The Evergreen State College Institutional Repository

Ferguson_PMESthesis2015.pdf

Media

Part of Predicting Historical Logging Camp Locations in the Capitol State Forest, Wa

extracted text (extracttext:extracted_text)

PREDICTING HISTORICAL LOGGING
CAMP LOCATIONS IN THE
CAPITOL STATE FOREST, WA

by
Patrick J. Ferguson

A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
June 2015

©2015 by Patrick J. Ferguson. All rights reserved.

This Thesis for the Master of Environmental Studies Degree
by
Patrick J. Ferguson

has been approved for
The Evergreen State College
by

________________________
Kevin Francis, Ph.D.
Member of the Faculty

________________________
Date

Abstract
Predicting Historical Logging Camp Locations in the Capitol State Forest, WA
Patrick J. Ferguson
Historical logging camps represent an important period in the resource extraction history
of the United States. Logging camps can provide historical context about the people who
inhabited these camps. Identifying and locating historical logging camps allows
archaeologists to collect data from them, furthering research about these specific site
types. Finding historical logging camps also helps land managers protect those sites from
major disturbances. Due to their temporary nature, many historical logging camps were
undocumented, making them difficult to locate and manage around. Demonstrating there
is a measurable variable for site locations would improve the ability of archaeologists and
land managers to identify and protect undocumented sites. One possible variable is the
distance between camps. The Mason County Logging Company (MCLC) was the largest
logging company that operated in the Capitol State Forest located near Olympia,
Washington, during the late nineteenth and early twentieth centuries. To date, 15 known
and assumed logging camps used by MCLC have been identified, but there are large gaps
in the forest where no MCLC camps are known to have existed. Calculating the average
distance, plus or minus one standard deviation, between known and assumed MCLC
camps could identify other undocumented logging camps used by the company. Average
distances were obtained by mapping the known and assumed MCLC logging camps in
ArcMap and calculating the distance between them by rail line. Each distance segment
was field verified for signs of past habitation. Although artifacts and features were
discovered during field verification, no definitive evidence of historical logging camps
was found. Factors such as topography and proximity to water sources, among others,
may have been more central to camp location than distance from previous camp sites.
There remain more opportunities to test this average distance theory; however, spatial
modeling using common site characteristics may prove more successful.

Table of Contents
Table of Contents ............................................................................................................... iv
List of Figures .................................................................................................................... vi
List of Tables ..................................................................................................................... ix
Acknowledgements ............................................................................................................. x
Chapter 1 – Introduction ..................................................................................................... 1
Chapter 2 – Literature Review ............................................................................................ 4
Introduction ..................................................................................................................... 4
Section I........................................................................................................................... 6
Logging to Build a Nation ........................................................................................... 6
Logging in Washington State .................................................................................... 11
Section II ....................................................................................................................... 16
Historical Logging Camps ......................................................................................... 16
A Family Affair ......................................................................................................... 19
Ethnic Differences ..................................................................................................... 23
Section III ...................................................................................................................... 27
Locating Historical Logging Camps.......................................................................... 27
Remote Sensing/Spatial Analysis for Archaeological Sites ...................................... 30
Spatial Modeling........................................................................................................ 34
Section IV...................................................................................................................... 42
Cultural Resource Management (CRM) .................................................................... 42
Determining Significance .......................................................................................... 42
Conclusion..................................................................................................................... 45
Chapter 3 – Research Area Background ........................................................................... 46
Introduction ............................................................................................................... 46
Logging Companies in Capitol Forest ....................................................................... 49
Capitol Forest Logging Camps .................................................................................. 51
Chapter 4 – Methods and Analysis ................................................................................... 70
Creating Spatial Data ................................................................................................. 70
Grade Measurements ................................................................................................. 72
iv

Field Verification Process ......................................................................................... 76
Logging Camp Site Identification ............................................................................. 77
Potential Complications and Limiting Factors .......................................................... 86
Chapter 5 – Results and Discussion .................................................................................. 92
Results ....................................................................................................................... 92
Aggregate Segment Findings..................................................................................... 94
Discussion................................................................................................................ 116
Chapter 6 – Conclusion................................................................................................... 121
References ....................................................................................................................... 125
Appendix A – Aggregate Segment Information and Findings ....................................... 132

v

List of Figures
DISCLAIMER: Figures and maps depicting the locations of historical sites or artifacts
are intended for academic purposes only and may not be reproduced without the
author’s consent. These figures and maps are not intended to provide locations for
artifact collectors. Knowingly removing artifacts from historical sites is considered a
misdemeanor; civil penalties and costs necessary to investigate and restore disturbed
archaeological sites can be imposed and artifacts can be seized.
Figure 1 ............................................................................................................................. 14
Figure 2 ............................................................................................................................. 25
Figure 3 ............................................................................................................................. 33
Figure 4 ............................................................................................................................. 38
Figure 5 ............................................................................................................................. 47
Figure 6 ............................................................................................................................. 48
Figure 7 ............................................................................................................................. 50
Figure 8 ............................................................................................................................. 52
Figure 9 ............................................................................................................................. 53
Figure 10 ........................................................................................................................... 54
Figure 11 ........................................................................................................................... 55
Figure 12 ........................................................................................................................... 56
Figure 13 ........................................................................................................................... 57
Figure 14 ........................................................................................................................... 59
Figure 15 ........................................................................................................................... 60
Figure 16 ........................................................................................................................... 61
Figure 17 ........................................................................................................................... 62
Figure 18 ........................................................................................................................... 63
Figure 19 ........................................................................................................................... 63
Figure 20 ........................................................................................................................... 69
Figure 21 ........................................................................................................................... 70
Figure 22 ........................................................................................................................... 71
Figure 23 ........................................................................................................................... 75
Figure 24 ........................................................................................................................... 77
Figure 25 ........................................................................................................................... 79
vi

Figure 26 ........................................................................................................................... 79
Figure 27 ........................................................................................................................... 82
Figure 28 ........................................................................................................................... 83
Figure 29 ........................................................................................................................... 84
Figure 30 ........................................................................................................................... 85
Figure 31 ........................................................................................................................... 87
Figure 32 ........................................................................................................................... 89
Figure 33 ........................................................................................................................... 90
Figure 34 ........................................................................................................................... 94
Figure 35 ........................................................................................................................... 95
Figure 36 ........................................................................................................................... 96
Figure 37 ........................................................................................................................... 97
Figure 38 ........................................................................................................................... 97
Figure 39 ........................................................................................................................... 98
Figure 40 ........................................................................................................................... 99
Figure 41 ......................................................................................................................... 100
Figure 42 ......................................................................................................................... 101
Figure 43 ......................................................................................................................... 102
Figure 44 ......................................................................................................................... 102
Figure 45 ......................................................................................................................... 103
Figure 46 ......................................................................................................................... 104
Figure 47 ......................................................................................................................... 105
Figure 48 ......................................................................................................................... 105
Figure 49 ......................................................................................................................... 106
Figure 50 ......................................................................................................................... 107
Figure 51 ......................................................................................................................... 109
Figure 52 ......................................................................................................................... 109
Figure 53 ......................................................................................................................... 110
Figure 54 ......................................................................................................................... 111
Figure 55 ......................................................................................................................... 112
Figure 56 ......................................................................................................................... 113
vii

Figure 57 ......................................................................................................................... 114
Figure 58 ......................................................................................................................... 114
Figure 59 ......................................................................................................................... 115

viii

List of Tables
Table 1 .............................................................................................................................. 37
Table 2 .............................................................................................................................. 43
Table 3 .............................................................................................................................. 66
Table 4 .............................................................................................................................. 67
Table 5 .............................................................................................................................. 73
Table 6 .............................................................................................................................. 92
Table 7 .............................................................................................................................. 93
Table 8 ............................................................................................................................ 118
Table 9 ............................................................................................................................ 120

ix

Acknowledgements
I would first like to thank my friend and colleague Leland Stilson, without whom
this project would never have become a reality. I am indebted to Lee for his support,
guidance, and professional knowledge, which were invaluable in helping me complete
this idea, which began as a rambling thought in the woods.
I am extremely grateful for the encouragement and patience of my wife, Rhonda,
and children, Payton and Gavin, throughout this process. I love you all very much. Also,
thank you to my parents and extended family for their support during this endeavor.
I would also like to thank my colleagues at the Washington State Department of
Natural Resources for their support during this research project. Specifically, Rolin
Christopherson, Steve Teitzel, Duane Emmons, Tom Shay, Tom Heller, Darin Cramer,
Stephen Slaughter, Maurice Major, Richard Bigley, Julie Sackett, Noelle Nordstrom,
Donelle Mahan, Brule Burkhart, Chris Snyder, Candace Montoya, Derwood Duncan,
Nick Cronquist, and Jon Olson.
A big thank you to my thesis reader, Dr. Kevin Francis, for his positive feedback
and support during my three years in the MES program. Also, thank you to the following
current and former MES faculty members from The Evergreen State College for their
help and guidance during my time at the College; Martha Henderson, Carri LeRoy, Ted
Whitesell, Erin Martin, Greg Stewart, Dina Roberts, and Gail Wootan. Finally, thank
you to my fellow classmates of the 2012, 2013, and 2014 MES cohorts for providing
feedback on this project as well as helping me through the challenges it presented.
Apologies if anyone was missed or not mentioned by name, but know that if I
discussed my thesis with you, your interest and/or support was important and I thank you.

x

Chapter 1 – Introduction
Historical logging camps represent an important period in the resource extraction
history of the United States. As one of America’s first exports, logging provided hope of
prosperity and financial security to thousands of early immigrants and settlers. Many
early logging operations developed into thriving towns, but as the easily accessible
timber was removed, logging companies had to travel farther into the wilderness and
mountains to acquire timber to feed the lumber mills. As railroads became the standard
method of transporting logs to mills in the 1850s, accessing more remote sites became
increasingly easier; however, transporting loggers to these remote sites from towns was
becoming more costly to company owners in relation to work time lost. To address this
issue, logging camps were constructed away from company towns to reduce travel times
to work sites.
Thousands of loggers from the mid-nineteenth through the mid-twentieth
centuries spent time living in these isolated logging camps connected to the rest of
society, including their families, only by railroads. These logging camps can provide
historical context about those people who inhabited the camps and what life may have
been like for them during, possibly, the greatest logging era in America. Studying the
artifacts and features of these sites could provide valuable information to archaeologists,
anthropologists, historical ecologists, historians, environmental historians, genealogists
and more.
Most logging camps were temporary and remain undocumented and unsurveyed.
Identifying historical logging camp locations has typically been completed by researching
historical maps and documents as well as interviews with former camp inhabitants or

1

family members. The number of historical maps and books which depict logging camp
locations is limited and, as time goes on, the number of people who could provide firstor second-hand knowledge of these camp sites becomes smaller and smaller. Therefore,
in the absence of these historical sources, establishing a potential method to better locate
historical logging camps based on patterns would be valuable to archaeologists and land
managers, allowing them to document and protect these historical sites. Can the distance
by rail line between known camps be used to identify the location of additional
undocumented camps? Also, does it matter how distances between camps are calculated:
from the edges of a camp extent, from a central point within a camp, or a combination?
This research aims to identify historical logging camp locations used by the
Mason County Logging Company (MCLC) in the Capitol State Forest near Olympia,
Washington, based on the distance between known and assumed camps along the
railroads that connected them. Assumed camps were inferred based on historical
evidence and by comparing the characteristics existing in the assumed camp area with
those common among the known MCLC camp sites. The reasoning behind using an
average distance is because logging company owners and managers selected camp
locations in advance of operations, they likely determined an approximate distance a
logging camp should be built from the company town or previous camp in order to
maximize production. This approximate distance may have been calculated through
some sort of cost analysis based on one or more of the following factors: how long it
takes to transport workers to a work site; how long it takes to harvest an area based on
topography, timber size (e.g. diameter and height), and logging technology; or how long
it takes to deliver the timber back to the mill in order to keep the mill operating at

2

capacity. Engineers traversing new rail lines through the forest would have used this
approximate distance to identify the most feasible camp site nearest that distance. The
average distance plus or minus one standard deviation provides a range where one might
expect that engineers would have identified the best location(s). It is predicted that other
undocumented MCLC camp sites could be found within this average distance range.
In the case of the MCLC, the distance between logging camps by rail may not
have been a determining factor for logging camp location. Insufficient evidence was
discovered in any calculated distance range to definitively label a site as a logging camp.
Average distance in relation to MCLC camp spacing appears to be coincidental and
logging camp locations may be related to other factors such as topography, proximity to
water resources, technology changes, and land ownership. Also, there was no clear
difference in success among the three methods for calculating distance between camps,
likely due to the small differences between averages. Altering the methods used in this
thesis to determine distance between camps may provide more definitive results;
however, spatial modeling based on common logging camp site characteristics may prove
to be more successful in determining potential historical logging camp locations in the
absence of historical documentation.
There remain more opportunities to test this theory in Washington and other states
where large-scale logging operations occurred. Ideally, this theory should be tested in an
area that was operated on or owned by a single logging company and was a large enough
area to require multiple logging camps. An average distance would vary by logging
company; therefore, two or more logging camp locations need to be known in order to
calculate a potential, company specific, distance between camps.

3

Chapter 2 – Literature Review
Introduction
Although logging is no longer looked upon as favorably as in the past and many
towns and cities once supported by logging have faltered in recent history, it can be
argued that logging is as important to America as any other long-standing industry.
Logging helped construct early America by providing fuel, shelter, and transportation
along with countless other products. Logging provided a glimpse of the American Dream
to thousands of early Americans. The companies who harvested the vast tracts of forest
that once covered much of the United States provided homes to many of those people
chasing financial security. After railroads became the primary mode for transporting logs
from forest to sawmill, areas once inaccessible to logging became available, occasionally
at great distances from existing sawmills (Cox, 2010, pp. 62 & 138). Large landholding
and logging companies built satellite work camps away from main sawmill sites because
the cost of transporting workers to the timber, as well as the valuable work time lost
during transport, outweighed the cost of building the work camps (CALTRANS, 2013).
These work camps, or logging camps, can provide details about the lives of early loggers
who harvested America’s timber.
Federal archaeological guidelines require the identification and protection of sites
with cultural or historical significance (NPS, 2002). Following these federal guidelines,
archaeologists can determine the significance of a site only after obtaining important site
information such as who used a site, when, and for what reason. Locating, documenting,
and, if significant, protecting historical logging camps should be primary goals of
archaeologists as forest managers begin harvesting the second or third growth timber in

4

the forests where early loggers once worked. Locating these historical sites is not always
an easy task since many logging camps were short-lived and often went undocumented
(CALTRANS, 2013). Some historical logging camps have been documented in literature
written by former loggers while other camps, having still been in existence at the time of
drafting, are denoted on historical maps. Other potential methods for locating historical
logging camps include remote sensing and spatial modeling. Management of
archaeological sites can only take place once they are located and varies depending on the
relevant significance of each site.
Section I of this literature review will examine the early history of logging in
America to demonstrate the important role logging played in the founding of the nation.
Attention will be paid to the methods and technology used by early loggers to cut trees
and to transport felled timber to markets. Section II will focus on logging camps and will
include details about the necessity of such camps as well as what life may have been like
for workers inhabiting logging camps in order to detail the significance of logging camps
in the lives of thousands of early American settlers. Section III will introduce
archaeological methods such as remote sensing and spatial modeling used to identify sites
of cultural or historical significance. The literature review will conclude with a look at
Cultural Resource Management (CRM) in Section IV and will touch on methods used by
archaeologists and land managers to protect cultural and historical sites.

5

Section I
Logging to Build a Nation
The history of logging in North America starts well before the founding of the
United States. Literature discussing the history of logging in America can often be
separated into two categories; those that discuss the vast environmental degradation as a
result of irresponsible practices and those that glorify the logging era and the people
involved. Studies of the environmental impacts caused by logging have been ongoing
following the major logging era as problems related to poor management became evident.
The history of logging in regards to the people and the methods they used can be found
primarily in books written during or not long after the major logging era was over.
Books discussing logging history are often written by people who took part in
those early logging operations or who were fascinated by early loggers and the ingenuity
they showed providing timber to the nation with limited, and sometimes unsophisticated,
technologies. The trio of books written by Ralph Andrews: This was Logging!, Glory
Days of Logging, and Timber: Toil and Trouble in the Big Woods, provide nearly every
detail of the inner workings of life and labor in the logging towns and camps of western
North America during the late nineteenth and early twentieth centuries. Andrews (1954,
1956, & 1968) speaks fondly about the people he met during his time in logging camps,
showing the regard he had for them and the job they did.
There have been numerous books providing insight into the great logging era in
the Pacific Northwest. These books include Railroads in the Woods (Labbe & Goe,
1961), When Timber Stood Tall (Pierre, 1979), Capitol Forest: the Forest That Came
Back (Felt, 1975), and Logging Railroads in Skagit County (Thompson, 1989). Books

6

written primarily about single logging companies also provide great details about logging
history and the people involved. These books include The Pine Tree Express
(Henderson, 1990) about the Cascade Lumber Company in eastern Washington, The
Oregon-American Lumber Company: Ain’t No More (Kamholz et al., 2003), Time, Tide
and Timber: A Century of Pope and Talbot (Coman & Gibbs, 1949), and Family Trees,
Simpson's Centennial Story (Spector, 1990) about the Simpson Logging Company in
western Washington. There are few sources, however, that attempt to cover both
environmental impacts and the historical aspects of logging in America. One such work,
Thomas Cox’s (2010) book The Lumberman’s Frontier, provides an objective and
comprehensive history of logging in the United States from pre-colonial times through
the early twentieth century while focusing on the people and their practices.
Timber was the main source of fuel and a common material for structures and
goods as well as a prime source of material for Royal Navy ships; however, it was not the
vast acres of forest that drew settlers to North America (Cox, 2010, pp. 1-3). Many earlyAmerican farmers wanted to model their life in The New World after what they knew in
Europe; open land free of forest except for scattered trees, an “agricultural society” (Cox,
2010, p. 1). Incoming settlers cleared much of the forests of North America to create
farms because they viewed the forests as an impediment to successful agriculture;
however, not all farms were successful in these early times. Many farmers settled away
from larger towns, which allowed them few opportunities to trade their goods for other
necessities (Cox, 2010, pp. 1-11). For this reason, many farmers began to utilize the
timber in and around their land to supplant income not gained from agriculture. Farmers
could sell their timber, for money or goods, to other settlers or to budding towns where

7

they needed wood to supply builders and craftsman (Andrews, 1954, p. 78; Labbe & Goe,
1961, p. 9; Cox, 2010, pp. 1-11).
Although colonists mainly viewed the never-ending forests of North America as a
hindrance to the life they wanted to create for themselves, the value of the timber in those
forests quickly shifted the focus of many early settlers from agriculture to logging. In the
seventeenth century, demand for wood products in North America and England became
so great that settlers began constructing sawmills in great numbers all over the east coast
of North America (Cox, 2010, pp. 23-9). Early sawmills first supported only those who
harvested and milled the timber along with their families, but as demand and production
of lumber increased, so too did the needs of the mill workers. Shops and manufacturers
met these demands by providing essential items to families supported by sawmills. The
introduction of commercial goods and services around thriving sawmills created towns
and new markets, which, in turn, drew families from other areas in search of opportunity
(Coman & Gibbs 1949, pp. 163-73; Cox, 2010, pp. 23-9).
As sawmills increased production and profits, and as the towns being erected
around sawmills grew, so too did the demand for commercial timber. Early loggers
moved through the forests of eastern North America to satisfy the needs of the flourishing
timber industry quickly and with little care for anything other than profit (Cox, 2010, p.
28). Logging operations removed forests in close proximity to sawmills and, by the early
eighteenth century, the logging machine moved into the interior forests along the East
Coast.
Technology required to get logs to mills changed rapidly during this time. A
common practice until the late nineteenth century, horses and oxen provided the power to

8

pull the logs once cut (Andrews, 1956, p. 64 & 1968, pp. 69-70 & 85; Labbe & Goe,
1961, p. 9; Cox, 2010, pp. 12-5, 57-72). To make moving them easier, loggers stockpiled
logs until the winter months when they could be sledded over frozen ground (Cox, 2010,
pp. 12-5). The abundance of timber near large rivers allowed logs to be floated to mills
located along those rivers or their tributaries (Andrews, 1956, pp. 130-2). This method
included the use of splash dams where a build-up of logs and water are held behind a dam
typically constructed from logs. When the time came to transport logs to mills, splash
dams were removed, allowing the excess water once impounded by those dams to carry
the logs downstream. Large amounts of cut timber could be kept until needed at the mills
by organizing them into collections of floating logs or ‘booms’ (Labbe & Goe, 1961, pp.
9-10; Cox, 2010, pp. 12-5 & 57-72).
Getting timber to rivers became increasingly difficult as loggers removed the
more accessible forests; they needed new technologies to access forests located further
from the rivers and streams used to transport the timber (Spector, 1990, p. XV).
Railroads allowed the transportation of cut timber from the most remote forests and
initiated the greatest logging era in United States’ history (Labbe & Goe, 1961, pp. 5 &
9-10; Cox, 2010, pp. 62 & 138). Logging operations supported by railroads moved even
more swiftly through America’s vast forests. The logging practices and methods used to
transport timber to mills had major impacts on the landscape of the United States. The
construction of logging railroads left miles of scars across once forested hills where
workers filled in areas to make level grades and blasted or cut through hills to make way
for rail lines (Labbe & Goe, 1961, p. 29). Rivers had their normal processes disturbed or
altered by the dams used to build up water supplies for mills and to provide adequate

