THE WISCONSIN ARCHEOLOGIST Volume 96, Number 2 July–December 2015 Published biannually by the Wisconsin Archeological Society Contents Editors’ Corner Constance M. Arzigian, Katherine P. Stevenson, and Vicki L. Twinde-Javner ....................................................2 Petrographic Analysis of Pottery Thin Sections from Three Late Woodland Sites in South-Central Wisconsin James B. Stoltman, Marlin F. Hawley, and Norman M. Meinholz .......................................................................3 The Rodney Clark Site (47MR146): A Quartz Biface Workshop in the Wausau Locality Robert F. Boszhardt ..............................................................................................................................................25 Aztalan’s Northeast Mound: The Construction of Community Thomas J. Zych .....................................................................................................................................................53 Fisher Mounds Site Complex (47VE825): 2001–2006 Investigations in Stoddard, Wisconsin Constance M. Arzigian ......................................................................................................................................119 Ground-Penetrating Radar Survey of Selected Portions of First Fort Crawford (47CR249), an Early Nineteenth Century American Military Post, Prairie du Chien, Wisconsin Daniel J. Joyce and Jonathan T. Van Beckum ....................................................................................................214 Ground-Penetrating Radar Survey of Selected Portions of First Fort Crawford (47CR249), an Early Nineteenth Century American Military Post, Prairie du Chien, Wisconsin Daniel J. Joyce and Jonathan T. Van Beckum Abstract In the fall of 2009, the authors conducted a limited ground-penetrating radar survey in Prairie du Chien, Wisconsin on the remains of the First Fort Crawford (47CR249). First Fort Crawford was built in 1816 and abandoned in 1832. In the 1930s, the fort was accidentally rediscovered. At that time, Leland R. Cooper of Hamline University excavated a majority of the fort’s rooms as a Works Progress Administration project. Largely undocumented, these excavations have left us without an accurate idea of the extent of the excavations and, more to the point, what was not excavated. Sketchy surviving records suggest that the powder magazine and latrines were either not located or not excavated at the time. Comparing forts with similar architecture offered clues as to where to look for those features. Cooper’s accession books and photographs allowed us to reconstruct much of the excavation and its features. Geophysical Survey Systems SIR-3000 ground-penetrating radar was used in an attempt to relocate these features for future excavation and to determine the extent of previous excavations. The results of these investigations indicate that there are intact rooms and features that were not previously excavated. Introduction Located in southwestern Wisconsin near the confluence of the Wisconsin and Mississippi Rivers, the First Fort Crawford (47CR249) was built upon the remains of Fort Shelby, a War of 1812 fort that was captured by the British in 1814 and renamed Fort McKay (Scanlan 1998; Twinde-Javner 2009a). Fort Crawford was built on the floodplain of the Mississippi River. It was frequently flooded during its occupation and was eventually abandoned in 1834. During this same time, the second Fort Crawford, a stone fort on higher ground, was being built and was completed in 1834 (Mahan 2000:137; Twinde-Javner 2005:61, 2009b:124). For the purpose of this report, the term “Fort Crawford” refers to the First Fort Crawford (Figure 1). In the late 1930s the fort was accidentally rediscovered during a ditching operation for placement of a water main that revealed a stone-lined cellar room. At that time, Leland R. Cooper of Hamline University was called in to assess the find. Securing support, Cooper excavated the majority of the fort as a Works Progress Administration (WPA) project starting in 1937. Limited surviving records indicate that the powder magazine and latrines were not located or excavated then. The extent of Cooper’s investigations is suspected but not verified, and it appears that areas within the parade ground and outside of the limits of the rooms were not excavated. Thus, this project assessed the potential for locating unexcavated outbuildings, latrines, and other features of the fort. To the best of the authors’ knowledge, nothing has been published on Cooper’s excavation. Rumors persist that there may have been a written report, but nothing other than a table of contents has been located, despite the efforts of Joyce, Van Beckum, and Twinde-Javner in archival research at the Wisconsin Historical Society (WHS); the Minnesota Historical Society; Fort Crawford Museum in Prairie du Chien, Wisconsin; Stonefield Village in Cassville, Wisconsin; or Hamline University, where Cooper taught in the 1950s (Twinde-Javner 2005, 2009a; Van Beckum 2006). Both the junior and senior authors cataloged Cooper’s collection, which had been assigned over 14,000 catalog numbers (Van Beckum 2006). The total number of artifacts actually far exceeds this number due to the bulk bagging of like items and assigning a single catalog number to the entire bag. Fortunately, the collection includes a database listing of the original object Daniel J. Joyce, Kenosha Museum System, Kenosha, Wisconsin; Jonathan T. Van Beckum, Brown Deer, Wisconsin. The Wisconsin Archeologist, 2015, 96(2):214–230 Vol. 96, No. 2 The Wisconsin