, Potassium, Rubidium, Samarium, Scandium, Sodium, Strontium,
Tantalum, Terbium, Thorium, Ytterbium, Zinc
mm: none reported
Used: all

' : Cerium, Cesium, Europium, Iron, Lanthanum, Samarium,
Scandium, Sodium

Hancock, et al. (1989):

W (1 minute or longer, based on sample weight) with
W (counwd after 19 minutes of decay):

Aluminum, Barium, Calcium, Chlorine, Dyaprosium, Magnesium, Manganese, Sodium.
Titanium, Uranium
Excluded: Barium, Chlorine, Magnesium, Uranium
Used: Aluminum, Calcium, Dyaprosium, Manganese, Sodium, Titanium
W: none reported
Lang mm 'm (10—20 minutes) with
W (counted after overnight decay):
Europium, Potassium, Samarium, Sodium [sic]
Excluded: none
11m: all
Wm: none reported

175
APPENDIX F (cont'd)

Elements Detected by INAA and Used in Pottery Paste Analyses

Minc, et at. (1989):
MW (used intermediate and long-lived isotopes):

Antimony, Barium, Calcium, Cerium, Cesium, Chromium, Cobalt, Europium,
Gadolinium, Hafnium, Iron, Lanthanum, Lutetium, Neodymium, Nickel, Potassium,
Rubidium, Samarium, Scandium, Sodium, Tantalum, Terbium, Thorium, Titanium,
Uranium, Ytterbium
Excluded: none W
listed: all

:Cerium, Cesium, Chromium, Cobalt, Iron, Hafnium, Lanthanum,
Rubidium, Scandium, Tantalum, Thorium, Titanium, Uranium, Ytterbium

Olin and Blackman (1983):

W
Stream (counted after 5 day decay)

Antimony, Arsenic, Barium, Bromine, Calcium, Lanthanum, Lutetium, N eodymium,
Potassium, Samarium, Sodium, Tungsten, Uranium, Ytterbium,

W3 nonemporwd
um- all
Diaenosficlilemems: none

I 1.. 11 I: 'l I
Imam! (counted after 30 day decay)

Cerium, Cesium, Chromium, Cobalt, Europium, Hafnium, Iron, Rubidium, Scandium,
Strontium, Tantalum, Terbium, Thorium, Zirconium, Zinc,

Excluded: none reported
11m all
Diagnostic Elements: Chromium, Iron, Scandium

176
APPENDIX F (cont'd)

Elements Detected by INAA and Used in Pottery Paste Analyses

Bieber, et al. (1976):
W (irradiation and decay times not reported):
Lanthanum, Manganese, Potassium, Sodium
W: none reported
Used: all
121W: Lanthanum, Manganese, POtassium, Sodium

WW (irradiation and decay times not reported):

Antimony, Barium, Cerium, Cesium, Chromium, Cobalt, Europium, Hafnium, Iron,
Luthenium, Rubidium, Scandium, Tantalum, Thorium

Excluded: nonereported
Used: all
Diamsficiilennnts: Cerium. Hafnium

LIST OF REFERENCES

177

LIST OF REFERENCES

Arnold, Dean E., Hector Neff, and Ronald L. Bishop .
1991 Compositional Analysis and "Sources" of Pottery: An Ethnoarehaeologrcal
Approach. American Anthropologist 93: 70-90.

Bicber, A. M., Jr., D. W. Brooks, G. Harbottle, andE. V. Sayre
1976 Application of Mutivariate Techniques to Analytical Data on Aegean
Ceramics. Archaeometry 18(1): 59- 74

Bishop, Ronald L. and Hector Neff
1989 Compositional Data Analysis in Archaeology. Advances in Chemistry Series
220, Archaeological Chemistry IV: 57-86.

Brashler, Janet
1981 Early Late Woodland Boundaries and Interaction: Indian Ceramics of
Southern Lower Michigan. Publications of the Museum, Anthropological
Series 3 (3). Michigan State University, East Lansing, MI.

Brose, David S. James A. Brown and David W. Penney
1985 Ancient Art of the American Woodland Indians. Harry N. Abrams, Inc.
New York.