9

transport for logs. The construction of canals to connect major water ways in order to
transport logs to specific mills and markets left permanent reminders of past logging
activities (Cox, 2010, pp. 57-63 & 88-91).
The California Department of Transportation (CALTRANS, 2013) recognizes
technology as being the major factor influencing costs of labor for early logging. Horses
and oxen required fewer workers than those operations using rivers to transport timber.
When the common method for transporting timber to rivers and sawmills shifted to
railroads in the mid-nineteenth century, the need for a “larger and more highly skilled
workforce” became apparent (Labbe & Goe, 1961, pp. 5, 9, & 29; CALTRANS, 2013, p.
96). The industry now required engineers to design rail lines and more workers to
construct the miles of rail lines throughout the forests. These workers commonly
occupied construction camps located in remote locations. Construction camps were often
temporary, moving along with the progression of construction (Labbe & Goe, 1961, pp.
29 & 149; CALTRANS, 2013).
As logging operations moved further from mill sites, the costs to transport loggers
to timber stands increased as did production time lost due to that transportation
(CALTRANS, 2013). The high costs of constructing railroads in the forest – Cox (2010)
states mainline construction could cost as much as $50,000 per mile (p. 317) – needed to
be balanced by finding measures to save costs and increase production. This led to the
construction of satellite logging camps, which reduced the amount of potential work time
lost from transporting loggers long distances. Camp locations were selected before
logging operations began as part of log transportation planning to ensure a profit could be

10

made (Andrews, 1954, p. 74). This fact could point to there being a more systematic
approach for choosing camp locations.
Logging practices that began along the east coast of the United States eventually
spread to the Great Lakes states, the southern United States, and finally to the West Coast
(Labbe & Goe, 1961, p. 5; Robbins, 1985; Cox, 2010, pp. 125-89 & 213-89). Common
themes arose in every location, beginning with removal of trees for agriculture; however,
the main drivers of the industry became timber speculation and utilization. Companies
and entrepreneurs from states such as Maine and Minnesota hired people to find untapped
stands of timber near burgeoning markets where greater profits could be made (Cox,
2010, pp. 125-89 & 213-89). With its massive timber and coastal access, Washington
State came to the forefront of the logging industry in the nineteenth century (Labbe &
Goe, 1961, p. 5).
Logging in Washington State
By the time major logging began in Washington State, railroad logging had
become the primary extraction method; however, some small landowners still used horse
and oxen to transport timber (Labbe & Goe, 1961, pp. 5 & 9). With less favorable
topography and fewer navigable rivers, delivering logs by river may not have been as
common in many areas of Washington. Puget Sound, however, provided a suitable
means for transporting logs to mills (Labbe & Goe, 1961, p. 211; Felt, 1975, p. 26). In
reviewing maps showing historical logging rail lines, logging railroads appear to be
constructed along many of the major stream channels in Washington (Thompson, 1989;
Henderson, 1990; Hannum & Hannum, 2002 & 2006). These areas often provided the
gentle slopes required for adequate train movement; railroads require sustained slopes of

11

three to five percent. This slope limitation not only restricted the amount of available
routes through a forest, but, consequently, limited the number of sites available for
logging camps. Rail lines along streams also provided access to a water source, a
necessity for workers and operations.
Many small logging companies operated in a single drainage because the cost of
land claims was high and many of the larger forested areas had already been purchased or
claimed by other timber speculators (Robbins, 1985). These small companies delivered
logs to privately-owned mills and may have only had a single logging camp as the base of
operations. Small companies with one camp, or no camp at all, would be difficult to
locate as there may be no recordings of such small operations. Research on these smaller
companies and their associated camps has been limited mainly because of this lack of
evidence. Instead researchers have focused primarily on the “large and midscale”
operations because of the abundance of evidence left behind on maps, in company
records, and remnants on-site (CALTRANS, 2013, p. 98).
Companies that owned or had rights to operate on large acreages constructed
sawmills in close proximity to those lands with rail lines spreading throughout their
surrounding landholdings (Carlson, 2003, pp. 6-15). Large rail networks were often only
financially feasible to large-scale logging companies (Cox, 2010, p. 317). Similar to
developments along the East Coast and upper Midwest during the early years of
mainstream logging in the United States, major sawmill locations became industrial town
sites (Carlson, 2003, pp. 6-15; Cox, 2010, p. 285). Company towns grew to
accommodate employees, their families, and the commerce needed to supply its residents.
Mills, being in set locations, also required permanent living quarters for workers, leading

12

to companies constructing towns with enduring structures (Carlson, 2003, pp. 6-15; Cox,
2010, p. 285). Having exhausted timber in close proximity to sawmills and company
towns, costs of transporting loggers to work sites became increasingly more expensive
than constructing satellite work camps (CALTRANS, 2013).
Depending on the logging company and location, logging camps may have been
constructed with more permanence in mind, containing more structures with concrete
foundations. More often than not, these camps had to be moved as quickly as work
progressed through the forest (CALTRANS, 2013). Because of the ephemeral nature of
logging camps, their construction allowed for quick removal and transportation to another
location (Figure 1; Labbe & Goe, 1961, pp. 29 & 149). In fact, Labbe & Goe (1961)
state structures used to house the bachelors in logging camps had to be built on “runners”
or skids to facilitate this rapid movement (p. 149). Kitchens could also be built on skids
or railcars so they could be loaded onto trains and easily moved. Temporary logging
camps may have only existed in a given location for one to three years and many factors
likely influenced the duration of camp use, including the amount and size of timber as
well as the feasibility of removing it (Cox, 2010, p. 284). More research can be done
regarding the duration a camp existed in a specific location, which may also help predict
locations of undocumented camp sites.

13

Figure 1. Example of a logging camp bunkhouse on a railcar. Having moveable
structures allowed for quick movement of logging camps to new locations. Photograph
by Clark Kinsey (UWL, 2015).

Although temporary, these camps still had to accommodate the needs of the
workers for varying amounts of time. Other than a kitchen providing sustenance for the
workers, many goods needed for everyday life had to be obtained from company stores
(Ayers, 1996; Carlson, 2003, pp. 6-15). This meant workers had to ride the trains back to
town or hike back through the forest (Pierre, 1979, pp. 56-64 & 79-80). Dishes from
kitchens and bottles once containing condiments or personal use items were indicative of
the temporary nature of camps and were commonly discarded at camp sites once they
broke or had served their use.
In the 1920s, a shift in technology to logging trucks allowed workers to commute
to work, but the shift took time to develop into common practice and railroad logging
continued to be the main method for log transportation in the West until the 1940s (Cox,
2010, p. 314; CALTRANS, 2013). Logging camps would no longer be needed as
14

transportation costs fell and technologies such as the chain saw improved tree removal
rates. As the amount of available old-growth timber slowly fell to the saws and axes of
the twentieth century, so too did the number of railroads and workers needed to harvest
the trees (Robbins, 1985). All that remains are the scars on the landscape and remnants
of the wares left by the people who spent their lives working in the forests during a great
period in the history of the United States. Locating former logging camps and analyzing
what remains could provide great insight into the lives of former inhabitants. This
research project attempts to identify a method to locate more of these sites in what is now
known as the Capitol State Forest, Washington, United States.

15

Section II
Historical Logging Camps
To be eligible for listing in the National Register of Historic Places (NRHP), a
site must contain information that can contribute to the overall understanding of a
specific aspect of human history. The specific criteria for a site to be eligible for the
National Register are (NPS, 2002, p. 2):
The quality of significance in American history, architecture, archeology,
engineering, and culture is present in districts, sites, buildings, structures, and
objects that possess integrity of location, design, setting, materials, workmanship,
feeling, and association,
and:
A. That are associated with events that have made a significant contribution to the
broad patterns of our history; or
B. That are associated with the lives of persons significant in our past; or
C. That embody the distinctive characteristics of a type, period, or method of
construction, or that represent the work of a master, or that possess high
artistic values, or that represent a significant and distinguishable entity whose
components may lack individual distinction; or
D. That have yielded, or may be likely to yield, information important in
prehistory or history.
It could be argued that historical logging camps fit all criteria except criteria B. Logging
camps are representative of a very finite, but important time period in the history of the
United States. Identifying and locating these sites can improve our understanding of the
living conditions and lives of these early-American forest workers. Judge et al. (1988)
notes archaeologists have been solely focused on identifying sites and the artifacts found,
spending little time working out possible reasoning for site locations and what life may
have been like for the inhabitants of those sites. Locating sites and describing remnants
and artifacts are a necessary facet of archaeology; however, determining the overall
context of a site will better “contribute to [the] scientific understanding” of
archaeological sites (Judge et al., 1988, p 3).
16

Work camps such as logging camps offer archaeologists an opportunity to analyze
“discrete data…related to ethnicity, assimilation, acculturation, rural life, immigration,
labor, and socioeconomics” and attempt to answer questions about the daily lives of camp
residents (CALTRANS, 2013, p. 11). This section will cover some dynamics of camp
life and include findings from research related to each factor discussed for the purpose of
demonstrating the significant historical context logging camps represent. Much of the
research in relation to logging camps has focused on the social aspects, including the
inclusion of workers’ families, differences in ethnic backgrounds, and the general
lifestyle in rural work camps.
Some of the most common data related to logging and other work camps discuss
the workers, typically men, living in the camps. Ayres (1996) details information about
work camps of the Standard Timber Company who operated in Utah. The report
discusses the types of workers inhabiting camps as well as the inclusion of women, and
possibly children, in some camps. Ayres (1996) mentions that two types of men worked
in the camps: professional and amateurs. Amateurs were often farmers needing to
substitute their income during the “winter months when their agricultural responsibilities”
had been reduced (p. 180). The Standard Timber Company preferred professional
loggers because they “clearly out-produced” the seasonal loggers (p. 180). Ayers (1996)
also notes that Swedish immigrants comprised the majority of the professional loggers
and represented the “largest ethnic group” of all the workers (p.180). Ayers (1996) finds
that the ‘Swedes’ typically worked longer hours than other workers, which could point to
either their financial needs or work ethic (p. 180).

17

A common theme in much research is that single men in need of work comprised
much of the logging work force, some of these men were former soldiers returning from
overseas (Andrews, 1968, p. 61). Many bachelors enjoyed the lifestyle of a logger,
working simply for money to spend at the closest tavern on alcohol, women, or both
(Andrews, 1954, pp. 53 & 101 & 1968, p. 55). Andrews (1954) also notes many bachelor
loggers were as “touchy as prima donnas,” quick to leave camp if the quality of food
declined or after the first instance of mistreatment (p. 53). Because of this lifestyle, some
managers thought of single male laborers as unreliable loners, only working to fulfill a
need (Ayres, 1996; Carlson, 2003, p. 11; CALTRANS, 2013).
Camp managers and company owners sometimes attempted to curtail the drinking
aspect of the logger lifestyle by not allowing liquor to be sold in the company stores
(Carlson, 2003, pp. 10-1). Company owners attempted to keep camps as dry as possible
by not allowing saloons to be constructed or liquor to be sold in town. As Carlson (2003)
notes, “dry camps” typically had fewer occurrences of fights and also resulted in fewer
accidents (p. 11). Not all companies tried to limit alcohol sales in town; some company
owners allowed drinking based on the understanding that activities such as drinking kept
the single workers happy and coming back to work (Carlson, 2003, p. 11). Regardless if
a company allowed liquor to be sold in their town or not, rarely were locations for
purchasing alcohol far enough away to deter loggers from finding a drink.
Hiring single men may have been a necessity based on the dangerous type of
work involved, work for which a family man might not be willing to risk his life.
Because of the view that single, often transient, workers were unreliable loners, some
logging companies specifically hired married men. Other workers traveled with their

18

families simply because they had nowhere else to go or had emigrated from a distant
location. The inclusion of workers’ families adds another aspect relating to the overall
significance of historical logging camp life.
A Family Affair
As discussed above, loggers of the era tended to be single men trying to earn a
wage. A fair amount of people going to work in the forests of the West came from other
states in an attempt to make a better life for their family (Rohe, 1994); this often entailed
traveling with their families to work sites and settling in company towns. Major mill and
company town sites made accommodations for families, but satellite work camps mainly
remained free of women and children. Exceptions existed and Ayres (1996) notes that
the Standard Logging Company listed a total of 20 women living in seven different work
camps.
Some logging companies specifically hired married men for multiple reasons; the
main reason being that companies viewed married men as more dependable because they
had to ensure steady employment to provide for their families (CALTRANS, 2013). The
inclusion of families in camps could be dependent on job status as well. Maniery (1996,
as cited by CALTRANS, 2013) notes workers such as managers had their families with
them, but the common workers did not. An account written by George Woodward (1894)
confirms this stating, by rule, the foreman was a married man and his family occupied a
home in the camp. CALTRANS (2013) mentions that proper accommodations remained
limited in camps, especially satellite work camps, leading to the exclusion of families.
Some companies may have encouraged hiring men with families because the managers
considered them “less mobile” and because the women and children could provide free or

19

relatively cheap labor for the camps (Brashler, 1991; CALTRANS, 2013, p. 66).
Andrews (1968) mentions that skid greasing, where grease was applied to the logs on the
ground over which harvested timber was pulled, was a task for "boys exclusively,"
sometimes as young as 14 (p. 64).
In reviews of other sources such as Brashler (1991) and CALTRANS (2013),
women played major roles in everyday camp life. Women usually handled the “domestic
economy” in camps and evidence found demonstrates that wives made mindful decisions
about certain household needs by paying close attention to family budgets (CALTRANS,
2013). In reviews of historical photographs, women appear to be common fixtures in the
kitchens associated with mill sites. Women also had a need to provide for their family,
but their labor often went unpaid. Responsible for upkeep of the home, women also
prepared meals for their husbands and sometimes other men who boarded with the family
(Brashler, 1991; CALTRANS, 2013). Women and children in camps became more
prominent during World War II when the government removed restrictions allowing
logging companies to hire women and boys in high school. Some women earned
minimum wage based solely on managers’ disapproval of having women involved in
logging operations. Logging operators based this low pay on the skills and strengths of
the women, but perceived levels of skill and strength seem to be based on gender biases
(Kamholz et al., 2003, pp. 215-16).
As more workers began bringing their families with them to company town sites,
companies had “little choice” but to build infrastructure to support the families, including
the building of schools to teach the children (Cox, 2010, p. 285). Ayres (1996) notes that
the Standard Timber Company operated a school for the children living in the main camp

20

and many other companies such as the Oregon-American Lumber Company did the same
(Kamholz et al., 2003, p. 75). The Mason County Logging Company (MCLC) also built
a school within the mill town of Bordeaux in 1903 (OAHP, 1985). The school in
Bordeaux remained in operation until the mill, as well as the town, shut down in 1941
when the company had exhausted its timber resources in the Capitol State Forest (Felt,
1975, p. 32; OAHP, 1985).
Being isolated deep in the forest, logging camp occupants had little contact with
the outside world. In addition to schools, some companies owned and operated stores
and blacksmith shops. Company stores provided household goods and necessities, but at
a cost. Stores typically provided food, clothing, hardware, and other personal items, but
companies viewed stores as more a form of profit for the company than convenience for
camp residents (Ayres, 1996; CALTRANS, 2013). Company stores often dealt in store
credit based on a worker’s production and pay, and, because of the isolated locations,
workers and their families had “little choice but to patronize” the company store (p. 182).
Unfortunately for the families, prices at company stores tended to be greatly inflated,
something the companies could also get away with because of the isolation of the camps
(CALTRANS, 2013).
Families of MCLC workers lived either in the mill town of Bordeaux or in the
Hollywood family camp. Field surveys of both Bordeaux and the Hollywood family
camp found numerous household items not commonly found at satellite work camps
within Capitol Forest (Boire & Stilson, 2006; Ferguson 2011A). Some of these items
included colorful ceramic wares and decorative ceramics from countries such as
Czechoslovakia, France, Germany, and Japan (Boire & Stilson, 2006; Ferguson, 2011A).

21

Other artifacts discovered at the Hollywood site included women’s and children’s shoes
and a small wheel, possibly from a tricycle (Ferguson, 2011A). The Mud Bay Logging
Company also had women and children living in logging camps (Felt, 1975, p. 35). Felt
(1975) writes that Mud Bay Camp 2 contained flowers such as daffodils and lilacs where
a woman attempted to “bring civilization to a primitive wilderness” (p. 35). Field
surveys of Mud Bay Camp 2 also uncovered a piece of a porcelain doll, indicating the
possible presence of a child (Stilson, 2010A).
Amenities constructed in company towns made life more comfortable and
appealing to the families of the workers. Companies occasionally constructed churches
in town, sometimes even determining the church’s denomination (Carlson, 2003, pp. 8 &
11). Companies provided locations for recreation within the town, including recreation
halls or baseball diamonds (Carlson, 2003, pp. 6 and 11). Some company towns even
had a theater to provide entertainment for workers and their families (Carlson, 2013, pp.
8 & 11). Recreation in the satellite camps remained simple, usually involving card games
such as poker (Andrews, 1954, p. 26 & 1968, p. 56).
The presence of families in work camps provides another layer to the significance
of life in such camps. Analysis of women and children in camps could further
demonstrate their importance as well as the significant role they played in the function of
logging camps not only in the industry as a whole, but in different locations as well.
Research on loggers can also provide similar details about their life. For most of these
workers, everyday life followed common themes. Brashler (1991) points out the
historical context of family camps in different states such as the Great Lakes states and
Appalachian states can vary. This may also be true of these types of camp sites in

22

Washington State as life may have been different for the early logging families of
Washington compared to other locales. Some of these differences can be related to the
many cultural or ethnic differences.
Ethnic Differences
Logging camps of the late nineteenth and early twentieth centuries became homes
to thousands of workers representing a myriad of ethnic backgrounds. There is a growing
amount of literature on the topic of ethnicity in relation to logging camps. Identifying
more sites for examination will only further the understanding of what role ethnicity
played among camps and companies. CALTRANS (2013) points out work camps
attracted many immigrants because wages in such camps were much higher than what
could be earned in their home country. Employers turned to native tribesmen, farmers,
miners, “sailors who had jumped ship,” transients, soldiers, and immigrants for laborers
in the woods (Andrews, 1968, pp. 55 & 61; Cox, 2010, p. 285). In a discussion of
varying ethnic work forces in the Great Lakes area, Rohe (1994) finds that workers from
other countries and states, often states along the east coast of the United States where the
logging era was coming to an end, comprised the majority of camp residents (Rohe,
1994). In the case of the Great Lakes states, immigrants mainly came from Canada
(Rohe, 1994), but other locations included immigrants from Sweden, Finland, China,
Germany, New Zealand, and Ireland, to name a few (Andrews, 1954, p. 59 & 1968, p. 55;
Franzen, 1992; Rohe, 1994; CALTRANS, 2013).
Franzen (1992) researches logging camps in Michigan, examining diet and
ethnicities in the camps. The author finds, in relation to ethnicity, workers tended to
retain family and ethnic ties as well as materials significant to their specific cultures

23

(Paullin, 2007). ‘Finns,’ a major immigrant group in early-Michigan logging, formed
their own “religious, literary, and socialist clubs” (Franzen, 1992, p. 84). Finnish
immigrants became major players in the social reform of Great Lakes work camps,
leading towards a better organized and educated workforce in regards to working
conditions and pay (Hoglund, 1960, as cited by Franzen, 1992). Immigrants from
Finland helped start camp reform because they required good food and sanitation in their
camps. Finnish immigrants also imported the idea of steam baths into logging camps,
something Hoglund (1960, as cited by Franzen, 1992) notes was a “distinctive transfer of
folkways” from their home country (p. 24). Varying ethnicities in work camps also
resulted in different foods being supplied at the camps. Finnish immigrants included
“fish and meat stews” more often as part of their diet (Franzen, 1992, p. 84). The
inclusion of traditional elements, such as saunas and ethnic recipes, demonstrates how
immigrants tended to “influence their new environment” with parts of their culture
(Franzen, 1992, p. 94). Finnish immigrants are believed to have made up a significant
portion of the logging workforce in southwestern Washington as well.
References to singular ethnic camps can occasionally be found in historical books
and maps. The Oregon-American Logging Company hired Japanese workers to construct
railroads, but segregated the workers into their own camp; segregating minorities into
their own, often isolated, camps was a common practice during this time period (Carlson,
2003, pp. 14-5). The ‘Jap’ camp, approximately a quarter mile from the main logging
camp, still included all the amenities of the main camp (Kamholz et al., 2003, p. 75).
Japanese workers participated in logging operations for the Oregon-American Logging

24

Company until Pearl Harbor was bombed, at which time the Japanese workers were sent
to internment camps (Kamholz et al., 2003, p. 215).
In Washington, Japanese workers were often employed to construct and maintain
railroads. The Rock Creek Lumber Company, later renamed Walville Lumber Company,
owned land and a company town named Walville in Lewis and Pacific counties. The
Walville Lumber Company employed many Japanese workers, listing 74 in 1909 alone
(Stilson, 2004). Japanese workers and their families lived in a section of the town named
‘Jap Town’ which also had its own cemetery (Stilson, 2004). Walville’s Japanese
residents introduced part of their culture to other residents of Walville, taking part in
Sumo wrestling tournaments (Figure 2). One satellite camp of the Walville operation, a
tunnel construction camp, was also found to have been occupied by Asian workers
(Stilson, 2010B).

Figure 2. Image from Stilson (2004) showing Walville’s Sumo

wrestling tournament in 1910. The author notes long johns may have
been worn so as not to offend women in attendance.