Archeologist 215 Figure 1. Nineteenth-century American fortifications in Prairie du Chien, Wisconsin (Smith 1973:99). Wisconsin map 1:100,000 scale (Wisconsin State Cartographers Office 2004). numbers and the room from which they were excavated, along with black-and-white photographs of the excavations. Unfortunately, however, no overall site map exists that indicates exactly what was excavated or from where. In addition, no officially drawn plan of the post has been found, such as was commonly made during fort inspections, showing the exact layout of the fort at the time of construction and occupation. There are some contemporary verbal accounts from military 216 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin inspection tours, as well as descriptions of the fort and artwork from the period, but the authors and TwindeJavner (2005, 2009a) have not located any plan of the fort in archival research at the National Archives in Washington D.C. The fort’s foundations can be ascertained using the reconstructed bastion in the northwest corner, the limestone markers outlining the foundations to the south and east of the bastion, a lithograph by Henry Lewis (Figure 2), and the measurements of the fort provided by Major Long when he visited Fort Crawford in 1817. Long described the fort as square, with walls 340 feet on each side (Scanlan 1998). Figure 3 is a topographical view of Fort Crawford and the Borough of Prairie du Chen from 1821. Figure 4 is a sketch of Fort Crawford made from Long’s description of the fort. All of these sources were used to reconstruct what Fort Crawford looked like when it was an important military fortification in the early nineteenth century. The authors used ground-penetrating radar (GPR) to secure additional information about the physical structure of the fort and to test the accuracy of historic artwork and descriptions. Due to the lack of a site Joyce and Van Beckum report by Cooper, the authors wanted to find out what areas of the fort were previously excavated by the WPA and also hoped to locate previously unknown features such as the powder magazine, the latrines, and perhaps other detached buildings. Geophysical survey techniques are a cost- and time-effective method to gain information about Fort Crawford and can serve as a base for future work at the fort. GPR, electrical resistivity, gradiometry, and other remote sensing techniques have been shown to be effective noninvasive survey techniques with the potential to locate subsurface archaeological features that would otherwise require excavation (Hargrave et al. 2002; Kaufmann and Kean 2002; Parrington 1979). In addition, numerous GPR studies of forts have been made in Iowa, Illinois, and Wisconsin with some success (Whittaker 2005, 2006a, 2006b, 2007, 2009a, 2009b). These examples demonstrate that geophysical survey techniques can be used to locate building foundations in order to map out a site plan. In the case of Fort Crawford, the rough dimensions of the fortification were already known and are partially exposed in one location. The fine-grained, sandy soils of this Figure 2. “Old Fort Crawford” from a color lithograph published in Das Illustrirte Mississippithal in 1857 by Henry Lewis (Wisconsin Historical Society, WHS-42291). Vol. 96, No. 2 The Wisconsin Archeologist 217 Figure 3. Map dated 1821 showing the location of First Fort Crawford on St. Feriole Island, Prairie du Chien, Wisconsin. Map is oriented with north to the left; the left end of the staff that bisects the fort points north (Wisconsin Historical Society, WHS-42267). bottomland area site are well drained and have low clay content (Slota and Garvey 1961). This type of soil works well for GPR and should reflect the buried foundations of the fort (Conyers 2004; Conyers and Goodman 1997; Hargrave et al. 2002). A geophysical survey of the Fort Crawford site would go a long way toward accurately describing this historic site and would add a valuable component to the history and archaeology of Wisconsin. Site Description Fort Crawford (47CR249) is located in Crawford County, southwestern Wisconsin, north of the confluence of the Wisconsin and the Mississippi Rivers (Figure 1). Situated on St. Feriole Island, it is part of the present-day city of Prairie du Chien (Figures 1 and 3). Crawford County is in an area of the state that was not glaciated during the most recent Wisconsinan glaciation (Martin 1965:83), and the area is characterized by higher tracts of land dissected by deep gorges (Finley 1970:7; Martin 1965:33). Crawford County is bordered on the western edge by the south-flowing Mississippi River and on the southern edge by the west-flowing Wisconsin River. The Mississippi River at Prairie du Chien is about 1,400 feet wide and is bracketed by terraces rising 500 feet above the floodplain (Martin 1965:145, 154). Site History Prairie du Chien was situated at one of the ideal geographic locations in the Upper Midwest for the fur trade. The Mississippi River allowed trade and commerce to flow from northern Minnesota all the way to New Orleans. The Wisconsin River, emptying into the Mississippi just south of Prairie du Chien, allowed trade to flow from Quebec through the western Great Lakes to the Gulf of Mexico. The city was also just north of the important lead mining resources of southwestern Wisconsin and northwestern Illinois. Thus, the area around Prairie du Chien had long been of economic and strategic importance to the fur trade (Mahan 2000; Smith 1973). Native Americans 218 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin Joyce and Van Beckum Figure 