Brown, Edgardo and Richard B Firestone
1986 Table of Radioactive Isotopes. Virginia S. Shirley, Editor. John Willey and
Sons, New York.

Bunzel, Ruth
1972 The Pueblo Potter: A study of Creatrve Imagination tn Primitive Art. Dover
Publications, Inc. New York. (Originally Published by Columbia
University Press, New York, in 1972.)

Clark, Caven P.
1991 Group Composition and the Role of Unique Raw Materials in the Terminal
Woodland Substage of the Lake Superior Basin. Ph.D. Dissertation.
Michigan State University, East Lansing.

Cleland, Charles E.
1966 The Prehistoric Animal Ecology and Ethnozoology of the Upper Great
Lakes Region. University of Michigan Museum of Anthropology,
Anthropological Papers 29. Ann Arbor.

Door, John A. and Donald F. Eschman
1977 Geology of Michigan. The University of Michigan Press, Ann Arbor.

178

Ellis, H. Holmes
1969 Surface Exposrnes of Flint and Chert in Michigan. In Selections From The
Michigan Archaeologist Volumes 140, Michigan Archaeological Society
Special Publication No. I, assembled by James E. Fitting, pp. 109-120.
Ann Arbor. (Manuscript originally published in The Michigan
Archaeologist, volume 6 number 4.)

Erdtrnan, Gerhard
1976 Neutron Asctivation Tables. Kernchemie in Einzeldarstellungen, Volume 6.
Verlag Chemie, Weinheim and New York.

Erdtrnan, Gerhard
1979 The Gamma Rays of the Nuclides: Tables for Applied Gamma Ray
Spectrometry. Kernchemie in Einzeldarstellun gen, Volume 6. Verlag
Chemie, Weinheim and New York.

Fitting, James E.
1973 The Schultz Site at Green Point: A Stratified Occupation Area in the
Saginaw Valley of Michigan. University of Michigan Museum of
Anthropology, Memoirs 4. Ann Arbor.

1975 The Archaeology of Michigan: A Guide to the Prehistory of the Great
Lakes Region, Second Edition. Cranbrook Institute of Science, Bloomfield
Hills, MI.

Griffin, James B. -
1983 The Midlands. In Ancient North Americans, edited by Jesse Jennings, pp.
243-301. W. H. Freeman and Co., New York.

Gunneweg, J. and H. Mommsen
1990 Instrumental Neutron Activation Analysis and the Origin of some Cult
gzbjfcts andsEdomite Vessels from the Horvat Qitmit Shrine. Archaeometry
1): 7-1 .

Hambacher, Michael
1992 The Skegemog Point Site: Continuing Sudt'es in the Cultural Dynamics of the
Carolinian-Canadian Transition Zone. Ph.D. Dissertation. Michigan State
University, East Lansing, MI.

Hancock, R. G. V., S. J. Flemming and W. D. Glanzman
1989 Analysis of Neolithic Iranian Ceramics. Advances in Chemistry Series 220,
Archaeological Chemistry IV: 1 13-124

Holman, Margaret B. and Kathryn C. Egan
1985 Processing Maple Sap With Prehistoric Techniques. Journal of
Ethnobiology 5 (1): 61-75.

Kinietz, W. Vernon
1965 The Indians of the Western Great Lakes 1615-1760. The University of
Michigan Press, Ann Arbor.

Lovis, William A., Jr.
n.d. Minimum Vessel Sheets for the O'Neil Site. Data on file at the Anthropology
Division, Michigan State University Museum, East Lansing, MI

179

Lovis, William A., Jr.
197 3 Late Woodland Cultural Dynamics in the Northern Lower Peninsula of
Michigan. Ph.D. Dissertation, Michigan State University. East Lansing.

1991 Site Formation Processes and the Organization of Space at the Stratified
Late Woodland O'Neil Site. In The Woodland Tradition in the Western
Great Lakes: Papers Presented to Elden Johnson, edited by Guy Gibbon,
pp. 195-211. University of Minnesota Publications in Anthropology
Number 4. Minneapolis.