25

Similarly to immigration being a key factor in the founding and development of
the United States, immigrant workers also shaped the life and working conditions of early
logging. This included incorporating traditional foods and technologies as well as
working towards better pay and living conditions (Hoglund, 1960, as cited by Franzen,
1992). Logging camps remain a significant source of data relating to immigrant labor
forces during the major logging era of the nineteenth and early twentieth centuries
(Franzen, 1992; Paullin, 2007). With the exception of the MCLC Hollywood family
camp where ceramic dishware pieces were found originating from Czechoslovakia,
Germany, and Japan, no evidence was found during research related to worker ethnicity
in MCLC logging camps (Ferguson, 2011A). As mentioned above, other companies in
Washington State had segregated camps for specific ethnic groups; therefore, locating
additional logging camp sites could add to the growing body of evidence regarding the
assimilation of immigrant workers to the culture of logging in America as well as
American life.

26

Section III
Locating Historical Logging Camps
As discussed in Section I, logging and railroad work camps tended to be
temporary and could be moved as quickly as the work progressed (CALTRANS, 2013).
Due to the temporary nature of logging camp sites, numerous camp locations likely
remain undocumented. Up to now, no literature or method exists to predict the location
of these undocumented historical sites. There are sources from which a logging camp
site can be identified, including historical maps, books written either about the logging
industry or about specific logging companies, and personal accounts from the people who
worked and lived in these camps. Books written about specific logging companies
provide some details about the loggers and the camps they lived in. Books about the
history of the Oregon-American Logging Company (Kamholz et al., 2003) and the
Simpson Logging Company (Spector, 1990), are two such books written specifically
about the history of each company. These books, however, focus mainly on the overall
logging operations and the major players involved with the companies rather than on
specific goings on in the work camps or those who inhabited them.
Books related to specific companies often contain maps of the areas worked,
displaying the location of many if not all of the work camps used, occasionally with dates
of usage. Maps found in Dennis Blake Thompson’s book Logging Railroads in Skagit
County (1989) and Eugene Henderson’s The Pine Tree Express (1990) depict logging
camp locations and, in many cases, the dates each camp existed. Camp locations in these
books are valuable pieces of information as they can lead to the locating and surveying of
those sites; however, map and symbol scales can make it difficult to find exact locations.

27

For instance, the location of a camp on a large scale map may appear to cover a much
larger area than it does on the ground, leading to a much larger survey area. Dates
associated with camp locations can be especially important in determining other factors
involved with those work camps. Dates of usage can help approximate the amount of
time it took the loggers in those camps to harvest a given area near the camp based on the
known technologies of that time. It is unclear if research related to this theory has been
undertaken.
There are a number of resources available containing first-hand accounts from
people who inhabited work camps of the late nineteenth and early twentieth centuries.
Many works choose to spend more time discussing logging and railroading operations,
but some talk specifically about life in the work camps. One such account, written by
George Woodward in 1894, discusses the everyday life of loggers in great detail.
Woodward (1894) provides details about the different buildings often present at historical
work camps, including kitchens and bunkhouses. A good portion of his account relates to
the type of people, typically men, who lived in logging camps and the customs they
adhered to (Andrews, 1968, pp. 56 & 76). Hazing was, at one time, a common practice
in work camps along with theft. The author notes occasions where socks would be stolen
off the feet of sleeping loggers (Woodward, 1894). There are also interesting facts about
recreation in work camps and the etiquette, or lack thereof, practiced by the workers
(Andrews, 1968, pp. 56 & 76). These types of accounts are important for discussing the
significance of logging camps, but they mainly focus on everyday life and relate little
information as to exact locations of a camp.

28

Field surveys remain the main method for locating historical camp sites. These
methods rely on local knowledge or begin by following a railroad grade or other clue
such as the location of a camp on a historical map. Even if a site is displayed on a
historical map or some other resource, it remains a simple geographical point. More
analysis can be done through site surveys to discern site dimensions. One method
involves mapping the occurrences and locations of artifacts or manipulated landscapes
(e.g. artificially flattened areas). Another method, attempted by Paullin (2007), to
determine the size and dimensions of logging camp sites could possibly be done through
dendrochronology, a method of studying the age of trees found on and around a known
logging camp site. During their use, camp sites are kept clear of trees while the harvested
areas surrounding them would be left to regenerate naturally. Once abandoned, trees
reclaim the camp sites, but these trees would be younger than the surrounding areas
where regeneration had occurred earlier. Based on this, Paullin (2007) tests whether the
dimensions of a logging camp site could be determined by an age difference in the trees
in and adjacent to the camp; however, Paullin (2007) could not prove this age difference
theory for the research area.
The outcome of Paullin (2007) seems logical since many satellite work camps
commonly existed for only one to three years. It is common to have a varying age
distribution in naturally regenerated forest stands, making an identifiable line where trees
are only a few years younger than the rest of a mixed stand nearly impossible to discern.
Determining the extent of a camp site may still be best completed by analyzing
modifications in the landscape in concert with artifact accumulations; however, artifact
accumulations, or middens, may not be located within the footprint of a camp site.

29

Schiffer (1983 & 1986) outlines site formation processes, or actions that alter an
archaeological site after its establishment, and methods to decipher and understand those
processes. The author’s points on the positioning of artifacts displays sound logic,
specifically noting that large accumulations of artifacts are not always an indication of a
site; rather, the location of artifacts could simply be a dump site (Schiffer, 1983 & 1986).
Researchers must consider the ideas presented by Schiffer (1983 & 1986) in relation to
artifact accumulations and the reasoning behind their formation when attempting to
locate historical sites.
Luckily for archaeologists and cultural resource managers, much information has
been recorded about locations and dates of usage for sites such as logging camps;
however, there remain numerous undocumented historical logging camp sites.
Archaeologists and cultural resource managers have developed new methods for locating
unknown cultural and historical sites. Remote sensing has been a useful tool for
archaeologists for a number of years and analytical tools such as Geographic Information
Systems (GIS) allow archaeologists to implement spatial statistical models where
multiple variables are compared in attempts to predict possible site locations. Spatial
modeling has become an increasingly valuable tool for identifying possible site locations
when there are no historical records available.
Remote Sensing/Spatial Analysis for Archaeological Sites
Remote sensing, a way of collecting information about a location without
physically visiting that location, has been in use by archaeologists since the 1930s after
the first aerial photographs were taken of the United States landscape (Lasaponara &
Masini, 2013). Since remote sensing involves analyzing information from afar, it could

30

be argued that archaeologists studying historical maps have been practicing remote
sensing techniques well before the 1930s. First used in the 1960s, the term remote
sensing provided a name for the “unified technical field” of data collection methods
scientists had been using to identify not only archaeological sites, but also environmental
information (Judge et al., 1988, p. 430).
Data for interpretation can be obtained from more than aerial photographs and
historical maps. New technologies emerging in the 1950s, such as infrared imagery,
direct current resistivity, and magnetometry provided archaeologists an abundance of
new information to better interpret and identify archaeological sites (Lasaponara &
Masini, 2013). Becoming available in the 1980s, satellite imagery greatly added to the
sources available for interpretation and further improved the remote sensing capabilities
of archaeologists (Lasaponara & Masini, 2013). Image resolution improvements and the
development of digital terrain models from Light Detection and Ranging (LiDAR)
imaging have occurred in the past 10 years, providing even more interpretable data
(Lasaponara & Masini, 2013). The collection of maps and imagery, especially in the past
century, allows archaeologists to utilize every source possible to remotely measure and
interpret data in order to locate sites and to identify site patterns.
Remotely sensed data are obtained by human interpreters who employ different
methods to analyze various sources. The simplest method is to look at maps and images
to identify sites. For instance, a historical map depicting the location of a site from 1930
can be compared with an aerial photograph from 1940 to check for any physical evidence
of a site. Scale can be a common problem of aerial photography, especially satellite
imagery, since these types of images are taken far from the area of interest.

31

Magnification is often needed in order to view small details in these circumstances.
Besides issues of scale, it can also be difficult to remotely survey for archaeological sites
due to vegetation depicted on aerial imagery. A source of information that helps combat
this limitation is LiDAR (Lasaponara & Masini, 2013). LiDAR uses laser pulses
between the laser equipment and the surface of Earth, thus eliminating the issue of
vegetation. LiDAR provides a fantastic tool for the identification of archaeological sites
because manipulated landscapes are easily visible, even to the untrained observer.
Recently, an abundance of research involving LiDAR has occurred. This research
includes extrapolating tree heights and other forest stand characteristics, but LiDAR has
also been used to locate cultural sites. A study conducted by Hare et al. (2014) used
LiDAR to map structures and features associated with the prehistoric city of Mayapán in
Yucatan, Mexico, a site that had been the subject of numerous archaeologic studies.
Using high-density LiDAR, the researchers mapped even the smallest features, including
benches (Figure 3; Hare et al., 2014). The level of resolution allowed researchers to
easily identify structural features and small scale landscape manipulations in comparison
to the surrounding landscape features. Results are only preliminary and more data has
yet to be analyzed, but at the time of publishing, Hare et al. (2014) had identified 3,429
new features in the research area.

32

Figure 3. Example of results from the LiDAR analysis completed by Hare et al. (2014).

The research completed by Hare et al. (2014) still had limitations based on the
quality of current LiDAR. Field verifications of the site determined multiple features
with low-relief could not be identified from the LiDAR data used (Hare et al., 2014).
Literature and research utilizing LiDAR for archaeological purposes is still in its infancy,
but the possibilities are great. As the data resolution improves, analyses can be
completed to determine smaller landscape manipulations with greater accuracy in order
to better identify prehistoric and historical sites, including logging camps.
Many of the available map and photograph sources are now accessible digitally
for manipulation and analysis in computer programs. GIS allows for the “collection,
storage, retrieval, manipulation, and display of spatial data” and has become an
invaluable tool for archaeologists (Ebert, 2004, p. 319). Ebert (2004) lists three
33

hierarchal levels applied by archaeologists in regards to GIS: visualization, management,
and analysis (p. 320). Visualization, as considered by Ebert (2004), is the “lowest level”
of GIS application and does not help archaeologists produce theories as much as it creates
aesthetically pleasing images (p. 320). Management simply refers to data management;
cultural resource managers and archaeologists utilize this level of GIS usage to track and
manage site locations (Ebert, 2004). Both visualization and management do not utilize
the “full analytical capabilities of GIS” whereas the third level, analysis, is the best
method for developing hypothetical theories (Ebert, 2004, p. 320). Although the use of
GIS in archaeological work is increasing, analysis remains the least used level of GIS
application (Ebert, 2004). In attempts to predict or locate unknown archaeological sites,
many archaeologists have used the analysis capabilities of GIS to create spatial models.
The next section will discuss spatial data and describe some of the model types used by
archaeologists and cultural resource managers.
Spatial Modeling
Modern archaeology operates under the principles that there are patterns in the
behavior of humans and site locations “exhibit non-random tendencies” (Brandt et al.,
1992, p. 269). Because of this non-random tendency, archaeologists are able to create
spatial models to predict possible site locations. Spatial models use patterns found
between the location of a site and variables such as topography and water resources
(Brandt et al., 1992, p. 269). Environmental variables such as soil types, geologic and
hydrologic patterns, and topography have all been found to influence the settlement
patterns of native and post-contact settlers (Brandt et al., 1992). These variables are also

34

considered continuous, which allow models to use “powerful” statistical methods (Brandt
et al., 1992, p. 270).
Numerous modeling methods have been constructed with the sole purpose of
predicting potential cultural resource site locations. These models range in complexity
from simple point-specific analysis to multivariate statistical analyses (Judge et al.,
1988). Ebert (2004) discusses two main types of data commonly analyzed spatially:
point and areal. Point data is the spatial location of a site, an artifact, or a feature, while
areal data can include entire locations or regions (Ebert, 2004, p. 321). Point data can be
used to study trends in data as well as discern pattern distributions through “density
mapping and interpolation” (Ebert, 2004, p. 321). Density mapping is used to display the
dispersal of a particular variable (e.g. artifacts or features) over a given area and would be
considered a visualization method in Ebert’s (2004) hierarchal applications of GIS.
Interpolation uses multiple “mathematical procedures to convert point distributions to a
continuous surface” (Ebert, 2004, p. 322).
Another interpolation method is called kriging and is based on the idea that sites
in closer proximity to an area being analyzed have more impact than sites at greater
distances (Barceló & Pallarés, 1996; Ebert, 2004). Kriging takes into account both the
presence and absence of sites during statistical analysis in order to better predict site
locations (Judge et al., 1988; Barceló & Pallarés, 1996; Finke et al., 2008). More indepth interpolation methods such as kriging are not widely used by archaeologists, but
Judge et al. (1988) notes this may be due to a lack of technical understanding by
archaeologists.

35

Judge et al. (1988) goes further in describing spatial models, by grouping
predictive models based on their operability, proposing two distinctive categories:
intuitive and objective models. Intuitive models use “inductive or deductive logic” based
on analyses of patterns of either “human behavior” or known sites (Judge et al., 1988, p.
64). Judge et al. (1988) points out that intuitive models are the primary method used by
archaeologists and consequently have a “very high accuracy rate” of predicting site
locations (p. 65). This high prediction accuracy is due to archaeologists identifying a
specific variable that a singular site has in common with other related sites and then
looking for other areas with that same variable. As a result, many sites may remain
unrecorded because no one has thought to look in a location that does not share a specific
characteristic of known sites (Judge et al., 1988).
Objective models can be broken down into three separate sub-categories based on
the characteristics of a site’s dependent variable, the procedural method used, and the
weighting of independent variables (Judge et al., 1988). The three sub-categories are
termed: “associational, areal, and point-specific models” (p. 63). Table 1 provides a
concise description of these three types of objective models. There is overlap among the
three sub-categories in regards to the characteristics of objective models that Judge et al.
(1988) provides.
Associational models look at the relation of a site to another variable, such as
vegetation type or aspect. These types of models utilize statistical methods such as a
goodness-of-fit test to determine a level of significance and can be used as a predictive
tool (Judge et al., 1988). Areal models are used to “predict certain characteristics” of
different sites (p. 68). Specifically, areal models can be used to determine the amount of

36

sites in a given area and are similar to kriging in the sense that areal models use the
existence of a variable at one site to predict the occurrence of that variable in neighboring
locations (Judge et al., 1988). Areal models often predict site density over a large
geographical area based on relationships between variables, both dependent and
independent, in a smaller sample area. In contrast to areal models, point-specific models
focus on exact locations for potential site predictions. Point-specific models have
become the most commonly used model for archaeologists and cultural resource
managers because, rather than the model predicting the potential number of sites in a
given area without specifying locations, these models can predict whether or not a site
exists in a given location (Judge et al., 1988).

Table 1. Sub-categories of objective models (Judge et al., 1988, p. 64).

Espa et al. (2006) completed an archaeological site prediction model for Cures
Sabini located in the Tiber Valley, an area where sites from the Roman historical era have
been found. This model utilized both known archaeological sites and sites where no
37

archaeological evidence existed. The model took the point-specific data of these known
and “absent” sites and compared them with environmental data such as elevation, slope,
aspect, rock type, and water network (p. 151). The researchers created GIS layers for site
locations and each environmental variable. Researchers then completed a Classification
and Regression Tree (CART) analysis using coded data derived from GIS analysis. The
results of this analysis ranked each variable’s relative importance to site location. Based
on these rankings, researchers created a map in GIS symbolizing areas by a low to high
probability of containing a site (Figure 4). Espa et al. (2006) concludes that CART
models are unaffected by outliers and make a better analysis tool than logistic regression.
Figure 4. Grayscale map showing
the probabilities of potential site
location based on multiple
environmental variables in the
Cures Sabini area of the Tiber
Valley. Black polygons represent
a high probability of an
undocumented site location. Site
location probability decreases as
the grayscale approaches white
(Espa et al., 2006, p. 154).

The vast majority of archaeological modeling has been completed in attempts to
better locate prehistoric archaeological sites, which are sites dating to before the
European settlement of America. There has been little modeling completed to predict
historical site locations. Some of the reasoning behind this lack in modeling is historical
38

sites are often well documented on historical maps, in state and county archives, in
company records, and from the personal accounts of former inhabitants or workers of
historical sites (Judge et al., 1988). Judge et al. (1988) says time spent developing
models to predict historical site locations could be better used researching historical
documents. Also, historical site locations may not be based on the same factors as
prehistoric sites because of the way lands were surveyed by sections and parceled to land
claimants, and thus not as easy to model (Judge et al. 1988).
Effective models can be created to identify potential historical site locations for
areas where no historical documentation of such sites exists. These types of models can
use multiple environmental “predictors” such as proximity to water sources and
topography (Judge et al. 1988, p. 330). Pattern recognition models based on pointspecific data may not directly take into account multiple variables when being built,
while settlement pattern research can look at site locations based on a site’s relation to
variables, environmental or other (Judge et al. 1988). Settlement pattern research can
also examine the spatial relationship between sites to predict potential occurrences of
similar sites (Judge et al. 1988); however, pattern recognition models cannot predict
variations in site formation processes that lead to artifacts not being located at an actual
habitation site, as mentioned earlier (Schiffer, 1983 & 1986). It could be argued that
finding artifact deposits located at a distance from the site of habitation is just as
important to understanding what life may have been like at the actual site; therefore
modeling site patterns would still lead to further research being needed based on what is
known about human waste disposal patterns.

39

Regardless of the spatial model used, Judge et al. (1988) discusses that success of
a model in predicting sites is less important to an archaeologist than if a model makes
erroneous predictions. The authors discuss two main types of errors and relate them to
Type I and Type II statistical errors. If a predictive model has a null hypotheses claiming
an area does not contain a site, a Type I error would occur if one rejected this hypothesis
when indeed a site did not exist; a Type II error would occur if one accepted the null
hypothesis and the area did in fact contain a site (Judge et al., 1988). In regards to
predictive modeling, Judge et al. (1988) refers to Type I errors as “wasteful errors” (p.
62). Wasteful errors are termed as such because management of a site where no actual
site exists leads to an “inefficient” use of money and resources, hence wasting these
resources (p. 62). Type II errors, or “gross errors” can lead to unintentional damages to a
site because a model predicted, and the result accepted, there would not be a site in that
location (p. 62). In relation to archaeology and cultural resource management,
committing gross errors is more detrimental because it can result in the destruction of a
cultural resource. Judge et al. (1988) points out the “ideal predictive model” functions to
reduce both types of errors by making accurate predictions (p. 62). Archaeologists and
cultural resource managers must be thorough when constructing predictive models and
cautious of the results to avoid making either type of error.
In short, spatial models minimize the level of analysis exerted researching
historical maps for potential site locations. Spatial models are able to analyze single or
multiple variables to determine potential cultural resource site locations for a given
geographical area. Whichever method is used to locate a culturally significant site,
historical maps, LiDAR, or spatial modeling, the key parts should be recording site

40

dimensions and the data gathered from sites. As discussed in Section II, these data are
especially important for historical context. Site locations and dimensions are equally
important for cultural resource managers for developing management and protection
plans. The next section will discuss cultural resource management in more detail.

41

Section IV
Cultural Resource Management (CRM)
The National Historic Preservation Act (NHPA) aims to preserve cultural and
historical sites throughout the United States. Section 106 of the NHPA requires potential
impacts to cultural resources be avoided and mitigated. Section 110 further requires
agencies to ensure culturally or historically significant sites are not unintentionally sold,
damaged, or destroyed (Judge et al., 1988). Preserving and protecting culturally and
historically significant sites is the primary concern of cultural resource managers.
CRM goals also vary depending on location and site type. Some managers may
only be concerned with locations of sites strictly for managing their protection while
other managers may be more concerned with specific site types more than others (Judge
et al., 1988). For the latter type of CRM, site types that require greater attention and
protection measures depend on a determination of their significance. Site significance
goes back to the NRHP Criterion D discussed in Section II and relates to the information
content provided by a site and its features.
Determining Significance
Logging railroad grades located in forests represent a site type that is not typically
eligible for listing in the NRHP. These types of sites are often little more than a grade
through the forest with occasional through cuts and stream crossings that demonstrate the
engineering and construction of the era. On the other hand, logging camps are typically
eligible as they meet more than one of the criteria as discussed previously. Conners
(1990) developed a ranking system to provide an indicator of significance for logging
railroad grades, and associated logging camps, constructed and in use between 1890 and

42

1930 (Table 2). By following this ranking system, cultural resource managers can easily
apply a level of integrity to a site to ensure proper management. Researchers considered
sites ranked as ‘excellent’ or ‘good’ to be eligible for NRHP listing under criteria A, C,
and D (Conners, 1990).

Table 2. Railroad logging camp integrity ranking system (Conners, 1990).

Regardless of whether a site is eligible for protection based on NRHP criteria, if a
site has been recorded and contains relevant historical or cultural information, the site
should be managed as a cultural resource. Site protections can vary due to the
significance or integrity of a site and the nature of work in or near a site. Grounddisturbing activities such as excavation often require more explicit protection guidelines
and occasionally require an archaeologist or cultural resource manager on-site to ensure
protection of cultural resources. Other activities, such as logging, can greatly disturb a
site, but protection measures such as limiting equipment operation in certain areas can
mitigate disturbance. Since the potential for more information may be available at
previously recorded sites, possible destruction is usually avoided. Cultural resource
managers and archaeologists continue to follow federal and state guidelines using

43

multiple methods for site protection, but as Judge et al. (1988) states, more time is being
spent locating and managing these sites rather than analyzing the data within sites.
Ground or site disturbance can occasionally uncover previously undocumented
sites or new artifacts which may have been missed in initial surveys. Christopherson
(2008) and Stilson (2010C) surveyed known camp sites following a timber harvest. In
both cases, the removal of timber and understory greatly improved the ability of
surveyors to locate and document artifacts and landscape manipulations. The Schafer
Brothers logging camp recorded by Stilson (2010C) was shown on a 1938 USGS
topographical map, but surveys prior to timber harvest were unable to locate any artifacts
due to thick understory brush.
Another interesting finding at these sites was that ground-disturbing activities did
not greatly disturb or damage artifacts. It could be assumed that large tracked machinery
would completely destroy delicate artifacts such as glass bottles and earthenware;
however, in both Christopherson (2008) and Stilson (2010C), it could not be discerned
whether damage to artifacts occurred through natural events or during harvest activities.
With the knowledge that timber harvest may have a lesser impact than believed and that
removing timber and brush only enhances the ability to record a site, management of
these historical sites can possibly become less stringent. The best management in relation
to timber harvests may actually be limiting disturbance in known or presumed areas of
artifact accumulations and requiring full surveys to be completed following timber
harvest in or near recorded sites to look for artifacts or portions of the site that may have
been missed in initial surveys.