4. Undated sketch made from Major Long’s description of the fort and archaeological excavations, probably ca. the 1940s. The northwest bastion is in the foreground. The northeast corner, which is GPR grid one, is on the left of the drawing. Villa Louis now sits where the southeast bastion, in the rear of the sketch, is located. The rifle points north (Wisconsin Historical Society, WHS-42236). saw the area as a neutral trading spot where they could trade with other Native peoples and Europeans. European and American forces, in order to gain control of this area and influence fur trade activity, built fortifications and fur trade posts throughout the region. Prairie du Chien alone has been the site of a French fort, a British fort, and three separate American forts (Twinde-Javner 2009a). The American government eventually established fortifications from Lake Michigan to the Mississippi River in the early 1800s (Prucha 1953; Wesley 1935). The first American fort to be constructed at Prairie du Chien was Fort Shelby, completed in June of 1814 (Kellogg 1935:315). The American presence at Fort Shelby was reported to the British by their Indian allies, and British troops were sent from forts in the area. Fort Shelby was then surrounded and attacked, and finally surrendered to the British on July 20, 1814 (Geister and Shaw 1998; Kellogg 1935:319). Britain renamed the captured fortification Fort McKay. At the end of the War of 1812, after a nine-month occupation by the British, the fort was abandoned and burned (Kellogg 1935:325). While the Treaty of Ghent officially removed British influence from the upper Great Lakes, British-allied trappers and Indian agents still remained a strong presence in the fur trade of the region. To counteract this foreign influence, the United States government built Fort Howard at the mouth of the Fox River at present-day Green Bay, Wisconsin, in July 1815, and a larger military post, Fort Crawford, at Prairie du Chien in June 1816 (Smith 1973:97–98). The establishment of Forts Howard and Crawford, along with Fort Edwards, Armstrong, and Snelling, allowed the United States to control access from the Canadian fur trade centers to the Mississippi River, and to protect the civilian settlements and businesses at these locales (Smith 1973:101–102). Determining what Fort Crawford looked like during its occupation was somewhat daunting. Although sketches, paintings, and descriptions exist, their accuracy was uncertain. What was needed to confirm these sources was a good period map. After an exhaustive search, one map of the fort was found, dated to 1821 (Figure 3). Although military inspections of the fort were made, a search of the National Archives did not reveal any associated maps, despite maps being a common byproduct of such inspections. However, Major Stephen H. Long described the fort in detail in 1817: The work is a square of 340 feet upon each side and is constructed entirely of wood…except the magazine which is of stone. It will accommodate five companies of soldiers....The faces of the Vol. 96, No. 2 The Wisconsin Archeologist work are flanked by two Block Houses, one of which is situated in the SE and the other in the NW Corner of the fort… The Block Houses are two stories high with Cupolas, or turrets, upon their tops…The first stores (sic)…(and)…the second…being placed diagonally upon the first…. The Quarters, Store Houses &c. (sic) are ranged along the sides of the Garrison, their rear walls constituting the faces of the work….The buildings are constructed with shed roofs, sloping inward, so that their outward walls are raised 20 feet from the ground….The rooms are generally about 19 feet square….The magazine is 24 by 12 feet in the clear with walls 4 feet thick….The works are, fore (sic) the most part, constructed of square timber, and the crevices in the walls of the building plaistered (sic) with lime mortar [Kane et al. 1978:88]. Long’s description and the color lithograph by Henry Lewis (Figure 2) allowed an unknown artist (possibly Cal Peters) to sketch a probable likeness of Fort Crawford based upon their descriptions and Cooper’s WPA archaeological excavations (Figure 4). As the territory that eventually became Minnesota, Iowa, and Wisconsin grew in population, the value of Fort Crawford began to be questioned. Fort Snelling in the Minnesota territory was better situated to control the fur trade. Also, the land where Fort Crawford was located was unhealthy, low-lying ground prone to flooding. The floods of 1822 and 1826 forced the garrison to evacuate the fort and make costly and time-consuming repairs upon their return (Mahan 2000). An 1826 inspection report showed that the fort was in very poor condition. These problems prompted the military to abandon the fort in October 1826 (Mahan 2000:103; Scanlan 1998:129). A Ho-Chunk Indian uprising in the summer of 1827 prompted the military to rethink their abandonment decision, and Fort Crawford was reoccupied in the summer of 1827 (Mahan 2000:111; Scanlan 1998:133). In April of 1829, official orders were received to find a new location for the fort (Mahan 2000:124). In May, construction of a second Fort Crawford, built of stone, began on higher ground overlooking the city of Prairie du Chien and the Mississippi River (Mahan 2000:126). In August of 1834 the second fort was completed, and the 16-year history of the First Fort Crawford came to an end (Mahan 2000:136–137). The property passed from government to private hands in the nineteenth century, and back into government hands in the twentieth century. Currently the Villa Louis house, an historic site owned by WHS, resides on part of the old fort’s foundations. 