Luedtke, Barbara
1976 Lithic Material Distributions and lnteracttion Patterns During the Late
Woodland Period in Michigan. Ph. D. Dissertation, University of
Michigan. Ann Arbor.

McPherron, Allan
1967 The J untunen Site and Late Woodland Prehistory of hte Upper Great Lakes
Area. University of Michigan Museum of Anthropology, Anthropological
Papers 30. Ann Arbor.

Meyers, Thomas and Mark Denies
1972 Longterm and Peakscan: Neutron Activation Analysis Computer Programs.
University of Michigan Museum of A nthropology Technical Report
Number 2. Ann Arbor.

Michigan State University Museum, Anthropology Division
n.d. Accession Cards. Unpublished catalogue cards of the Michigan State
University Museum, East Lansing, Michigan.

Minc, Leah, Mary G. Hodge, and M. James Blackman
1989 Neutron Activation Analysis of Early Aztec Orange Wares in the Valley of
Mexico. Paper presented at the 12th Annual Midwestem Conference on
Mesoamerican Archaeology and Ethnohistory, March 11, 1989, Ann
Arbor, MI.

Mommsen, H.
1981 filters to Sort Out Pottery Samples of the Same Provenience From a Bank
of Neutron Activaiton Analyses. Archaeometry 23(2): 209-215.

Mommsen, H., A. Kreuser and J. Weber
1988 1;“ Method for Grouping Pottery by Chemical Composition. Archaeometry
1): 47-57.

Moran, Emilio F.
1979 Human Adaptability. Westview Press, Boulder, Colorado.

Nuclear Reactor Laboratory, Michigan Memorial Phoenix Project
n.d. Radiation, Analytical, Test and Training Services. Manuscript on file at the
Michigan Memorial Phoenix Project, University of Michigan, Ann Arbor,
MI.

180

1988 The University of Michigan Nuclear Reactor Laboratory Quarterly Review,
Spring, 1988. Unpublished Manuscript on file at the Michigan Memorial
Phoenix Project, University of Michigan, Ann Arbor, MI.

Olin, Jaquelin S. and M. James Blackman
1989 Compositional Classification of Mexican Majolica Ceramics of the Spanish
Colonial Period. Advances in Chemistry Series 220, Archaeological
Chemistry IV: 87-112.

Orlove, Benjamin S.
1980 Ecological Anthropology. Annual Review of Anthropology 9: 235-73.

Sayre, Edward V.
1977 Brookhaven Procedures for Statistical Analyses of Multivariate
Archaeometric Data. Manuscript on file (file number BNL-21693) at the
Chemistry Department, Brookhaven National Laboratory, Upton, NY.

Trigger, B.G., L. Yafi’e, M. Diksi, J.L. Galinier, H. Marshall, and J. F. Pendergast
1980 Trace Element Analysis of Iroquoian Pottery. Canadian Journal of
Archaeology 4: 119-145.

Vayda, Andrew P. and Bonnie J. McCay
1975 New Directions in Ecology and Ecological Anthropology. Annual Review of
Anthropology 4: 293-306.

Vogel, Thomas A., John C. Eichelberger, Leland W. Younker, Benjamin C. Schuraytz,
John P Horkowitz, Harlan W. Stockman, and Henry R. Westrich
1989 Petrology and Emplacement Dynamics of Intrusive and Exrusive Rhyolites
of Obsidian Dome, lnyo Craters Volcanic Grain, Eastern California. Journal
of Geophysical Research, 94 (BIZ): 17,937-17,956.

Wilkinson, Leland
1989 SYSTAT: The System for Statistics. SYSTAT, Inc., Evanston, IL.

Wahla, Edward J.
1981 Indian Projectile Points Found in Michigan. Manuscript published by the
Clinton Valley Chapter of the Michigan Archaeological Society, Second
Printing. (Originally published in 1969.)

Yarnell, Richard
1964 Aboriginal Relationships Between Culture and Plant Life in the Upper Great
Laked. University of Michigan Museum of A nthropology, Anthropological
Papers 23. Ann Arbor.