44

Conclusion
Historical logging camps represent an important period in the resource extraction
history of the United States. Logging camps provide historical context about the people
who inhabited those camps as well as a glimpse into the lives of the many workers and
their family members during the late nineteenth and early twentieth centuries. Historical
logging camp sites highlight the social, cultural, and gender differences occurring during
this specific time period, which could be representative of the country as a whole.
Locating historical logging camps can be accomplished through a number of means,
including research of historical maps and books as well as acquiring narratives from
former inhabitants.
Unfortunately for archaeologists, many logging camps were not documented on
historical maps and other sources due to their ephemeral nature and the number of people
with first- or second-hand knowledge of camp locations is rapidly decreasing. This lack
of documentation coupled with the loss of local knowledge requires research to focus on
new methods to better locate these sites. Utilizing LiDAR and spatial modeling may
provide methods to better locate historical logging camps; however, spatial modeling
does not account for the site formation processes as laid out by Schiffer (1983 & 1986)
and, therefore, may only provide limited success. Locating and documenting historical
logging camps must be done to ensure the preservation of data only those sites can
provide and allow cultural resource managers and archaeologists to better manage the
protection of these sites.

45

Chapter 3 – Research Area Background
Introduction
The focal point of this research, the Capitol State Forest, is located in Grays
Harbor and Thurston counties at the very southern tip of Puget Sound. This chapter will
discuss the research area in more detail. First, the technologies used to extract timber
from the Capitol State Forest during the late nineteenth and early twentieth centuries will
be detailed, followed by information about the companies who logged the forest during
that same time period. Next, the known and assumed Mason County Logging Company
(MCLC) logging camps used in this research will be discussed in greater detail.
Characteristics appearing to be common among the known and assumed MCLC logging
camps will be identified in order to demonstrate the potential importance those variables
may have for locating other camp sites. This chapter will conclude with an introduction
of the theory behind using an average distance to locate historical logging camp sites.
Capitol Forest is representative of the peak, and subsequent decline, of logging in
Washington State. The history of Capitol Forest was well documented in a publication
produced by the Washington State Department of Natural Resources (WADNR), the
current manager of the majority of the forest. In this publication, Margaret Felt (1975)
detailed the recent history of Capitol Forest, which included early settlers, early logging
history, and the replanting and reclamation of the forest by what is now WADNR. The
logging history is of great value to this research because it provides details about the
companies involved in logging the forest, including images. Felt (1975) also contains a
map of the forest with the locations of some of the known logging camps.

46

Despite its location and proximity to Puget Sound and Olympia (Figure 5),
logging in Capitol Forest, or the Black Hills, did not begin until the 1870s (Felt, 1975, p.
17). By the time of the logging boom in the Capitol Forest area, railroad logging was the
primary method of delivering logs to mills. As discussed in the previous section,
railroads were typically located along streams because of the low grade needed for
adequate train movement. Capitol Forest is no different and railroad grades can be found
along nearly every major stream in the Black Hills.

Figure 5. Current USGS topographic map (1994) of the Capitol State Forest in relation to
Olympia, Washington.

The topography of the Black Hills ranges from gentle slopes to steep mountainous
terrain. This topography required loggers to use a piece of equipment called a steam
donkey to pull logs through the forest to where they were to be loaded on trains (Figure
47

6; Labbe & Goe, 1961, p. 61; Spector, 1990, p. XV). Scars from logs being dragged up
hillsides by steam donkey operations can still be found covering the landscape of the
forest. Typically comprised of two log skids and a steam-powered engine, steam
donkeys used a cable pulley system, which allowed the donkey to pull logs to the
machine and pull itself through the forest (Andrews, 1954, p. 64). Steam donkeys could
also be built on train cars to pull logs directly to the trains.
Figure 6. Example of a steam
donkey. The timber of the
West provided an ample
supply of large skids such as
the ones shown here.
Photograph by Clark Kinsey
(UWL, 2015).

The main method for falling trees was to use a cross-cut saw after a large ‘face
cut’ was chopped into the tree. Initially used for ‘bucking’ trees into lengths suitable for
transport, cross-cut saws became more heavily used once loggers encountered the
massive trees of the West as these saws were found to be more useful in falling
operations (Cox, 2010, pp. 137 & 274). Existing old-growth stumps are often found with
the remnant notch in them from where loggers used spring boards as sawing platforms.
Spring boards, something unique to western United States logging operations, were
essential to get loggers above the dense ground cover and ‘butt swell’ at the base of the
tree (Cox, 2010, pp. 137 & 274).
48

Technologies of the time required loggers to spend long periods of time in a given
area and, as stated previously, the cost of transporting workers to the site, in terms of
production time lost, was more expensive than building a camp closer to the work site
(CALTRANS, 2013). Technological improvements in logging such as the chain saw and
the standard usage of trucks to deliver logs rather than trains did not become common
practices until the 1940s, a time when logging in the Black Hills was winding down (Felt,
1975, p. 32). If these technologies had come into use sooner, fewer camps may have
been needed throughout the forest because trucks could move more easily around the
landscape, being less bound by road gradient than trains. The incorporation of logging
trucks would also have sped up the transportation of workers to a site. Power saws
allowed workers to cut trees more quickly, meaning more acreage could be cut in a
shorter period of time. Moving faster through stands of timber would reduce the amount
of time a camp was in existence, but likely would have eliminated the overall need for
temporary logging camps. Setting up a logging camp when these technological
advancements were available would have been fiscally irresponsible; however, removing
existing rail infrastructure to provide access to log trucks may have been more expensive
than the financial benefits log trucks could provide.
Logging Companies in Capitol Forest
A number of logging companies were involved in the early logging of the Capitol
State Forest. These included the Vance Lumber Company, Mumby Lumber and Shingle
Company, Union Timber Company, Lytle Logging and Mercantile Company, Mud Bay
Lumber Company (partly owned by the Weyerhaeuser Timber Company) , and MCLC
(Felt, 1975, pp. 29-35; Carlson, 2003, pp. 123 & 214). Each company removed the

49

timber from their parceled land using their own railroad system. These companies also
constructed their own work camps within the forest; the Mud Bay and Mason County
Logging companies operated multiple satellite work camps, while the Union and Lytle
companies had one apiece (Felt, 1975, pp. 29-35 & map insert; Blum, 2000).
The most prominent logging company in the Capitol Forest area, as discussed by
Felt (1975, pp. 31-2), was the MCLC. In 1902, the Mumby Lumber and Shingle
Company, initially a subsidiary of the MCLC, constructed a mill at what would become
the town of Bordeaux on the east edge of the forest (Figure 7; OAHP, 1985). More than
200 workers were employed by the MCLC for their operations in the Capitol State
Forest, most at the mill, but many others in the work camps along 85 miles of railroad
through the forest (Felt, 1975, pp. 23 & 31). In 1924, MCLC purchased the Vance
Lumber Company’s operations near Malone adding another 10 miles of rail line to their
operations on the west side of the forest (Felt, 1975, p. 32; Carlson, 2003, p. 214).

Figure 7. Picture of Bordeaux Washington circa 1910 looking northwest. The large building to
the left is a hotel and the buildings to the north are living quarters for the workers. The mill
included all the structures in the southern half of the image. Taken from the Thurston County
webpage: http://www.co.thurston.wa.us/history/

50

For nearly 50 years, the seemingly endless supply of old-growth Douglas-fir
(Pseudotsuga menziesii) and other species was removed at a staggering pace from the
forest. Similar to many forested areas throughout America, the timber supply in the
Black Hills was exhausted, causing the mill at Bordeaux to be shut down. With the
closure of the mill in 1941, people who once worked in the hills and lived in the town left
and Bordeaux fell into ruin (Felt, 1975, p. 32). What remains is a common sight in the
industrial forests of the nineteenth and twentieth centuries, a ghost town representative of
the great logging era in Washington. Also left in the forest are the scars of this bygone
era, including the countless miles of railroad grades used to haul timber out of the woods.
The tracks are no longer present because the iron was typically moved after an area had
been harvested (Brashler, 1991) and because iron was salvaged for use in World War II.
Many of the wooden trestles constructed throughout the forest have also been removed or
destroyed, including some that were demolished by the United States Army during
“explosive experiments” (Felt, 1975, p. 35).
Capitol Forest Logging Camps
Another remnant of the decades of logging in the Capitol State Forest is debris
left by early loggers at sites where satellite work camps were located. Nearly all the
camps scattered throughout the forest were temporary camps where structures could be
loaded on train cars and moved to the next location. Much of what remains in these areas
is trash consisting of bottles, ceramics, and cans left behind by the loggers. Trash can be
the key to approximate camp occupation dates; for all of the camps discussed below,
these dates are approximately between 1900 and 1950. Identifying logging camp sites in
Capitol Forest was a key element of this research and involved combing through multiple

51

historical resources, including books, historical maps from the United States Geological
Survey (USGS), county timber cruise maps, and historical photographs from people who
worked in the forest or from prominent photographers of the era such as Clark and Darius
Kinsey. Researching logging companies and their associated camps in the Capitol State
Forest identified 13 known and nine assumed camp sites; many of these camp sites are
discussed in more detail below (Figure 8).

Figure 8. Map depicting all the known and assumed camp sites in the Capitol State Forest.

A map contained within Felt (1975, map insert) provided the location of some of
the logging camps within Capitol Forest. All four camps associated with the Mud Bay
52

Logging Company, who operated in the northern portion of the forest, are noted as well
as one logging camp site for MCLC, Camp 4. MCLC Camp 4 is also depicted on a 1938
USGS topographic map, and represents the beginning of this research project (Figure 9).

Figure 9. MCLC Camp 4 location shown on the 1938 USGS topographic map.

Camp 4 was located on Waddell Creek and was the only known permanent camp within
the forest. Concrete foundation pieces, which were found on-site, would not have been
associated with a temporary camp that could be transported by rail (Stilson, 2009).
MCLC Camp 4 is adjacent to Camp Four Creek and it was this creek name which led to
the discovery of MCLC Camp 7. Camp 7 was located along Camp Seven Creek,
southwest of what is now the Cedar Creek Correction Center work camp (Stilson, 2009 &
2010D). The location of Camp 7 was also discussed in research completed by Blum
(2000); this research was essential to the discovery of Lost Valley Camp and many of the
camps discussed below (Stilson, 2010D).

53

Another site shown on the map in Felt (1975, map insert) was named Hollywood.
Located just south of MCLC Camp 4, Hollywood was not a logging camp, rather it was a
home to many of the families of the loggers working in the forest (Figure 10). For this
reason, Hollywood was not used in distance measurements. Family camps could be a
necessity for those loggers who traveled with their family for work, but whose families
could not live in the satellite work camps or company town. In the 1950s, long after the
logging had ended and families had left, Hollywood was converted into a campground
(Ferguson, 2011A). The conversion of the Hollywood site to a campground along with
research completed by Blum (2000) led to the discovery of another former logging camp
in what is currently the North Creek campground (Figure 11; Ferguson, 2011B).
Figure 10. Photograph of
informational sign that once
stood in the Hollywood
Campground discussing the use
of the camp by workers’
families (unknown
photographer; Ferguson,
2011A).

54

Figure 11. Image of North Creek Campground, which was the location of
MCLC North Creek Camp. Note the abundance of English ivy. Ivy is
common among MCLC camps. Photograph by author (Ferguson, 2011B).

In completing research for site surveys of the camps mentioned above, the book
written by Joseph Pierre in 1979 titled When Timber Stood Tall helped identify and locate
additional MCLC camps. This book discusses some of the early logging of western
Washington, including two areas within Capitol Forest. One area, Gibson Creek (Camp
5) along the western edge of the forest is located on private property, but the camp
location can be inferred remotely from other sources (Pierre, 1979, pp. 79-80). Gibson
Creek Camp was not used to measure distances between camps because it is not directly
connected to other camps and there is not enough rail line beyond the Gibson Creek camp
site to warrant another camp in that area.
The other site discussed in Pierre (1979, pp. 56-64) was High Camp 7, a camp
located at the top of an incline. An incline is a stretch of railroad going directly uphill at
a grade too steep for trains to climb under their own power (Labbe & Goe, 1961, p. 123).
A steam donkey, which was used to pull trains up the grade and to slowly lower them
back down, typically sat at the top of the incline. The incline discussed in Pierre (1979,
55

pp. 58-9) and shown on the map in Felt (Figure 12; 1975, map insert) was located and
surveyed in 2010. During the survey, remnants of camps were found at the base of the
incline (MCLC Incline Camp) and at the top (High Camp 7). Remnants of the steam
donkey, likely used to aid trains up and down the incline, were also found (Ferguson,
2010A & 2010B).
Figure 12. Portion of map from
Felt (1975) depicting the location
of a MCLC incline.

The last MCLC camp identified through research, MCLC Camp 2, was found
after reading a transcript from an interview with a former MCLC train engineer
discovered while researching previously surveyed camps in Capitol Forest. In the
transcript, Harlan Smith referenced an old road, which has since been abandoned, when
he mentioned “[t]his goes up to Camp 2” (Baldo & Coombs, 1991). Remotely following
the old road grade, a site was identified as the best potential spot due to it being in close
proximity to a large ponded area near the end of the railroad grade along a major stream.
A field survey of the probable site location confirmed the location as a logging camp,
likely Camp 2, uncovering numerous artifacts from the early twentieth century
(Ferguson, 2011C).
56

It was the discovery of MCLC Camp 2 where a pattern in camp location was first
encountered. A ring of four satellite logging camps appeared to be within a similar
distance radially from the mill site at Bordeaux (Figure 13). This finding is similar to
what Maniery et al. (1996, as cited by CALTRANS, 2013) found in a survey of historical
sites in the Shasta-Trinity National Forest. Research by Maniery et al. (1996) notes that
work camps were within a radius of approximately one to two miles of mill sites. MCLC
camps were farther than one to two miles from the mill site at Bordeaux, but each
visually appeared to be at roughly the same distance away. It is possible the theory
provided by Maniery et al. (1996) that satellite camps were at a specific distance radially
from a mill site is still applicable, but distances likely vary by logging company and
topographic variability.
Figure 13. The ring of
satellite camps appear to be
at similar distances radially
from the mill site of
Bordeaux. These camps also
appear to be located at
similar distances from each
other. Note locations of
camps along major streams.

57

One camp location used in this research, Goliath Creek Camp, was found based
on a theory similar to the findings of Maniery et al. (1996, as cited by CALTRANS,
2013). After observing the ring of satellite camps at similar distances radially from the
mill site of Bordeaux, remote sensing analysis focused on an area to the south of
Bordeaux where no camp locations were known. Two potential locations were identified,
both within a similar distance to Bordeaux as the other camps in the ring and both having
similar characteristics of other camp locations in the forest. One potential site was
located at the junction of four railroad grades and the other was located just downstream
of a large water impoundment in a wide floodplain along Goliath Creek. Field
reconnaissance of the location at the grade junctions found no evidence of a camp, but
reconnaissance of the Goliath Creek location found a small midden, water pipes, and
English ivy (Hedera helix) (Ferguson, 2011E). Although the Goliath Creek location did
not contain artifacts consistent with other camps, specifically an absence of kitchenware,
the site shared multiple characteristics with other camp locations in the forest such as
being located adjacent to a major stream and near impounded water (Stilson, 2009,
2010D, & 2010E; Ferguson, 2010A, 2011B, & 2011C). For this reason, Goliath Creek
Camp and the distance between the camp and Bordeaux were used in this research.
There are two additional known MCLC camps where the sites have not been
surveyed fully. One site, referred to as the D-Line Ivy Spot (Figure 14), was believed to
have had a powder house where dynamite was kept and a “donkey sled operation” where
the footings of steam donkeys were constructed (Blum, 2000, p. 6). A portion of this area
was covered when the road through the forest was paved and thick English ivy now
makes surveying the site difficult (Blum, 2000). The D-Line Ivy Spot was used to

58

calculate the average distance between known camps because the site dimensions and
distances to other camps can be determined.

Figure 14. The D-Line Ivy Spot location discussed as a MCLC camp by Blum
(2000).

The second known camp, MCLC Camp 1, is believed to have been located
somewhere north of both the North Creek Camp and the D-Line Ivy Spot locations.
Although Blum (2000) provided specific details about the location of this camp, field
reconnaissance failed to adequately pinpoint the location of MCLC Camp 1 and camp
dimensions could not be ascertained. For this reason, distances between MCLC Camp 1
and other camps in close proximity were not included in calculations.
There are also two sites which have not been positively identified as logging
camps, but for the purpose of this research will be assumed to have been logging camps
based on the history of those sites and their relation to other positively identified camp
sites. Wedekind, once the site of a Civilian Conservation Corps (CCC) camp used to
house workers hired to replant Capitol Forest, was located at a very logical camp
59

location. Wedekind was located at the junction of five rail lines in a topographical saddle
and remnants of steam donkeys used to log Capitol Forest were found at the site (Figure
15; Ferguson, 2012A & 2012B). It is highly likely Wedekind was the site of a steam
donkey construction operation (Ferguson, 2012A).
Figure 15. 1941 aerial
photograph showing the
Wedekind area. At least three
(3) structures that were part of
the CCC camp are visible in
the photograph (designated by
star). The water body to the
northeast of the camp area is
where Wedekind Dam was
located (Ferguson, 2011D &
2012B).

A dam created by a large tree with planks set against it was found within close
proximity to the Wedekind site and may have supplied water to the camp (Figures 15 &
16). Technology used to create the dam is consistent with early twentieth century
logging. The dam is shown on a company map of the forest from 1924, a time when
logging in the forest was at its peak and logging camps were common. The 1924 map
does not cover the full Wedekind area; however, since the map does not indicate logging
camp locations in an area where two camps are known to have existed, it is likely that a
camp would not have been shown at Wedekind. Trees growing on top of the dam were
cored and found to date near 1954, likely after the CCC planting camp was abandoned
(Ferguson, 2011D & 2012B). This finding does not eliminate the notion of the dam
60

being related to a logging camp at Wedekind because the dam was likely kept clear of
vegetation while in use and it would have taken some time for trees to establish
themselves on top of the dam once it was no longer needed.
As discussed in the previous section, camps were often located at major railroad
junctions and as mentioned above, many campgrounds in Capitol Forest are located at
former logging camp sites; Wedekind became a campground in 1967 and is still in
limited use today (Ferguson, 2012B). Some artifacts consistent with other logging camps
of the time were also discovered in the Wedekind area and these artifacts in conjunction
with the steam donkey remnants and conclusions about the dam lead to the qualification
of Wedekind as a logging camp site. It is extremely likely the subsequent uses of
Wedekind as a CCC planting camp and recreation site led to the destruction of possible
camp remnants.
Figure 16. Portion of the
Wedekind Dam. The dam
was constructed from one
large log with planks placed
against the log to impound
water and sediment
(Ferguson, 2011D).

The other assumed camp site used in this research is currently the Sherman Valley
Campground. Five factors led to the inference of Sherman Valley Campground as a
logging camp site. The first three factors are site characteristics common among other
61

former camps in the forest; the site is now a campground, was located at a major railroad
junction, and is situated in a wide floodplain next to Sherman Creek. Additionally, there
is a structure shown at this location on the 1938 USGS topographic map during a time
when MCLC was still operating in the forest (Figure 17). The final finding was the
location of a powder house, a site where dynamite was stored for work involving
blasting, depicted on the map included in Felt (Figure 18; 1975, map insert). The powder
house location was at a safe distance, directly east of the site. The powder house can be
seen on 1941 aerial photographs of the area as well (Figure 19). Also visible on the 1941
photograph is a large opening in the forest at the same location as the structure shown on
the 1938 USGS map and the present-day campground. These findings provide the basis
for inferring that a logging camp was located at this site.

Figure 17. Sherman Valley Camp location shown on the 1938 USGS topographic map;
the black dot (indicated with arrow) depicts location of a structure.

62

Figure 18. Portion of the map from Felt (1975, map insert) depicting the
location of a powder house directly east of a four-way junction and the
location of the current campground.

Figure 19. Portion of 1941 aerial photograph of Sherman Valley Campground area. Star at
right indicates powder house location. Note rectangular opening in the forest along the
grade just past the major rail junction. This opening is also the location of the dot depicting
a structure on the 1938 USGS map (Figure 17).

63

Two other camps had been previously surveyed in Capitol Forest, both on the
west side of the Capitol State Forest. Bozy Creek Camp and what will be referred to as
B-Line Camp were quite a distance from the other MCLC camps located in the northwest
portion of the forest. It is unclear which company operating in Capitol Forest was
responsible for these two camps. The B-Line Camp may have been constructed by the
Vance Lumber Company based on its proximity to the town of Malone where the Vance
mill and headquarters were located, but artifacts found at the site date to after the
purchase of the Vance Lumber Company by MCLC in 1924 (Christopherson, 2008;
Stilson, 2011). Based on this finding, the B-Line Camp will be included as a MCLC
camp and part of this research model. Bozy Creek Camp located north of the B-Line
Camp was at a junction of three rail lines and a road labeled “Mox Chuck Truck Trail” on
the 1938 USGS topographical map. The main route out of the forest from this camp was
to the west where logs could have been delivered north to the town of Elma or south to
Malone and the Vance Lumber Company mill.
Many of the known logging camps in the Black Hills share common features. As
previously discussed, many camps were located close to major streams, but camps
located high in the hills were not. Conversion of many former logging camps into
campgrounds and camps being located at or near major rail junctions are other common
factors already mentioned. Two other variables common at some camp locations are
being located in areas with a wide floodplain and the existence of ponds or wetlands
immediately adjacent to camp sites. The wide floodplain may be due to railroad grades
being built along major streams, but the association of camps with bodies of water may
be more than coincidental. This is because camps located near ponded water were also

64

along streams that were not very wide and a greater source of water may have been
needed. The pond near MCLC Camp 7 appears to have been man made, possibly to
provide a water resource for the camp (Stilson, 2010D). MCLC Camp 2 and Goliath
Creek Camp were also located near large ponds (Ferguson, 2011C & 2011E). These
ponds are currently related to beaver activity, but may have initially been manmade
impoundments.
Each of the factors discussed above should be taken into account when attempting
to identify logging camp locations, but each variable may not apply to every location
(Table 3) depending on topography and access to suitable water sources, as noted by
CALTRANS (2013). Topography can be a key factor to specific site type locations (De
Reu et al., 2011) and may be the most important factor related to logging camp site
location. Railroads followed specific gradient requirements for efficient movement and
were located in sites with similar topography. In the case of Capitol Forest, railroad
grades were mainly located up the draws of major streams. Brashler (1991) notes a
similar occurrence of logging camp sites in West Virginia, finding that sites were mainly
located at the “headwaters of shorter tributaries” to main rivers and streams (p. 61).
Likewise, the logging camps of the MCLC in Capitol Forest appear to be situated in
similar topographical locations, a factor that should be taken into account when looking
for potential sites. The Mud Bay Logging Company had to construct “impossible grades”
and tall trestle crossings in order to access the steeper terrain of their ownership (Felt,
1975, p. 25). Two of Mud Bay Logging Company’s four camps are located in the steep
hills quite a distance from the main camp, demonstrating how topography impacted camp
location.

65

x

x

x

x

Consumables4
5
Metal Pieces
Steam Donkey
Water Pipes
Structural Remnants6

x
x

x
x

x
x

x
x
x

x
x
x
x
x
x
x

x7
x
x

x

x

x
x

x

x
x

x
x

x
x
x
x

x
x

x
x

x
x
x

x
x

x
x

x

2

x
x

x

x

Wedekind

Incline Camp

1

x
x
x
x

x

x
x

Hollywood

High Camp 7

Goliath Creek Camp

D-Line Ivy Spot

x
x
x

Sherman Valley Camp

x
x
x
x

x
x

North Creek Camp

Near Major Landscape Manipulation3
Exotic Plants

x
x

Lost Valley Camp

x

Camp 7

x
x

Camp 4

x
x
x

Artifacts

Near Rail Junction
Near Stream/ Wide Floodplain
Impounded Water Presence
Current Campground

Camp 2

Bozy Creek Camp

B-Line Camp

Bordeaux

2

Camp Name

x
x

x
x

x

x

Table 3. Shared variables among known and assumed MCLC camps used in this research.
1
Hollywood was not a logging camp, but a family camp. 2These are assumed camp locations
3
Major landscape manipulations include large (>10 feet) railroad grade cuts or fills and trestle
crossings. 4Consumables artifacts include bottles, earthenware, enamelware, tin cans, boot
leather, etc. 5Metal pieces are related to heavy machinery, trains, and pieces related to railroad
operations. 6Structural remnants include bricks, concrete foundations, stone footings, and pilings
used for structures. 7Site identified by Blum (2000) as having a steam donkey construction area.

Felt (1975, p. 32) concludes the discussion about the logging history of Capitol
Forest by stating there are “some seven” logging camps located in Capitol Forest (p. 32).
Not including the company town of Bordeaux and Hollywood family camp, research and
surveys completed by Blum (2000) and those completed by Christopherson, Ferguson,
and Stilson between 2009 and 2012 have documented 12 logging camps. The locations
of another nine sites are known or assumed to be camps, but as yet, have not been
adequately surveyed (Table 4). Of these 21 known and hypothesized camp sites, 15 are
believed to have been used by MCLC and there remains potential for more MCLC camps
to be found within the forest.

66

Company
Lytle Logging and Mason County Mud Bay
Union Logging
Logging
Logging
Mercantile
Company
Company
Company3 Company
Total
Total Number of Camps 21
1
15
4
1
Known to have existed1
Assumed to have existed
Documented
Undocumented

2

19

1

2
12
9

1

13

4

2
10

2

4

5

2

1

1

4

On Public Lands
17
14
3
On Private Lands
4
1
1
1
1
Table 4. Total number of known and assumed logging camps in the Capitol State Forest, not
including the town of Bordeaux or Hollywood family camp. 1Camps known to have existed were
determined through research of historical documentation and field surveys. 2Assumed camps are
based on professional judgment of possible sites taking into account site variables shared with
known sites. 3Mason County Logging Company may include camps originally constructed by
Vance Lumber Company. 4Camp numbers include assumed camps.

Large gaps where no MCLC camps are known to have existed in Capitol Forest
are evident when looking at the locations of the 15 known and assumed MCLC camps
(Figure 20). Based on the knowledge that it was not cost effective to transport loggers
long distances to where logging activities were taking place, it could be concluded that
there are indeed more camp locations within the Capitol State Forest. The method
proposed by this research uses the average distance between known and assumed camps
based on three different measurements: from the edge of camp extents, from a central
point within camps, and from the mill site in Bordeaux to the edge of satellite camp
extents. The theory behind this average distance method is that camp locations were
typically identified before railroad construction and logging activities began as part of a
transportation plan to ensure a profit could be made from logging a given area (Andrews,
1954, p. 74). Logging company owners likely used a systematic method to identify camp

67

locations, using an approximate distance a camp should be built from a company town or
previous camp in order to maximize production and profits.
An average distance would have been derived from calculations related to a cost
analysis based on one or more of the following three factors: how long it takes to
transport workers to a work site; how long it takes to harvest an area based on
topography, timber size (e.g. diameter and height), and logging technology; and how long
it takes to deliver the timber back to the mill in order to keep the mill operating at
capacity. Engineers traversing potential new rail lines through the forest would have
used this approximate distance to identify the most feasible camp site nearest that
distance. If this were the case, then calculating the average distance plus or minus one
standard deviation between the known and assumed MCLC camps could pinpoint other
MCLC camp sites. This research used the company town of Bordeaux along with 10 of
the 15 known and assumed MCLC logging camps to calculate the average distance. Only
10 known and assumed camps were used to calculate distances because the remaining
five camps were not directly connected to the other known or assumed camps. These
calculations identified 36 areas where logging camps could potentially be located.

68

Figure 20. Known and assumed MCLC camps depicted on the current USGS topographic map
(1994). Ovals signify gaps in Capitol Forest where no logging camps are known to have
existed. House symbol indicates known and assumed MCLC logging camp locations.

69

Chapter 4 – Methods and Analysis
Creating Spatial Data
Field data from known and assumed Mason County Logging Company (MCLC)
camps was collected between 2009 and 2012 using a Garmin 60CSx handheld GPS.
Data collected included the locations of features such as railroad grades, structural
remnants (if any), trash accumulations, landscape manipulations, and site extents. Points
referencing site extent were used along with a 2-meter hillshade digital elevation model
(DEM) based on LiDAR data from Washington State Department of Natural Resources
(WADNR) to create a polygon feature in ArcMap 10 (versions 10.1 and 10.2) to
distinguish an approximate camp size (Figure 21).
Figure 21. Example of
polygon feature created to
display camp extent (crosshatched area) overlaid onto
a 2-meter LiDAR hillshade
DEM. Points represent
landscape manipulations or
artifact accumulations.

A layer of railroad grades known to have been used by the MCLC was created in
ArcMap 10 using multiple sources for accuracy (Figure 22). A digitized historical United
States Geologic Survey (USGS) map from 1938 provided by WADNR was compared to
a MCLC map from 1940 depicting all company railroad grades. Both of these map

70

sources were verified against the 2-meter hillshade DEM based on LiDAR data from
WADNR. Additional layers for streams, bodies of water, roads, trails, townships, and
sections were created in ArcMap 10 by clipping WADNR data for the research area.

Figure 22. Sources used to create GIS layer of grades used by MCLC; 1938 USGS topographic
map (A), MCLC railroad grade map compiled in 1940 (B), and 2-meter LiDAR hillshade DEM
(C). Map (D) shows the final grade location based on the three sources. House symbol
designates camp location.

71

Grade Measurements
Three methods were used to calculate average distance between camps; camp
extent, camp central point, and Bordeaux central point to satellite camp extent. The
edges of camp extent were easily distinguishable based on GPS data as discussed in the
previous section. Basing distance between camps on camp extent may not be the best
method because extent can grow or shrink over time; therefore, distances were also
measured from a central point within each camp. Using camp central points is based on
the notion that road mile markers are measured to or from a specific point location within
a town such as a post office or city hall. Choosing central points may be an arbitrary
exercise because, with the exception of Bordeaux where the mill could be used as a
central point, no logical central location exists in each camp. Therefore, the third
distance measurement method, Bordeaux central point to satellite camp extent, was
completed to eliminate the subjectivity of distinguishing a central satellite camp point.
Railroad lines between camps were merged into a single line segment and the
length of each segment between Bordeaux and 10 of the 15 known and assumed MCLC
camps was calculated using the Calculate Length tool within the ArcMap XTools
extension. Lengths were measured in feet based on the fact that American logging
engineers, both past and present, measure distances in Imperial Units / United States
Customary Units. Lengths of the segments between camps based on each of the three
methods were exported into a Microsoft Excel (2010) spreadsheet where the mean
distance between camps and the standard deviations were calculated (Table 5). One
standard deviation was then added to and subtracted from the mean distance to determine
a distance range to be field verified from each camp location.

72

CAMP EXTENT
Camp Grades
Bordeaux to Camp 2
Bordeaux to Goliath Creek Camp
Camp 7 to Sherman Valley Campground
Sherman Valley Campground to Lost Valley Camp
Sherman Valley Campground to Incline Camp
Sherman Valley Campground to North Creek Camp
North Creek Camp to D-Line Ivy Spot
High Camp 7 to Wedekind
Lost Valley Camp to Incline Camp

Length (ft)
11336.7132
9817.765
11974.0901
9302.8851
14702.1787
11843.3469
10439.8725
15884.5971
13406.3977

Difference
-741.9364
-2260.8846
-104.5595
-2775.7645
2623.5291
-235.3027
-1638.7771
3805.9475
1327.7481

Dif^2
550469.6052
5111599.1243
10932.6867
7704868.4978
6882904.9968
55367.3554
2685590.3471
14485236.4573
1762915.0466

Probable Camp Distance Range (ft)

-1 SD
9863.6472

Mean
12078.6496

+1 SD
14293.6520

Range

Standard
Deviation
4906235.5146

2215.0024

2215.0024

4430.0047

CENTRAL POINT
Camp Grades
Bordeaux to Camp 2
Bordeaux to Goliath Creek Camp
Camp 7 to Sherman Valley Campground
Sherman Valley Campground to Lost Valley Camp
Sherman Valley Campground to Incline Camp
Sherman Valley Campground to North Creek Camp
North Creek Camp to D-Line Ivy Spot
High Camp 7 to Wedekind
Lost Valley Camp to Incline Camp

Probable Camp Distance Range (ft)

Length (ft)
13010.1594
12046.8332
13286.8122
9751.2972
15072.0821
12557.3776
10866.0798
16201.4528
13683.2823

Difference
68.4509
-894.8753
345.1037
-3190.4113
2130.3735
-384.3309
-2075.6287
3259.7443
741.5738

Dif^2
4685.5244
800801.8226
119096.5519
10178724.1561
4538491.4232
147710.2555
4308234.5510
10625932.8217
549931.7361

-1 SD
10964.5385

Mean
12941.7085

+1 SD
14918.8785

Range

Standard
Deviation
3909201.1053

1977.1700

1977.1700

3954.3399

BORDEAUX CENTRAL POINT/CAMP EXTENT
Camp Grades
Bordeaux to Camp 2
Bordeaux to Goliath Creek Camp
Camp 7 to Sherman Valley Campground
Sherman Valley Campground to Lost Valley Camp
Sherman Valley Campground to Incline Camp
Sherman Valley Campground to North Creek Camp
North Creek Camp to D-Line Ivy Spot
High Camp 7 to Wedekind
Lost Valley Camp to Incline Camp

Length (ft)
12735.8146
12004.58546
11974.0901
9302.8851
14702.1787
11843.3469
10439.8725
15884.5971
13406.3977

Difference
258.7293
-472.4999
-502.9953
-3174.2003
2225.0933
-633.7385
-2037.2129
3407.5117
929.3123

Dif^2
66940.8461
223256.1554
253004.2274
10075547.2645
4951040.3900
401624.4305
4150236.2202
11611136.2863
863621.4329

Probable Camp Distance Range (ft)

-1 SD
10458.5337

Mean
12477.0854

+1 SD
14495.6370

Range

Standard
Deviation
4074550.9067

2018.5517

2018.5517

4037.1034

Table 5. Distance between individual camps, mean distance, and standard deviation calculations
using camp extents, central points within each camp, and Bordeaux mill site/satellite camp extent.
The standard deviation in the lower right corner was calculated from the Microsoft Excel standard
deviation function as a verification of the manual standard deviation calculations in the table.

Once the plus-1 and minus-1 standard deviation distances were determined, the
measure tool in ArcMap was used to measure distances along the grades from known and
assumed camps to plot potential sites. Measurements for the central point method were
73

initiated from the designated central point of each camp, while extent measurements were
made from the edge of determined camp extents. Measurements for the Bordeaux central
point method were initiated from the mill location at Bordeaux and from the defined
extents of each satellite camp. Due to the precision of measuring along a line in ArcMap,
measurements were made to the nearest whole foot rather than the ten-thousandth of a
foot as shown in the standard deviation calculations (Table 5). A point was placed at
each location marking the beginning and end of the range for each of the three methods.
Different colored symbols were used when there was overlap based on the direction of
measurement; for instance, if measuring north along a grade from one camp overlapped
locations measured moving south from another camp, the symbology would change to
designate the overlap (Figure 23). There were a total of 41 measured segments based on
extent measurements, 42 from central point measurements, and 44 from Bordeaux central
point measurements (Appendix A). Some measured segments contain more than one
beginning or end point due to segments spanning a railroad junction(s).
For ease of field verification, clumps of measured distance ranges were assigned
an aggregate segment identifier (Appendix A). These 36 aggregate segments (Figure 23)
were arranged into a Microsoft Excel (2010) spreadsheet where the location, access, site
probability, and field verification information was entered. Location refers to the public
land survey system or section, township, and range of each segment. Access provides
details of whether or not a segment is along a road or trail. Site Probability documents a
subjective opinion based on a remote site evaluation taking into consideration the factors
that appear to be common among MCLC camps previously recorded, as discussed in
Chapter 3 (Table 3). These factors include proximity to a railroad junction or stream,

74

presence of impounded water, if a campground is located in part of the segment, and
presence of a large landscape manipulation feature (i.e. grade cuts or fills over 10 feet).
Discussed in more detail below, the probability also took into account topographic
proximity to other MCLC camps and proximity to former MCLC property lines.
Segments were assigned a low, medium, or high probability based on these factors;
however, the probabilities were simply an educated opinion and had no impact on field
verifications. Field Verification lists the date of each site visit, the type of artifact or
feature found, and whether or not a camp site was identified.

Figure 23. Map of all measured segments with aggregate segments identified.

75

Field Verification Process
Field verification methods involved traveling to locations within Capitol Forest by
vehicle and hiking along old railroad grades to the locations indicated by distance
measurements. Many old railroad grades have been converted to trails or roads within
Capitol Forest, making travel along those grades easy. Less time was spent surveying
grades that have since been converted to roads due to the level of disturbance in those
areas. Trails were thoroughly surveyed because, as discussed below, trails can be helpful
for locating artifacts. Walking railroad grades included weaving from one side of the
grade to the other in order to survey the adjacent land because camps would not be built
directly on the grade, but rather along it in wide areas or flats. Besides local topographic
features, such as flats along grades, other areas receiving greater attention were draws
(features in the terrain where water may flow or drain), stream crossings, and steep areas
near flats along grades. Finding artifacts in these areas is common to both logging camps
and homesteads of the late nineteenth and early twentieth centuries and could be
considered typical topographic indicators of high site probability.
Major streams adjacent to grade segments were also surveyed in attempts to
locate artifacts consistent with logging camps of the early twentieth century. As
discussed previously, it was common for camps to be located near streams and
inhabitants frequently threw broken or empty ceramics and glassware into adjacent
streams (Figure 24; Stilson, 2009, 2010D, & 2010E; Ferguson, 2010A, 2011A, 2011C, &
2011E). Once in the stream, these artifacts were distributed over a larger area as pieces
washed downstream. For this reason, verification of grade segments adjacent to major
streams were completed by traveling upstream from below each segment.

76

Figure 24. Examples of trash deposited in streams and distributed by fluvial processes
(designated by stars). Photographs by author from MCLC Camp 2 (left) and MCLC Camp 7
(Stilson, 2010D; Ferguson, 2011C).

Data collection consisted of digital photographs and written notes of artifacts and
features of significance. GPS points were taken with a Garmin 60s handheld GPS where
artifacts or landscape anomalies were found and the points transferred to ArcMap. Some
segments were surveyed as part of WADNR timber harvest operations and not as a part
of this research. Reports completed by Christopherson (2009A & 2009B), Ferguson
(2012C), Nordstrom (2012 & 2014), and Vaughn (2013) were used to verify findings of
the segments covered in each report.
Logging Camp Site Identification
Knowing what to look for in the field is important in order to identify whether or
not a site was a logging camp. Common variables appear to be shared by multiple camps
used by MCLC (Table 3). Sites characteristics such as proximity to rail junctions,
floodplains of a major stream, or major landscape manipulations such as a large grade cut
or fill (greater than 10 feet) are shared by almost every known and assumed MCLC camp.
Consumables artifacts are a staple at every site except those that have not been fully

77

surveyed; however, some of these types of artifacts or site features could be related to
logging, railroad activity, or homesteads. There are a handful of homestead or cabin sites
denoted on historical maps of Capitol Forest, but these areas have either been
documented previously or the sites surveyed and nothing found. Determining whether
artifacts or site features are related to a logging camp or other human activity can be
difficult without an understanding of what those artifacts and features mean.
A single artifact found in the woods may mean very little by itself, but when other
features around that artifact are taken into consideration, a full picture can begin to take
shape. Likewise, comparing findings or recognizing features and patterns of those
features from known sites can help archaeologists better identify site type. Artifacts can
be the easiest pieces of evidence to identify site type. Logging camp artifacts can include
consumable products such as earthenware or china, glass bottles, tin cans, enamelware,
boot leather, gloves, tools, and more.
The difference between logging camp artifacts and artifacts found at a homestead
are the types of bottles, china, cans, etc. Logging camps have an abundance of basic
white china with little to no decorative properties because this type of china was cheaper
and loggers did not typically have a need for such sophistication (Figure 25). While more
common at homestead sites, decorative china has been found at logging camp sites, but
not with great frequency. Decorative china at logging camp sites may also provide
evidence for the presence of women or families. Dating china can be fairly accurate
provided a maker’s mark can be located and identified. Maker’s marks can provide the
location and date range a piece was manufactured. Finding a maker’s mark can be
difficult in the forest as ceramic fragments, also known as sherds, are all that remain.

78

Figure 25. Example of the plain white china often found at logging camps
(left) and blue transfer print decorative china more commonly found at
homestead sites. Photographs by author.

Figure 26. Example of bottle found at a historical camp site. This brown bottle
was manufactured in 1933 based on the Owens-Illinois maker’s mark shown. In
the absence of a maker’s mark, manufacturing variations can identify a possible
production date. The oval-shaped seam on the bottom (indicated by arrow) is
likely a cut-off scar where the bottle was cut from a suction machine; this type
of scar dates the bottle between 1904 and 1950. Photograph by author
(Ferguson, 2011A).

79

The types of glass bottles found at a site can also be indicative of a logging camp
site. The most common bottles found at MCLC camp sites are ketchup and liquor
bottles. These types of bottles can also be found at homesteads, but not in the numbers
found at logging camps. Bottles can provide the best methods for obtaining a date of
usage. Like china, bottles often have maker’s marks from which an exact manufacturing
date can be ascertained from bottle identification sources (Figure 26).
Other bottle features can help pinpoint an approximate manufacturing date.
Features such as seams, which can identify the type of mold used, how a lip or finish was
applied to the bottle, closure type (e.g. cork), letter embossing, and glass color can all be
used to identify a manufacturing date range. Dating a bottle cannot determine site type
unless a specific site type only existed during a given time period, which can be the case
for MCLC. Although whole bottles make identification much easier, they are not
required because glass fragments, also known as shards, can provide enough clues based
on the various dateable bottle characteristics listed above.
Artifacts made from metal were common items left in the forest following the
major logging era. Metal artifacts include cans, enamelware dishes, pans, saws, files,
barrels, stove parts, bed frames, and more. These artifacts can be indicative of any
habitation site type, but some can be more commonly found specifically at logging camp
sites. Certain tin cans are found on a regular basis in logging camps including condensed
milk cans and tobacco tins. Similar to bottles, can features such as closure type, seam
type, and dimensions can help date that can. Various tobacco brands existed in cans from
1907 to the 1960s; however, Prince Albert was the most common brand found in logging
camps. Unfortunately for many historical sites such as logging camps, tin cans end up

80

deteriorated beyond recognition due to time and climate. Metal artifacts such as
enamelware are not exclusive to logging camps, but can be if found with other artifact
types and features. Metal bed frames are a definitive indication of habitation and have
been found at multiple camps in Capitol Forest, but could also be related to a homestead
(Stilson, 2009; Ferguson, 2010B & 2011A). Cross-cut saws can be more common to
logging camps, but homesteads of the same era also used cross-cut saws to clear land.
Structural artifacts found at sites that were, at one time, home to forest workers
include bricks and foundational pieces. Bricks can be indications of a structure and are
very common at logging camp sites in Capitol Forest, including MCLC’s Camp 2, Camp
4, Camp 7, High Camp 7, and Lost Valley Creek Camp (Stilson, 2009, 2010D, & 2010E;
Ferguson 2010B & 2011C). Bricks can help ascertain approximate dates of production
based on manufacturing method or if a name was part of the brick mold; however, many
bricks were mass-produced and were often moved from site to site. For this reason,
bricks are not trustworthy artifacts by themselves for dating a site. Foundational pieces
such as concrete or stone footings also indicate former habitation. Concrete foundation
pieces demonstrate a more permanent structure location while stone footings may be
more representative of temporary residents (Stilson, 2009, 2010C, & 2011).
Similar to structural remnants, leather artifacts such as shoe parts may not provide
a time of usage for a site, but the style of shoe can be used to identify site type and
determine the presence of women and children (Ferguson, 2011A). Leather boots similar
to work boots worn by foresters and loggers today are common at logging camp sites.
Shoes with more decorative qualities and colors signify a woman’s presence in a camp
while smaller shoes can indicate the presence of a child (Figure 27; Ferguson, 2011A).

81

Shoe remnants can be found at various site types, but are indicative to some form of
habitation rather than a simple logging activity.

Figure 27. Leather shoe examples found at Capitol Forest logging camp sites. Plain boots (left)
were commonly worn by loggers while the decorative shoe likely belonged to a woman.
Photographs by author (Ferguson, 2011A).

All artifacts discussed above can help assign an approximate date range a site may
have been used as well as determine the type of historical site (e.g. logging camp versus
homestead). Artifacts, along with research into the history or ownership of a specific
area, can also provide site type determination. For instance, if numerous artifacts were
discovered in the middle of the forest near a railroad grade and there was no land grant
patent listed for that geographic location, the site is more likely to be a logging camp than
a homestead. An absence of the types of artifacts listed above, however, can make it
difficult to distinguish between a habitation site and a site where only logging or railroad
activity took place.
In the absence of consumables artifacts, other site features such as equipment
pieces can help identify a site as a potential logging camp. Equipment left behind may
signify a site where only logging operations took place rather than a camp site. For
instance, wire rope, hereafter referred to as logging cable, in varying diameter can be

82

found in abundance throughout the forest. Single strands of cable are likely related to
logging activities; however, large piles comprised of various diameters of logging cable
have been found associated with logging camp sites throughout Washington State,
including those used by the MCLC (Figure 28; Ferguson 2010A & 2010B).
Figure 28. Mound of logging
cable found at MCLC Incline
Camp. Photograph by author
(Ferguson, 2010A).

Steam donkey remnants can also be common for both logging camp sites and
logging activity only. A few steam donkey remnants appear to have been left in random
locations within Capitol Forest, but others seem to be connected to logging camp
locations. Discussed previously, three of the known and assumed camps contained or
were known to have had steam donkeys associated with them. High Camp 7 and
Wedekind both had remnants on-site while the D-Line Ivy Spot was known to have had a
steam donkey construction operation (Blum, 2000; Ferguson 2010B & 2012A). Steam
donkey remnants typically include the log skids with large iron pieces within each skid
and sometimes connecting skids. Corrugated metal from roofing elements, metal
waterlines, and logging cable are also common artifacts to find in conjunction with steam
donkey remnants.
83

Landscape manipulations, in the case of logging activities, include any
modification or addition to the landscape. Large amounts of earth were commonly
removed from or added to areas to keep the rail grade at allowable slopes for train
movement; these are called cuts and fills with through cuts being when earth is removed
from both sides of the grade (Figure 29). These features can range from one foot to
greater than 30 feet in depth or height. Cuts deeper than 10 feet required heavy
excavation work to move the vast amounts of material to reach grade and, prior to
bulldozers, this work was completed with tools such as a Fresno scraper (Labbe & Goe,
1961, pp. 62-3). Camps needed to house the workers constructing these large landscape
modifications have been found near large cuts in Capitol Forest (Christopherson, 2008;
Stilson, 2010A & 2011). Smaller cuts and fills could be completed more quickly and
likely did not require lodgings nearby. Even small cuts and fills are visible using LiDAR
and that data can be used to map grades as completed in the methods above. Some
smaller modifications such as flattened areas near grades and small rail spurs (short
lengths of grade) can indicate a possible camp site; however, without artifacts to
definitively identify past habitation, these features are likely logging related.
Figure 29. An extreme example for
Capitol Forest, this through cut near
Mud Bay Logging Company’s
Camp 2 is greater than 40 feet deep
through solid rock. Photograph by
Lee Stilson (Stilson, 2010A).

84

Another landscape modification common to late nineteenth and early twentieth
century logging was the wooden trestle. Bridges made from stacked logs or pilings and
cut lumber were constructed when there was no other route around a stream or gully
(Figure 30; Labbe & Goe, 1961, pp. 33-4). Logging camps in Capitol Forest have been
located near significant trestles (Stilson, 2009; Ferguson, 2010C); however, trestles are
not an adequate indication of a logging camp site. Due to the disposal patterns of humans
as discussed by Schiffer (1983 & 1986), areas where trestles once stood are good
locations to survey for artifacts that may have been thrown from passing trains. Artifacts
discarded in these locations can lead to a camp locations.

Figure 30. Trestle remnants found near MCLC Camp 4. These remnants are approximately
20 feet tall. Although many standing trestle remnants were removed or demolished in the
decades following the end of the major logging era in Capitol Forest, a number of examples
remain. Photograph by author (Ferguson, 2010C).

85

Exotic plants, which are non-native species either planted or spread by other
means in the forest, can also help identify a logging camp site. Fruit trees, typically apple
or cherry, and ornamental flowers often leave no question as to whether or not a site was
inhabited (Felt, 1975, p. 35; Stilson, 2010A; Ferguson, 2011A). English ivy (Hedera
helix) was found associated with the D-Line Ivy Spot, Hollywood family camp, North
Creek Camp, and Goliath Creek Camp sites and can be a good indicator of past
occupation (Ferguson, 2011A, 2011B, & 2011E). Another exotic plant species that can
help lead to a positive logging camp site determination is holly (Ilex aquifolium). Exotic
plants can also be related to homesteads as settlers commonly planted ornamental
species. Non-native plants can also appear in random locations from natural mechanisms
such as birds spreading seeds or berries to different locations. These plants can also be
spread when an apple core or cherry pit is tossed from a train as its moving through the
forest. For these reasons, exotic plants should not be used to define a site as a logging
camp site without the presence of other site variables.
Potential Complications and Limiting Factors
There are multiple factors that could complicate and limit the success of this
research. Factors unrelated to research methods such as landscape changes, timber
harvest, road construction, recreation, looting, local topography, and property boundaries
can all potentially lead to negative findings. Factors related to the measurement methods
used in this research or basing known camp identifications solely on artifact
accumulations rather than known habitation sites could also lead to unintended errors.
A large number of railroad grades in the Capitol State Forest have been converted
to forest roads since MCLC left the forest. In some places, former railroad grades have

86

been widened, paved with asphalt, or covered with gravel. Disturbance along these
converted grades covered, if not completely obliterated, any trace of artifacts or past
inhabitants as was the case with the D-Line Ivy Spot where road paving is said to have
covered artifacts (Blum, 2000). Many other grades have been converted into recreation
trails. This can be seen as a positive outcome, one which allows the public an
opportunity to enjoy these historical features; however, some trails located near known
logging camps can disturb artifacts found on-site (Figure 31). Numerous segments to be
verified are now either roads or trails, likely limiting success of locating artifacts.

Figure 31. Example of a recreation trail disturbing a known logging camp site. Turned
up by all-terrain vehicles, glass shards and earthenware sherds can be seen in the
footprint of the trail (designated by stars). Photograph by author.

87

How recently timber harvest occurred in potential locations can improve or
reduce the ability to locate or identify possible camp locations. As noted earlier, artifacts
can be easier to find following timber harvest as the understory brush is almost
completely removed and ground disturbance from equipment can uncover artifacts with
little damage (Christopherson, 2008; Stilson, 2011). Heavily stocked stands of 15-yearold or older reproduction timber with fully closed canopies and little ground cover also
improve the ability to find artifacts and potential camp sites (Stilson 2009 & 2010A).
This is because ground visibility is higher and the most recent harvest activity likely
turned up artifacts that were partially or fully buried. The period after recently harvested
areas have been planted, but prior to the canopy closing, can be a difficult stand
successional stage to survey for potential sites because the ground is nearly impossible to
see with the amount of vegetation, both trees and shrubs, covering the forest floor.
Likewise, mature stands, those greater than 60-years-old that are developing an
intermediate canopy, can be difficult to survey due to the amount of ground cover. As
discussed, railroads were constructed along major streams due to the gradients required
for train movement. This fact along with current timber harvest restrictions requiring
areas immediately adjacent to streams to be left for riparian protection, means that many
surveyed areas were within mature stands.
Local topography caused segments to be close in geodesic distance, which is the
straight line distance between two points ignoring topography, to existing camps. Many
segments were well below the average distance from known camps due to the topography
of the landscape (Figure 32). In these cases, loggers of the era could have simply walked
across the landscape to these potential sites in the time it would take them to be

88

transported by rail, making a camp in those locations unnecessary. Short geodesic
distances were one of the variables used in determining site probability (Appendix A) and
likely limited success of locating undocumented logging camps.

Figure 32. Example of grade segments that, due to topography, are too close to
existing camps in geodesic distance to make a logical camp location. In this
example, aggregate segments J4, K1, and K2 are all less than 6,000 feet from the
existing camp, which is well below the calculated average distance between known
and assumed camps.

A number of segments were at the end of known rail lines and up against
ownership lines between MCLC and the Mud Bay Logging Company. This is an
important factor because it does not seem plausible that a company would undertake the
time and cost of establishing a satellite logging camp when that company’s property only
extends a limited distance past where a camp would have been logical (Figure 33).

89

Segments close to known property boundaries are a limiting factor to success and also
contributed to lower probability predictions (Appendix A).

Figure 33. 1938 USGS topographic map showing measured distance segments in
close proximity to land managed by the Mud Bay Logging Company.

Looting may also make it more difficult to find artifacts. It is common for people
to go to historical sites such as logging camps to gather bottles and other artifacts with the
intention of selling those artifacts for profit. This has clearly happened at MCLC Camp
4; however, the identification of Camp 4 was easy given the sheer volume of artifacts
found on-site as well as the documentation of the camp on historical maps (Stilson,
2009). Much of the information collected by Blum (2000) was from a local resident who
knew the locations of many camps within the forest; it can be assumed that camp

90

locations are common knowledge among longtime residents of the Capitol Forest area.
Finding and recording historical camps may not deter looting; however, documenting and
revisiting sites could determine if looting has occurred. Knowingly removing artifacts
from historical sites is considered a misdemeanor crime in Washington State.
A potential complication specifically related to methods used in this research is
that of compounding measurement errors. Segments measured from the end of previous
measured segments can create potential location errors depending on where a camp may
have been located. For instance, if a camp is discovered within one measured segment,
segments beyond the newly discovered camp site should be adjusted based on that
finding and would result in a different set of potential locations. If no adjustments are
made after discoveries, this error would compound itself given the long stretches of
railroad grades between some camps (e.g. Wedekind to B-Line Camp).
A final potential complication related to the specific methods used in this research
is the misidentification of camp sites. Using camps that are either unproven to be camps,
such as Sherman Valley Campground and Wedekind, or, like the D-Line Ivy Spot, that
have not been sufficiently surveyed, can affect the average distance measurements. Also,
understanding the human disposal patterns discussed by Schiffer (1983 & 1986) could
lead to changes in camp locations. A site recorded based on the location of a midden
may not be an actual camp site, but rather a dump site. This could be the case with
Goliath Creek Camp since common logging camp artifacts such as dishware and other
consumables were not discovered. Misidentification of camp sites would affect
calculated average distances and cause shifts in the distance segments identified as
potential logging camp sites.

91

Chapter 5 – Results and Discussion
Results
Field verifications of distance segments were inconclusive, failing to definitively
identify any historical logging camp locations. Artifacts dating from the early twentieth
century and consistent with logging camps of that era were found within some of the 36
aggregate distance segments and within close proximity to others (Table 6; Appendix A).
Aggregate segments are clumps of measured distance ranges located in the same general
area.
Variable

2

Factors

Limiting

Artifacts

Total
Near Rail Junction
Near Stream/ Wide Floodplain
Impounded Water Presence
Current Campground
Near Major Landscape Manipulation
Exotic Plants
Consumables
Metal Pieces
Steam Donkey
Water Pipes
Structural Remnants

Within
Segments
36
20
21
4
4
7
4
5
5
1
2
1

Near Property Line

17

Close in Geodesic Distance

10

3

Nothing found

Outside
1

Segments

3
1
2

29

Table 6. Field verification findings in aggregate segments. 1Cells
without a value do not apply. 2Four aggregate segments were both near
property lines and close in geodesic distance. 3Includes segments where
artifacts were found outside, but close to that segment.

As previously mentioned, artifacts such as bottles, earthenware or china, and
enamelware can be the best indicators of site type. These types of artifacts were found
within five aggregate segments (A1, E1, E2, J5, and L3); however, artifacts found in one

92

segment (L3) were determined to be from after Mason County Logging Company
(MCLC) had shut down operations in Capitol Forest. Other findings discovered within
aggregate segments, which were similar to those artifacts and features discussed in
previous chapters, lead to more potential logging camp sites. In total, artifacts consistent
with historical logging camps in Capitol Forest were found within seven separate
aggregate segments (A1, E1, E2, E3, J5, J6, and L7) and within close proximity to five
additional aggregate segments (B1, D1, F1, J2, and M1). Specific findings, both within
and just beyond aggregate segments, are covered in more detail below.
Results in relation to the three different methods used for measuring average
distance were very close, likely due to the small differences among averages (Table 7;
Appendix A). Artifacts and features found within aggregate segments were often found
within segments of all three measurement methods; however, more artifacts and features
were found within extent segments. These findings are discussed in more detail below.

Extent
41
4
6
4
1
1
2

42
1
4
2
1
1
2

Near Property Line

19

21

21

Close in Geodesic Distance

12

10

11

32

35

37

Variable

Factors

2

Artifacts

Total
Exotic Plants
Consumables
Metal Pieces
Steam Donkey
Water Pipes
Structural Remnants

Limiting

Bordeaux
Central Outside
Point / Segments1
Extent
44
1
5
3
3
1
1
1
2
2

Central
Point

Nothing found3

Table 7. Field verification findings among three measurement methods. 1Cells
without a value do not apply. 2Four segments of each type were both near property
lines and close in geodesic distance. 3Includes segments where artifacts were found
outside, but close to segments.

93

Aggregate Segment Findings
Aggregate Segment A1 Findings
An enamelware pitcher was found within all three types of measured segments
within aggregate segment A1 (Figures 34 & 35; Ferguson, 2011F). The pitcher was
found along the edge of a beaver pond adjacent to the original railroad grade. The beaver
dam failed in 2011, releasing the impounded water. It remains possible more artifacts
could be uncovered as the sediment of the former pond is scoured. This pitcher is
unlikely to be an isolate because it is not an item typically discarded at a random location
like a broken dish or bottle.

Figure 34. Area of aggregate segment A1 where artifact was found. Note
location near junction and impounded water.

94

Figure 35. Enamelware pitcher found
within segment A1. Photograph by author
(Ferguson, 2011F).

Some logging cable was also discovered near another junction and large
impounded body of water at the southern portion of segment A1, but could be related to
logging activity only. It remains a possibility that artifacts also exist within this southern
body of water as it may not have been full of water during the time the forest was initially
harvested. This is because the adjacent stream has been altered downstream of the site,
likely causing changes in flow patterns. If the pond did exist at the time of original
harvest, the southern pond may have been used as a dump site similar to what was found
at the Hollywood family camp site (Ferguson, 2011A).
Aggregate Segment B1 Findings
A portion of a 1.5-inch iron water pipe was found during a survey by Vaughn
(2013) approximately 1,800 feet from the ends of extent and Bordeaux central point
measured segments (Figure 36). By itself, this water pipe does not indicate a logging
camp site. Given the location of this pipe at the end of a rail line, it may be more likely
this pipe was related to logging or railroad activity only.

95

Figure 36. Findings near aggregate segment B1.

Aggregate Segment D1 Findings
A cultural resource survey completed prior to a culvert removal uncovered a trash
accumulation consisting of ceramic, glass, and metal fragments during shovel tests,
where holes are dug to see if artifacts exist in the earth. Located just over 2,000 feet from
the northern edge of segment D1 in a wide floodplain of a major creek (Mill Creek), this
trash midden was from the early twentieth century and consistent with logging camps of
that era (Figures 37 & 38). The nearest distance range within segment D1 was based on
Bordeaux central point measurements. Surveys of the areas between the midden site and
the end of segment D1 failed to find any signs of habitation. This location is now a major
road junction, and it is possible the road work destroyed any signs of a former camp.
This trash site is likely too far from the town of Bordeaux or the Art Karlen homestead
(site of the current Cedar Creek Correctional Camp) to have been related to either.

96

Figure 37. Location of midden in relation to aggregate segment D1.

Figure 38. Ceramic sherds discovered during shovel tests north of aggregate
segment D1. Photograph by Maurice Major.

97

Aggregate Segment E1 Findings
In the initial survey used to record the railroad grades, a whiskey bottle dating
from between 1910 and 1920 was found near a bridge crossing within each of the three
measured segment types (Figure 39; Christopherson, 2009A). Christopherson (2009A)
also found the skids from a steam donkey outside this aggregate segment, approximately
1,000 feet west of a central point segment. Other surveys found a small china fragment,
also known as a sherd, and some metal pieces, likely related to logging or railroad
equipment, near a major junction within extent and Bordeaux central point measured
segments and roughly 400 feet from a central point measured segment. It is possible the
sherd and bottle are isolate artifacts; however, aggregate segment E1 is located in what
was deemed a high probability location because it is located in or near the floodplain of a
major stream, Monroe Creek, and near impounded water and a major rail junction.

Figure 39. Findings in aggregate segment E1 with donkey skids to
the west, china and metal at south, and bottle in the northeast.

98

Aggregate Segment E2 Findings
A ceramic insulator, a china sherd, iron pieces related to railroads, and an apple
tree were all found within all three measurement variations in aggregate segment E2
(Figures 40 & 41). Collective segment E2 was initially noted as having a high
probability for a logging camp location because it is located in or near the floodplain of
two major creeks, Falls Creek and Sherman Creek. A portion of this segment is also
located in the present-day Falls Creek Campground. As discussed at length, multiple
current and former campgrounds in Capitol Forest were at one time logging camp
locations (Ferguson, 2011A, 2011B, & 2012B). It is possible the insulator and fruit tree
are related to an attempted homestead or recreational activity, but the iron pieces and
china sherd add more weight to the site having been a logging camp.

Figure 40. Findings within aggregate segment E2. Findings are
within Falls Creek Campground.

99

Figure 41. Findings within
collective segment E2.
Ceramic insulator (A), china
sherd (B) found slightly
downstream of other findings,
and fruit tree (C), likely an
apple. Photographs by author

100

Aggregate Segment E3 Findings
Two features within a portion of aggregate segment E3 are important
characteristics when compared to known logging camp sites. A standing portion of a
railroad trestle, approximately 12 to 15 feet in height, and fallen remnants of the same
trestle were found at a large stream crossing within extent and central point distance
segments (Figures 42 & 43). Just past the southern end of the former trestle is an
approximate one-twentieth acre patch of English ivy (Hedera helix) growing on at least
seven trees (Figure 44). As noted, former camps have been located near large trestles and
the areas where trestles crossed can be valuable locations for finding artifacts; however,
no artifacts other than those related to the former trestle were discovered. Likewise,
English ivy can be a sign of past habitation, but no other artifacts or features were found
to identify the site as a logging camp. The grade where the trestle was located was not
originally known prior to field verifications as it was not shown on the MCLC map of
grades and was not clearly visible on LiDAR imagery.

Figure 42. Map of findings in aggregate segment E3. Star to the
north is the trestle location.

101

Figure 43. Looking south from end of grade at the standing portion of trestle remnants
found in aggregate segment E3. Photograph by author.

Figure 44. English ivy found along grade within segment E3. Photographs by author.

102

Aggregate Segment F1 Findings
Numerous glass shards, including manganese glass, which was manufactured
from the early 1800s to 1916, were found approximately 400 feet past the end of a central
point measured segment within aggregate segment F1 (Figures 45 & 46). The location of
artifacts found near aggregate segment F1 are likely from MCLC Camp 1, which was
detailed by Blum (2000). Even after multiple field visits, the exact location of MCLC
Camp 1 has yet to be determined and may be hidden in the dense brush. The fact that a
known camp is located relatively close to, if not within, a distance segment may be
evidence that camps were indeed constructed at or near a specific distance from other
camps.

Figure 45. Findings near collective segment F1.

103

Figure 46. Example glass fragments found near aggregate segment F1. At left is a manganese
glass shard from the Western Paint & Glue Company; manganese glass was manufactured from
early 1800s to 1916. At right is a bottle manufactured by the Whitall Tatum Company likely
between 1870 and 1901 (BLM & SHA, 2014). Photographs by author.

Aggregate Segment J2 Findings
A 2-inch iron water pipe was found coming out of the ground along a short rail
grade above an approximate 32-foot deep through cut. The pipe is located roughly 1,300
feet outside an extent measured segment of collective segment J2 (Figures 47 & 48;
Ferguson, 2012C). Similar pipes were found in the Lower Incline Camp and MCLC
Camp 7 (Stilson 2010D; Ferguson, 2010A). A cherry tree was also found approximately
50 feet south of the water pipe location. Although cherry trees occur naturally
throughout the forest, the location of this particular tree is suspicious in relation to the
grades and pipe. This location is also near a large flat in a topographic saddle; however
thick understory brush hindered a more thorough investigation of the saddle.

104

Figure 47. Findings near aggregate segment J2.

Figure 48. The 2-inch water pipe found in situ on a side grade above a deep
through cut near segment J2. Photograph by author (Ferguson, 2012C).

105

Aggregate Segment J5 Findings
Surveys of the segments in combined segment J5 found a few artifacts leading to
an inference that a camp may have been present. A piece of brick was found on the grade
within all distance measurement variations in the western portion of the aggregate
segment (Figure 49; Ferguson, 2010D). This brick piece is likely an isolate or unrelated
to historical activity. Surveys in the eastern portion of segment J5 found a small china
sherd and a small patch of English ivy within the current Porter Creek Campground
(Figure 50). Both the ivy and the broken china could, again, be related to recreational
activities, but when the history of campgrounds within the forest is considered, it seems
logical to infer that Porter Creek Campground may have been a logging camp. The
campground is only located within extent measured segments.

Figure 49. Map of findings within aggregate segment J5.

106

Figure 50. View of English ivy as seen looking south within present-day Porter Creek
Campground. China fragment was found to the east of this location in Porter Creek.

Aggregate Segment J6 Findings
Multiple features were discovered within all three measurement types in
aggregate segment J6 (Figure 51). Parallel to the grade near a large beaver pond are the
remnants of two sets of steam donkey skids. These two sets of skids are less than 50 feet
apart and both sets consist of logs roughly 35 to 40 feet in length and 36 to 40 inches in
diameter. Multiple 1-inch threaded iron posts spaced approximately three feet apart were
found protruding from each skid (Figures 52 & 53). Also found attached to each set of
donkey skids was a large iron washer or wheel (Figure 53) that was nine inches in
diameter for the northern skid set and seven inches for the southern set. Deteriorated
metal remnants, possibly roofing material, were found between the northern set of skid
remnants. The northern set of skid remnants also had a shorter log connecting the two

107

skids at one end; this was a common feature of steam donkeys. Strands of 1.5 and 2-inch
logging cable were also found with both sets of skids.
Other artifacts found within collective segment J6 include an abundant amount of
1, 1.5, and 2-inch logging cable adjacent to the large beaver pond. A 1-inch pipe was
found crossing the grade just to the south of the donkey skids (Figure 54). The alignment
of the pipe appeared to be coming directly from the large pond. Pieces of corrugated
metal, often used for roofing on both steam donkeys and structures, were discovered
along the edges of the beaver pond as well. A piece of corrugated metal was also found
approximately 1,200 feet from the skids and water pipe. This particular piece of metal
may be unrelated to historical logging or the similar pieces found near the skid remnants;
however, given its proximity to artifacts found near the pond, there is a high likelihood
the isolate piece of corrugated metal is related to historical logging operations.
Although no consumable type artifacts were found within segment J6, it remains
possible that a logging camp may have existed in this area. The two sets of donkey skids
aligned parallel to the railroad grade could lead to an inference that this site was used as a
steam donkey construction site similar to other camps in the forest; the D-Line Ivy Spot
and Wedekind are both believed to have had steam donkey construction operations
associated with a camp (Blum, 2000; Ferguson, 2012A). The skids are aligned as such to
allow them to be loaded onto rail cars and shipped to other locations; however, this could
also mean the two sets of skids were dumped at this location after breaking down.

108

Figure 51. Findings within collective segment J6. The two stars to the west
of the grade and beaver pond are the two donkey skids with the 1-inch pipe
located just to the south of them. Far west star is isolate corrugated metal.

Figure 52. The southern donkey skid as seen looking east towards the beaver
pond. Arrow designates a 1-inch diameter iron rod and strands of 1 to 1.5inch logging cable to the left of young tree. Photograph by author.

109

Figure 53. Photographs of northern donkey skids. Looking north along skid (top left) with stars
indicating locations of some of the numerous 1-inch diameter iron posts (top right) protruding
from skid. Southern end of northern skids (bottom left) with 9-inch diameter iron disc (also
bottom right) visible with perpendicular log across skids. Portion of 2-inch diameter logging
cable near machete is visible in bottom left. Photographs by author.

110

Figure 54. Image of the 1-inch water pipe
found on grade and now under water (left).
The arrow (right) designates the location and
orientation of the 1-inch water pipe, which may
be coming from the existing pond.
Photographs by author.

After reviewing 1941 aerial photographs of the area, a third reason for artifacts
found within aggregate segment J6 became clear (Figure 55). MCLC harvested and
milled their last remaining timber in 1941; in the aerial photograph, small stands of
mature timber are located immediately adjacent to the steam donkey locations. These
stands of timber likely represent those final trees harvested by MCLC. Scars on the
surrounding landscape show both very recent logging activity to the south as well as
areas that were harvested much earlier to the north. This is likely because areas to the
north were more easily accessible with Porter Creek, a major stream system, providing
topography more suitable for train movement while the area of segment J6 is located
higher in the hills of the forest which took longer for MCLC to gain access to. Another
interesting finding is that the skid roads, which are the trails created by logs being pulled
from where they were cut to the locations they were to be loaded onto train cars, appear
to be converging precisely at the location of the steam donkey remnants. From this
111

evidence, it seems clear that this location was not a logging camp, but was indeed the last
area logged by MCLC and these steam donkeys were left where they were last used after
parts of value were removed, leaving only the skids, metal, and logging cable.

Figure 55. Portion of a 1941 aerial photograph showing aggregate segment J6. Note mature
timber in the immediate vicinity as well as the skid roads (white lines appearing to create fan
shapes) converging at the location of the steam donkey remnants (indicated by star).
Wedekind is also shown at the bottom right of photograph.

Aggregate Segment L7 Findings
Segment L7 is located at the northwestern extent of MCLC ownership and
seemed unlikely to be a potential camp site; however, the segment is located near a major
rail junction and in close proximity to a large pond (Figure 56). Only one potential
artifact was discovered during field verifications; what appeared to be a metal stove pipe
or chimney piece was found partially buried in a dug out area immediately adjacent to the
grade within an extent measured segment (Figure 57). This area was likely excavated to

112

provide fill material for the grade. The metal artifact is approximately 8 inches in
diameter and extends roughly another foot into the ground.
Found throughout the immediate vicinity of the metal artifact was an abundance
of holly (Ilex aquifolium). This non-native species can be spread easily by birds and may
not be indicative of habitation; however, the sheer number of plants and one large
specimen found near the main junction of grades to the southwest of the metal artifact
could prove otherwise (Figure 58). The metal artifact may prove to be unrelated to
historical logging activity in the forest as some recent trash was also found in the
excavated trench. A large flattened area was discovered directly east of the metal artifact.
This flat was not immediately adjacent to any grades or roads and trees growing on the
flat were much younger than the adjacent stand. More investigation in the area of the
trench and flat should be completed.

Figure 56. Aggregate segment L7. Star represents location of the metal
artifact found in a trench adjacent to the grade.

113

Figure 57. Metal artifact found adjacent to the grade in segment L7. Photograph by
author.

Figure 58. Large holly discovered on grade in segment L7. Photograph by author.

114

Aggregate Segment M1 Findings
A fragment from a ceramic coffee cup was found approximately 1,550 feet north
of an extent measured segment in aggregate segment M1 (Figure 59). The area where the
sherd was found has flat topography near a rail junction; however, no other artifacts were
discovered in a field survey, leading to the inference of this artifact being an isolate.
Small dug out holes approximately three feet in diameter and roughly a foot deep were
found next to the grade as well. This artifact was found approximately 7,000 feet from
Bozy Creek Camp, putting it close in geodesic distance or straight line distance from that
camp, providing further evidence of this artifact being an isolate.

Figure 59. Coffee cup fragment finding in relation to aggregate segment M1.

115

Discussion
It was assumed from the beginning of this research that not all identified segments
would contain a logging camp site because it is unrealistic for the given area. A lack of
any definitive camp location identification was, however, unexpected. More
investigations should be done in the areas where artifacts and features consistent with
historical logging camps were discovered. Some of these areas are located in mature
timber stands which are available to be harvested; if harvest occurs in these areas, field
surveys should be completed following harvest activities. Future surveys could also
include a metal detector for those areas where brush was too thick to complete an
adequate survey at the time of this research.
It remains possible that camps existed within the segments where artifacts were
found in close proximity because it was common for trash to be dumped outside the
central living area. As discussed in the literature review, Schiffer (1983 & 1986) points
out accumulations of artifacts are subject to various formation processes in regard to their
positioning. Artifact collections may not pinpoint a habitation site, but rather a dump site
based on human disposal practices (Schiffer, 1983 & 1986; Paullin, 2007). For instance,
middens were found approximately 500 feet from MCLC Camp 4 and 700 feet from the
Lower Incline Camp (Stilson, 2009; Ferguson, 2010A). With an artifact accumulation as
close as 400 feet, measured segments inside collective segment F1 could potentially have
contained a logging camp. Measured segments within J2, M1, and D1 aggregate
segments, at more than 1,300, 1,500 and 2,000 feet respectively from artifact
accumulations, are unlikely to have contained logging camps that could have generated
the discovered artifacts. The water pipe found roughly 1,800 feet from the edge of

116

aggregate segment B1 is more likely to be related to logging or railroad operations than a
camp. More thorough investigations in each aggregate segment where artifacts, features,
or both were found within or near will deliver more insight into whether a logging camp
existed within those segments.
Potential complications discussed in Chapter 4 may have led to the overall low
success rate and could be accounted for in future testing of this theory. Twenty-nine, or
75 percent, of combined segments contained no artifacts (Table 6; Appendix A).
Approximately 47 percent of the aggregate segments are within close proximity to past
MCLC property lines and nearly 28 percent are close to known or assumed camps in
geodesic distance. It should be noted that three aggregate segments originally considered
to be close in geodesic distance or near former property lines contained artifacts or
features that may be related to historical logging activities or camps. Because no logging
camp sites were definitively identified, segments were not adjusted and compound
measurement errors could not be identified.
Using assumed camp locations was another complication related to the research
methods. Using the assumed camp locations of Sherman Valley Campground and
Wedekind did not appear to cause an error, but would have produced different results.
Completing average distance calculations based only from measurements between known
logging camp sites produced similar average distances for each measurement method;
however, the distance ranges based on adding or subtracting one standard deviation were
smaller (Table 8). Shorter distance ranges would create a different set of segments to
verify as well as different results. Roughly plotting segments based on the average
distances between only known camps did not improve results. Sherman Valley

117

Campground fell within all three measurement methods based on calculating distances
between known camps only; however, Wedekind was not located within any distance
segment calculated using only known camps. Removing distances calculated using
assumed camps reduced the amount of calculations from nine to four, possibly limiting
accuracy of predicting an average distance for MCLC. Although the use of assumed
camp locations did not affect the outcome of this research, using as many locations as
possible to provide more distances for determining a company specific average would be
preferred and should produce more accurate predictions for potential logging camp sites.
Measurement Method
Camp Extent
Central Point
Bordeaux Central Point

With Assumed
Average
Range
12078.65 4430.00
12941.71 3954.34
12477.09 4037.10

Without Assumed
Average Range
11250.19 3133.72
12401.59 2448.60
12146.67 2547.51

Table 8. Average distances and standard deviation ranges for each
measurement method calculated with and without using the assumed logging
camp locations of Sherman Valley Campground and Wedekind.

The potential limiting factors discussed in Chapter 4 also hindered more thorough
site surveying. As mentioned earlier, many old railroad grades have been converted into
roads or trails. Conversion of grades to roads can result in minor impacts to the integrity
of those railroad grades; however, widening, placing rock, or paving grades can lead to a
loss of historical artifacts or features. This was known to have happened near the D-Line
Ivy Spot (Blum, 2000). Creating trails on railroad grades also has minimal impacts to the
integrity of the grade, but can open sites up to looting and light damage from all-terrain
vehicles. Also, while many railroad grades go up drainages, roads and trails often cut
across drainages; misidentification of roads and trails as railroad grades while
interpreting LiDAR data could lead to errors in distance calculations.

118

Much of Capitol Forest is managed by the Washington State Department of
Natural Resources (WADNR) for timber production with few areas unavailable for
harvest. A number of segments surveyed were located in forest stands with abundant
understory brush, making surveying the ground difficult. Most field verification was
completed during the winter in an attempt to visit sites when the foliage of shrub species
was at a minimum, but species such as salal (Gaultheria shallon) and sword fern
(Polystichum munitum), with foliage persisting through the winter, were common among
all segments. A large number of segments were also along major streams where foliage
can be especially dense. As discussed in Chapter 2, timber harvest can have minimal
impacts on sites and can be beneficial to archaeologists by uncovering new artifacts.
Areas where artifacts and features were found as part of this research should be surveyed
following any future harvest to look for signs of past habitation or additional artifacts.
Total findings based on the three measurement methods were within five percent
of each other and there appears to be no significant difference among those methods.
Segments based on calculating average distance from the edges of logging camp extents
did produce more results than the other two methods. This could be because the
calculated range of extent segments was greater and a larger search area should yield
greater results. From these results, identifying and measuring from the edge of camp
extent may provide the best results due to longer segments; however, this may vary
depending on site location and logging company.
The probabilities assigned to each aggregate segment prior to field verifications
may have been successful (Table 9; Appendix A). As noted in the methods section
(Chapter 4), probabilities were assigned based on remote site evaluations taking into

119

account the factors common among known logging camp sites. Although originally
considered arbitrary, site probability assignments appear to have been a good indication
of potential camp locations.
Probability
High
Medium
Low

Total
4
14
18

Findings1
3
8
1

No Findings
1
6
17

Table 9. Success of probabilities assigned to aggregate segments. 1Findings include
artifacts and features consistent with historical logging camps discovered within and
outside all aggregate segments except L3 where artifacts found dated to after MCLC
operations had ended in the forest.

Comparing site characteristics shared among known camp sites in conjunction
with improved methods for calculating the average distance between camps could prove
valuable in future attempts to locate historical logging camps. Success based on
comparing the shared site characteristics could lead to the conclusion that spatial
modeling based on those common characteristics is the most promising method for
locating camps.

120

Chapter 6 – Conclusion
In the case of the Mason County Logging Company (MCLC), distance from other
camps along rail lines may not have been a determining factor for logging camp
locations. In regards to this research, average distance in relation to camp spacing of the
known and assumed MCLC camps appears to be coincidental and logging camp location
may be related to other factors such as topography, proximity to water resources, and
possibly land ownership. Although no logging camp locations were definitively
identified, artifacts and features found within some segments and within close proximity
to others may be evidence of how distance between camps was an important factor for
camp placement. More thorough investigation of areas where artifacts were discovered,
including shovel testing or metal detecting, could determine if there were logging camps
in those locations or if the artifacts discovered were isolates. Also, there was no obvious
difference among the three measurement methods used. This is likely due to the three
methods having relatively small differences in their calculated average distances and
standard deviation ranges.
This research did show that there are numerous site characteristics common
among historical logging camp sites; this finding could be useful for archaeologists
attempting to locate these sites in the future. Variables identified during this research
appearing to be common among logging camp sites include being located near rail
junctions, major streams, impounded water, and landscape manipulations (e.g. large cuts,
fills, or trestles), as well as sites that continue to have forms of habitation such as
campgrounds. Each of the known and assumed logging camp sites used in this research

121

were located near two to four of these variables, demonstrating the significant role these
characteristics may have in predicting sites.
Other factors beyond those variables discussed above likely influenced camp
locations in different geographic areas based on how long it took to harvest a given area.
Stand species composition, tree size, and harvest technology could all affect the duration
a camp was in use. If a logging company was primarily interested in harvesting the most
valuable timber species out of a mixed stand, or high-grading, that company may move
through a given area more quickly than if they were to harvest the entire area. Similarly,
if trees in an area are much larger in diameter than another area (e.g. western Washington
Douglas-fir (Pseudotsuga menziesii) and western redcedar (Thuja plicata) versus eastern
Washington pine species), it would take longer to fall, cut-to-length, and remove the
larger trees from the forest. Finally, as harvest technology improved from axes to crosscut saws and then to chainsaws, the amount of time spent falling timber would be greatly
reduced. This technology change would also reduce the amount of time spent logging in
a given area and the need for additional logging camps. The same point could be made
about techniques to yard or remove logs from a site; as methods progressed from horse
and oxen, to steam donkeys and trains, to log loaders and trucks, the need for logging
camps would be reduced. All of these factors should be accounted for when attempting
to predict where logging camps may have been located in different geographic areas.
The methods used for this research could be improved to possibly provide more
accurate predictions. Improving techniques in regards to calculating distance and plotting
the resulting distance segments may increase accuracy. Utilizing the network analyst tool
in ArcMap could potentially accomplish this predictive analysis as well as account for

122

other factors such as time of travel and direction. Even though using assumed camp
locations to calculate average distance measurements did not appear to impact the results
of this research, using only known camp locations would likely provide the most accurate
predictions. As far as improvements to field verifications, winter appears to be the best
time to survey potential sites due to low foliage levels. Where applicable, surveys
following timber harvest provide greater results due to removal of the ground cover
species and the low impacts on artifacts from harvest activity (Christopherson, 2008;
Stilson, 2010C & 2011).
There remains more opportunities to test this average distance theory due to the
number of large logging operations that took place throughout the United States during
the railroad logging era. One specific area where the methods used in this research could
be attempted are the areas formerly managed by the Cascade Lumber Company north of
Cle Elum, Washington, as discussed in The Pine Tree Express (Henderson, 1990). In this
book, Henderson (1990) notes the locations of a significant number of logging camps on
hand drawn maps (pp. 100-05); nearly all of these camps remain unverified and
unrecorded. These camps can be roughly digitized based on the hand drawn maps and
the railroad grades verified with the use of LiDAR once it becomes available for the area.
Once an adequate number of camp locations have been verified, average distance
calculations could be completed and compared to the remaining unverified camp
locations to see if those camps align with that average distance prior to additional field
verifications.
Locating historical sites such as logging camps could become easier to do as
mapping and spatial analysis technology improves. As noted in the literature review,

123

Hare et al. (2014) achieved success using high resolution LiDAR data, similar to
archaeologists who have used LiDAR to locate sites concealed by dense jungles in
Central and South America. These recent successes demonstrate how improvements to
LiDAR quality and resolution will further the abilities of archaeologists to map and
locate undocumented sites. Altering the methods used in this thesis to determine distance
between camps along with improved LiDAR data may provide more definitive results;
however, spatial modeling based on common logging camp site characteristics, which
can also account for issues such as geodesic distances, may prove to be more successful
in determining potential historical logging camp sites.
During the major exploitation era of logging in America, thousands of men,
women, and children lived in various logging towns and satellite camps associated with
those towns. Locating and documenting historical logging camps is essential for
understanding the lives of the people who called these towns and logging camps their
home. The best method for identifying historical logging camp locations is to research
historical maps and documents; however, due to their ephemeral nature, logging camps
often went undocumented. In this situation, it can be advantageous to locate a person
with direct knowledge of where camp sites may have been. Unfortunately, losses of the
people with first- or second-hand knowledge of camp locations will occur with more
frequency in the future. In the absence of historical documentation and local knowledge,
it is imperative a method be established to locate historical logging camps for the purpose
of collecting and preserving the data from these sites and to further our understanding
about the lives of early-American loggers and their families.

124

References
Andrews, R. W. (1968). Timber: Toil and Trouble in the Big Woods. Schiffer
Publishing, 181 pages.
Andrews, R. W. (1956). Glory Days of Logging. Superior Publishing Company, 176
pages.
Andrews, R. W. (1954). This Was Logging! Superior Publishing Company, 157 pages.
Ayres, J. E. (1996). Standard Timber Company Logging Camps on the Mill Creek
Drainage, Unita Mountains, Utah. Proceedings of the Society for California
Archaeology. 1996, Vol. 9, pp. 179-182.
Barceló, J. A., & Pallarés, M. (1996). A Critique of GIS in Archaeology. From Visual
seduction to Spatial Analysis. Archeologia e Calcolatori n. VII, pp. 1-32.
Baldo, C., & Coombs, R. (1991). Transcript of Interview with Harlan Smith; former
engineer for the Mason County Logging Company. Retrieved August 24, 2011,
from: http://www.rootsofmotivepower.com/html/Collection/Mason7.html
Blum, G. (2000) Capitol Forest Historical Research Manuscript. On file at the
Washington State Department of Natural Resources, Olympia, Washington.
Boire, K. & Stilson, M. L. (2006). Bordeaux Mill Site Archaeological Site Inventory
Form. On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Brandt, R. Groenewoudt, B. J., & Kvamme, K. L. (1992). An Experiment in
Archaeological Site Location: Modeling in the Netherlands using GIS
Techniques. World Archaeology, Vol. 24, No. 2, Analytical Field Survey, pp.
268-282.
Brashler, J. G. (1991). When Daddy was a Shanty Boy: The Role of Gender in the
Organization of the Logging Industry in Highland West Virginia. Historical
Archaeology, 25(4), pp. 54-68.
Bureau of Land Management & Society for Historical Archaeology (BLM & SHA)
(2014). Historic Glass Bottle Identification & Information Website, found at:
http://www.sha.org/bottle/index.htm
California Department of Transportation (CALTRANS). (2013) Work Camps: A
Historical Context and Archaeological Research Design for Work Camp
Properties in California. The California Department of Transportation
Sacramento, California, 280 pages.
125

Carlson, L. (2003). Company Towns of the Pacific Northwest. University of
Washington Press, 286 pages.
Christopherson, R. (2009A). Mason County Logging Company Monroe Creek Railroad
Grades Archaeological Site Inventory Form. On record at the Washington
Department of Archaeology and Historic Preservation, Olympia, Washington.
Christopherson, R. (2009B). Mason County Logging Company South Fork Porter Creek
Grade Archaeological Site Inventory Form. On record at the Washington
Department of Archaeology and Historic Preservation, Olympia, Washington.
Christopherson, R. (2008). Mason County Logging Company North Fork Porter Creek
Railroad Logging Camp Archaeological Site Inventory Form. On record at the
Washington Department of Archaeology and Historic Preservation, Olympia,
Washington.
Coman Jr., E. T. & Gibbs, H. M. (1949). Time, Tide and Timber: A Century of Pope and
Talbot. Stanford: Stanford University Press, 480 pages.
Conners, P. A. (1990). Stanislaus National Forest, Tuolumne County, California,
Comprehensive Evaluation Plan for the West Side Discontinuous District. U.S.
Department of Agriculture, Forest Service, Stanislaus National Forest, Sonora,
CA.
Cox, T. R. (2010). The Lumberman's Frontier: Three Centuries of Land Use, Society, and
Change in America's Forests. Oregon State University Press, 560 pages.
De Meo, A. A., Espa, G. G., Espa, S. S., Pifferi, A., & Ricci, U. (2000). Study of
Archaeological Areas by Means of Advanced Software Technology and
Statistical Methods. Journal of Cultural Heritage, 1, pp. 233-245.
De Reu, J. D., Bourgeois, J., De Smedt, P., Zwertvaegher, A., Antrop, M., Bats, M., De
Maeyer, P., Finke, P., Van Meirvenne, M., Verniers, J., Crombe, P. (2011).
Measuring the Relative Topographic Position of Archaeological Sites in the
Landscape, a Case Study on the Bronze Age Barrows in Northwest Belgium.
Journal of Archaeological Science, 38, pp. 3435-3446.
Ebert, D. (2004). Applications of Archaeological GIS. Canadian Journal of Archaeology,
28(2), pp. 319-341.
Espa, G. G., Benedetti, R. R., De Meo, A. A., Ricci, U. U., & Espa, S. S. (2006). GIS
Based Models and Estimation Methods for the Probability of Archaeological Site
Location. Journal of Cultural Heritage, 7(3), pp. 147-155.

126

Felt, M.E. (1975). Capitol Forest…The Forest That Came Back. Washington Department
of Natural Resources, 76 pages.
Ferguson, P. J. (2012A). Wedekind Donkey Remnants Archaeological Site Inventory
Form. On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Ferguson, P. J. (2012B). Wedekind Planting Camp Archaeological Site Inventory Form.
On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Ferguson, P. J. (2012C). C-2700 Railroad Grade Archaeological Site Inventory Form.
On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Ferguson, P. J. (2011A). Hollywood Archaeological Site Inventory Form. On record at
the Washington Department of Archaeology and Historic Preservation, Olympia,
Washington.
Ferguson, P. J. (2011B). Mason County Logging North Creek Camp Archaeological Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
Ferguson, P. J. (2011C). Mason County Camp 2 Archaeological Site Inventory Form.
On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Ferguson, P. J. (2011D). Wedekind Dam Archaeological Site Inventory Form. On
record at the Washington Department of Archaeology and Historic Preservation,
Olympia, Washington.
Ferguson, P. J. (2011E). Goliath Creek Grade and Midden Site Inventory Form. On
record at the Washington Department of Archaeology and Historic Preservation,
Olympia, Washington.
Ferguson, P. J. (2011F). Mason County Logging Company Railroad Grade Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
Ferguson, P. J. (2010A). Incline-Low Camp Archaeological Site Inventory Form. On
record at the Department of Archaeology and Historic Preservation, Olympia,
Washington.

127

Ferguson, P. J. (2010B). D-3000 Camp Archaeological Site Inventory Form. On record
at the Washington Department of Archaeology and Historic Preservation,
Olympia, Washington.
Ferguson, P. J. (2010C). Waddell Creek Railroad Grade and Trestle Archaeological Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
Ferguson, P. J. (2010D). A-Line/Porter Falls Railroad Grade Archaeological Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
Finke, P. A., Meylemans, E. E., & Van de Wauw, J. J. (2008). Mapping the Possible
Occurrence of Archaeological Sites by Bayesian Inference. Journal of
Archaeological Science, 35(10), pp. 2786-2796.
Franzen, G. (1992) Northern Michigan Logging Camps: Material Culture and Worker
Adaptation on the Industrial Frontier. Historical Archeology2 6(2), pp. 74-98.
Hannum, J. S. & Hannum, C. B. (2006) South Puget Sound Railroad Mania. Olympia,
Wash.: Hannum House Publications, 284 pages.
Hannum, J. S. & Hannum, C. B. (2002). Gone but not Forgotten: Abandoned Railroads
of Thurston County, Washington. Olympia, Wash. Hannum House Publications,
256 pages.
Hardesty, D. L., & Little, B. J. (2000). Assessing Site Significance: A Guide for
Archaeologists and Historians. Walnut Creek, CA: AltaMira Press, 184 pages.
Hare, T., Masson, M., & Russell, B. (2014). High-Density LiDAR Mapping of the
Ancient City of Mayapán. Remote Sensing, 6(9), pp. 9064-9085.
Henderson, E. M. (1990). The Pine Tree Express. A History of the Cascade Lumber
Company’s Pine Hauling Railroad in Kittitas County, Washington 1916-1946,
128 pages.
Hoglund, W. A. (1960). Finnish Immigrants in America, 1880-1920. University of
Wisconsin Press, Madison, 213 pages.
Judge, W. J., & Sebastian, L. (eds.) (1988). Quantifying the Present, and Predicting the
Past: Theory, Method, and Application of Archaeological Predictive Modeling,
U.S. Bureau of Land Management, Department of the Interior, Denver, CO, 669
pages.

128

Kamholz, E. J., Blain, J., & Kamholz, G. (2003). The Oregon-American Lumber
Company: Ain’t No More. Palo Alto, Calif.: Stanford University Press, 384
pages.
Kantner, J. (2008). The Archaeology of Regions: From Discrete Analytical Toolkit to
Ubiquitous Spatial Perspective. Journal of Archaeological Research, 16(1), pp.
37-81.
Kvamme, K. L. (1999). Recent Directions and Developments in Geographic Information
Systems. Journal of Archaeological Research 7, pp. 153–201.
Kvamme, K. L. (2006). There and Back Again: Revisiting Archaeological Locational
Modeling. In Mehrer, M., and Wescott, K. (eds.), GIS and Archaeological Site
Location Modeling, Taylor and Francis, Boca Raton, FL, pp. 3–38.
Karamanski, T. J. (1985). Logging, History, and the National Forests: A Case Study of
Cultural Resource Management. The Public Historian, 7(2), pp. 27–40.
Labbe, J. T., & Goe, V. (1961). Railroads in the Woods. Berkeley, California: HowellNorth, 269 pages.
Lasaponara, R., & Masini, N. (2013). Remote Sensing in Archaeology: An Overview.
Journal of Aeronautics & Space Technologies, 6(1), pp. 7-17.
Maniery, M. L., Baker, C., Syda, K., & Hamusek, B. (1996). National Register of
Historic Places Evaluation of 15 Historical Sites, Shasta-Trinity National Forest,
California. Manuscript, U.S. Department of Agriculture, Forest Service, ShastaTrinity National Forest, Redding, CA.
McCoy, M., & Ladefoged, T. (2009). New Developments in the Use of Spatial
Technology in Archaeology. Journal of Archaeological Research, 17(3), pp. 263295.
National Park Service (NPS). (2002). How to Apply the National Register Criteria for
Evaluation, revised for the Internet from the 1990 edition. National Register
Bulletin, 15. U.S. Department of the Interior, National Park Service, Washington,
DC, 54 pages.
Nordstrom, N. (2014). GH00211 Extension Archaeological Site Inventory Form. On
record at the Washington Department of Archaeology and Historic Preservation,
Olympia, Washington.
Nordstrom, N. (2012). A-Line Railroad Grade/Spur Archaeological Site Inventory Form.
On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
129

Paullin, P. K. (2007). Boring to the Core: The Archaeology, History, and
Dendrochronology of a Railroad Logging Camp, Ladee Flat, Clackamas County,
Oregon. Master’s thesis, Oregon State University.
Pierre, J. H. (1979). When Timber Stood Tall. Superior Publishing Company, 192 pages.
Robbins, W. G. (1985). The Social Context of Forestry: The Pacific Northwest in the
Twentieth Century. The Western Historical Quarterly, Vol. 16, No. 4 (Oct.,
1985), pp. 413-427.
Rohe, R. (1994). Geographers and the Lake States Lumber Era: A Prospectus. The
Great Lakes Geographer, 1, pp.47-65.
Rohe, R. (1985). Settlement Patterns of Logging Camps in the Great Lakes Region.
Journal of Cultural Geography, 6(1), pp. 79-107.
Schiffer, M. B. (1986). Formation Processes of the Archaeological Record. University of
New Mexico Press, Albuquerque, NM, 448 pages.
Schiffer, M. B. (1983). Toward the Identification of Formation Processes. American
Antiquity, Vol. 48, No. 4, pp. 675-706.
Spector, R. (1990). Family Trees, Simpson’s Centennial Story. Documentary Books
Publishing Company, Washington, 236 pages.
Stilson, M. L. (2011). Mason County Logging Company North Fork Porter Creek
Railroad Logging Camp Archaeological Site Inventory Form. On record at the
Washington Department of Archaeology and Historic Preservation, Olympia,
Washington.
Stilson, M. L. (2010A) Mud Bay Camp #2, Grade, Rock Cut and Trestle Archaeological
Site Inventory Form. On record at the Washington Department of Archaeology
and Historic Preservation, Olympia, Washington.
Stilson, M. L. (2010B). McCormick Tunnel Construction Camp Archaeological Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
Stilson, M. L. (2010C) Schafer Brothers Logging Camp Archaeological Site Inventory
Form. On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Stilson, M. L. (2010D). Mason County Logging Company Camp 7 Archaeological Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
130

Stilson, M. L. (2010E). Lost Creek Valley Logging Camp Archaeological Site Inventory
Form. On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Stilson, M. L. (2009). Mason County Logging Company Camp 4 Archaeological Site
Inventory Form. On record at the Washington Department of Archaeology and
Historic Preservation, Olympia, Washington.
Stilson, M. L. (2004). Wallville [sic] Archaeological Site Inventory Form. On record at
the Washington Department of Archaeology and Historic Preservation, Olympia,
Washington.
Thompson, D. B. (1989). Logging Railroads in Skagit County: The First Comprehensive
History of the Logging Railroads in Skagit County, Washington, USA.
Northwest Short Line, 297 pages.
University of Washington Library Digital Collection (UWL) (2015). Kinsey Brothers
Photographs of the Lumber Industry and the Pacific Northwest, 1890-1945.
Obtained January 2015, from:
http://content.lib.washington.edu/clarkkinseyweb/index.html
Vaughn, J. (2013). Waddell View TBS RR Grade Archaeological Site Inventory Form.
On record at the Washington Department of Archaeology and Historic
Preservation, Olympia, Washington.
Washington State Department of Archaeology and Historic Preservation (DAHP).
(2011). Washington State Standards for Cultural Resource Reporting, Survey and
Inventory Standards, 65 pages.
DAHP. (2009). The Washington State Historic Preservation Plan 2009-2011, 56 pages.
Washington State Office of Archaeology and Historic Preservation (OAHP). (1985).
Bordeaux House Historic Property Inventory Form. On record at the Washington
Department of Archaeology and Historic Preservation, Olympia, Washington.
Westcott, K. L., & Brandon, R. J. (2000). Practical Application of GIS for
Archaeologists: A Predictive Modeling Toolkit. Taylor & Francis Group. Social
Science, 160 pages.
Woodward, G. A. (1894). Life in a Lumber Camp. Munsey’s Magazine, pp. 604-610.

131

Appendix A – Aggregate Segment Information and Findings
Segment Identification Map

132

Aggregate Segment Location Information
Distance Segment Information
Location
Aggregate
Segment
A1
A2
B1
B2
B3
C1
D1

Measured Segments

Access

Section(s)

Township(s)

Range(s)

B03, B04, C03, C04, E03, E04

3, 4
33
16,21,28

16
17
17

03W
03W
03W

B05, C05, E05

28,32,33

18

03W

B06, C06, E06

32

18

03W

B07, E07

18

17

03W

B16, C15, E11

18,19

16

03W

B15, C14, E12

12

16

04W

B13, B14, C12, C13, E14

23, 25, 26

17

04W

B01, B02, C01, C02, E01, E02

E3

B12, C11, E13

24, 25

17

04W

B08, C08, C09, E08, E09

7, 18
13
5
33

17
17
16
17

03W
04W
04W
04W

B18, C21, E18

5, 7, 8

16

04W

B11, C16, E15

15

17

04W

B10, C16, E15

15

17

04W

B09, C10, E10

12, 13

17

04W

B19, C17, E16

8, 16, 17

17

04W

B20, C18, E19

20, 29

17

04W

B21, C19, E20

19, 30

17

04W

B23, C23, E22

7, 18

17

04W

B24, C24, E25

7
12
11, 12

17
17
17

04W
05W
05W

B17, C20, E17
F1

G1
H1
H2
H3
I1
J1
J2
J3
J4

J5

J6
J7
K1
K2
K3

No
Yes
No
Yes
No
Yes
Yes

Road Name or Number

Notes

Along trail. The portion in Section 4 is
privately owned
Waddell CK/Sherman Valley Also Middle Waddell Campground
NA

NA

Along trail

C-8000
NA

Along trail

E-5000

End of E-5000 at big opening

E-7000

Road to Cedar Creek CC

NA

Monroe Creek grade

No

E1

E2

On Road?

B26, B27, B31, B32, C26,
C27, E26, E27

Yes

C-6000

Yes

C-Line/C-7000

No

NA

17, 18

17

04W

B25, E23, C42

13

17

05W

B33, B34, C28, C29, E29, E30
B28, C25, E28

6
1
7

17
17
17

04W
05W
04W

B29, B30, C30, E31

9, 10

17

04W

Partially on the D-1070

No
Yes
No
Yes
Yes
Yes
No
Yes
No
No

D-1000
NA
C-Line
C-7200
C-3000
NA

NA
NA
NA

Yes
Yes
No

Yes
No

Partially along trail

C-Line/C-2020

No

B22, C22, E21

Falls Creek Campground area

Near Porter Falls partially along trail
from Porter Creek Campground

C-3100/C-3110
B-0100
NA
B-1000
NA

Along trail also
Along trail just off B-1000

133

Aggregate Segment Location Information
Distance Segment Information
Location
Aggregate
Segment
L1
L2

Measured Segments

Access

Section(s)

Township(s)

Range(s)

B36, C33, E33

33

18

04W

C32

32

18

04W

B37, C34, E34

31

18

04W

L5
L6

B38, C35, E35

29-32

18

04W

B38, B39, C36, E37

20, 29

18

04W

B40, C37, E36

19, 30

18

04W

B43, B44, C40, C41, E40, E41

19, 20

18

04W

L7
M1
M2
N1

No
No

Road Name or Number

Notes

NA

Along Swan Creek

NA

Along trail

NA

Along trail

No

L3

L4

On Road?

Yes
No
Yes

A-4000/A-5000

Some off road

NA
A-Line/A-4000

Some off road

A-Line
Yes

B42, C38, E38

2

17

05W

B41, C39, E39

36

18

05W

B35, C31, E32

4,5
33

17
18

04W
04W

Yes
Yes
Yes

A-Line
A-2050

Some off road

B-Line/B-2030

Some off road

134

Aggregate Segment Site Probability & Characteristics
Distance Segment Information
Site Probability
Aggregate
Segment

Probability

Near
Junction

A1

High

X

X

A2

Medium

X

X

B1

Low

X

X

B2

Low

X

B3

Low

X

C1

Medium

X

D1

Medium

E1

High

X

X

High

X

X

X

X

E2
E3

Medium

Water
Near Stream
Impoundment

In or Near
Campground

Landscape
Modification

X

Near Property
Line

Close in
Geodesic
Distance

X
X
X
X
X

X
X

X

X

X
X

Medium
F1

X

G1

Low

H1

Low

H2

Low

H3

Low

I1

Medium

X

J1

Medium

X

J2

Medium

J3

Low

J4

X
X

X

X

X

X

X

Low

J5

Medium

X

X

J6

Medium

X

X

J7

Medium

X

X

K1

X

Low

K2

Low

K3

High

X

X

X

X

X
X
X

X

X
X

X

X
X

135

Aggregate Segment Site Probability & Characteristics
Distance Segment Information
Site Probability
Aggregate
Segment

Probability

L1

Medium

L2

Low

L3

Low

L4

Low

L5

Low

L6

Low

X

L7

Medium

X

M1

Medium

X

M2

Low

N1

Low

Near
Junction

Water
Near Stream
Impoundment
X

X

In or Near
Campground

Landscape
Modification

X

Near Property
Line

Close in
Geodesic
Distance

X

X

X

X

X

X

X

X

X

X
X

X

X

X

X
X

X

X

X

X

136

Aggregate Segment Field Verification Findings
Distance Segment Information
Field Verification
Aggregate
Segment
A1
A2
B1
B2
B3
C1
D1

E1

E2
E3

F1

G1
H1
H2
H3
I1
J1
J2
J3
J4

J5

J6
J7
K1
K2
K3

Date(s)
8/9/2010

Consumables Metal Pieces
X

Steam
Donkey

Water
Pipe

Structural
Remnants

Exotic
Plants

Notes
Enamelware pitcher, logging cable

X

12/9/2009
1/14/2015
12/9/2013

Nothing found - heavily disturbed
Steel pipe found (J. Vaughn) outside segments, but nothing else.

X

6/10/2010

Nothing found

1/13/2015

Nothing found

9/30/2010
1/13/2015
9/18/2014

Nothing found

4/29/2009
4/25/2011
1/22/2015

10/20/2010

Trash pile found north of this segment by Mo

X

X

X

X

X

Original survey of grade with Rolin Christopherson found a
whiskey bottle from 1910-1920 near stream crossing structure and
donkey outside segment; grade was recorded. Second survey
found small china fragment. Third survey found nothing new.
X

1/13/2015
6/9/2010
9/18/2014

X

Insulator, china fragment, rail piece, and fruit tree found
Found trestle remnants and suspicious ivy spot, but no artifacts
Found glass outside segment that can date from 1870s to 1938;
second search revealed nothing new. Camp likely, but unable to
determine precise location.

X

9/18/2014

Nothing found

1/13/2015

Nothing found

1/13/2015

Nothing found

6/23/2010

Nothing found

1/14/2015

Nothing found

1/31/2014

Nothing found (N. Nordstrom)

10/20/2010
1/31/2014
1/14/2015

Pipe in ground just east of saddle, but outside range segment.
Nothing found in segment survey (N. Nordstrom).
Nothing found

X

5/18/2009

Surveyed by Rolin Christopherson. No artifacts found.

6/2/2010
1/16/2015

Surveyed and recorded railroad grade West of Porter Falls - only
trestle remnants and one brick found. Found small piece of china
(undetermined age) and a large spot of English ivy in Porter
Creek Campground
Located 2 sets of steam donkey skids 20 feet apart along with an
abundance of logging cable; no glass or porcelain artifacts
Nothing found

1/14/2015
1/14/2015
5/10/2010

X

X

X

X

X

X

9/18/2014

No artifacts - recorded railroad grades (6/22/10), major stream
crossing
Nothing found

9/18/2014

Nothing found

Camp
Present?
Possible
No
No
No
No
No
Possible

Possible

Possible
Possible

Possible

No
No
No
No
No
No
Possible
No
No

Possible

Possible
No
No
No
No

137

Aggregate Segment Field Verification Findings
Distance Segment Information
Field Verification
Aggregate
Segment
L1
L2

Date

L5
L6

M2
N1

Structural
Remnants

Exotic
Plants

Notes
Nothing found other than stream crossing pylons

1/16/2015

Nothing found

7/12/2012

1/22/2015

Artifacts found near the north end of segment. Artifacts date to
later than logging operations would have been in the forest (post1960); most likely a hunting cabin. Grade was previously
recorded.
Nothing found other than stream crossing pylons

1/22/2015

Nothing found

1/22/2015

Nothing found

1/22/2015

1/14/2015

Suspicious stove pipe found partially buried in dug out area
along grade; abundance of holly, also large flat directly east of
flat
Nothing found (N. Nordstrom). However, isolate coffee cup
found 1550 feet north of segment
Nothing found

1/16/2015

Nothing found

X

L7
M1

Water
Pipe

1/16/2015

L3

L4

Consumables Metal Pieces

Steam
Donkey

X
3/22/2012

X

X

Camp
Present?
No
No

No

No
No
No

Possible
Possible
No
No

138