219 Excavation History The rediscovery of Fort Crawford was made in 1936, when the city of Prairie du Chien dug a new swimming pool where the fort once stood. At that time, the city of Prairie du Chien owned the property where the fort once resided (Wisconsin Historical Society 2015). During digging of a trench for a water main, a stone-lined cellar room of the fort was found. By this time, the pool was largely completed, and the construction crew had unknowingly destroyed more than half of the west row of rooms. Between 1937 and 1940, Leland R. Cooper of Hamline University excavated the fort as a WPA project. These excavations uncovered the supposed west bastion of Fort Shelby/ McKay, the First Fort Crawford, and the nearby Boilvin or Indian Agent House. The limited archival documentation associated with Cooper’s excavation is currently stored at WHS and at Villa Louis, and Villa Louis is responsible for the care of the fort and its associated surroundings. The existing documentation includes the original accession books, black-and-white photographs, four plan-view feature drawings, one elevation drawing of units excavated at Fort Crawford, a table of contents from an archaeological report, correspondence between Cooper and other institutions, plus miscellaneous documents, maps, and newspaper clippings pertaining to the excavation. Photographic evidence indicates that the perimeters of the rooms were excavated first, followed by the interiors. The units excavated by Cooper corresponded with room designations and were labeled as such in his accession logs. It appears that Cooper excavated the northwest bastion, the north row of buildings, the northern and southernmost parts of the west row, the northern two-thirds of the east row, and much of the south row of buildings (Figure 5). Construction of the original Dousman House in 1843 and the later 1871 Villa Louis destroyed the southeast bastion of Fort Crawford and much, if not all, of Fort Shelby/McKay. By the summer of 1938, Cooper wrote a letter to Thomas M. Pitkin and reported that two-thirds of the fort had been excavated (Cooper 1938). Units were dug haphazardly and screens were used for recovery, but there are no known notes from Cooper’s excavation that indicate the sizes and location of the units or the screen size. No levels were recorded, and no artifacts were piece-plotted. Major features were mapped and numbered. Based on photographs, only the northwest bastion had any substantial stone foundations. What was left of the north, south, east, and probably west rows had little in the way of foundation stones, and what few stones still existed were 220 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin removed during Cooper’s excavation, although the cellar rooms did have well-made stone walls. This would explain why much of the west row was destroyed by construction and the fort walls went undetected until a Joyce and Van Beckum stone-walled cellar room in the east row was encountered. As a part of the WPA project, the outline of many of the fort rooms and fireplaces were rebuilt for public interpretation using new stone (Figure 6) (Cooper Figure 5. The shaded areas are known to have been excavated. All of the shaded room blocks were excavated no more than a meter outside of the walls, into the parade ground or to the exterior of the fort (Joyce 1997). Figure 6. North row and northwest bastion (looking west). Reconstruction of original foundations with new stones can be seen (Wisconsin Historical Society, Museum Accession File 1997.62, Cooper Correspondence Photo P-63). Vol. 96, No. 2 The Wisconsin Archeologist 1938). Where Cooper got these stones to construct these outlines remains a mystery. The excavation documentation found in the Curator of Anthropology’s office at WHS is extremely limited. The table of contents, though clearly related to Fort Crawford, has specific page numbers but no title. Letters from Cooper indicated that as late as November 1955, the report was still being planned (Cooper 1954, 1955). After extensive searching, it appears that Cooper may never have finished the report, or it may have been lost in the 1969 flood that damaged Villa Louis. Furthermore, in 1993, James Stoltman of the University of Wisconsin–Madison excavated at the adjacent Villa Louis site (47CR7) in order to better understand the relationship between the mound on which Villa Louis sat, the house itself, and the First Fort Crawford. Due to flooding at that time, the excavations were very brief, and no evidence of the fort was found (Whittaker 2012). Ground-Penetrating Radar Surveys and Methods On October 3 and 4, 2009, the authors conducted limited GPR investigations at the First Fort Crawford. All of the data were collected using a Geophysical Survey Systems SIR-3000 with a mobile three-wheeled cart and a 400-MHz antenna. In all seven grids, data was collected at .5-m transect intervals. Transects were made in either a back-and-forth (Grids 1, 3, 4, 5, 6) or a unidirectional fashion (Grids 2 and 7). Data were post-processed using RADAN 6.0 software. This last step is necessary for three-dimensional processing. Higher gain was applied for maximum anomaly visibility to improve visual contrast. These processing steps improve the resulting accuracy in determining size, shape, and depth of anomalies and subsurface features. Grids were laid to a minimum size, parallel to the building layout due to time constraints. They were oriented to magnetic north, the same as the fort when it was constructed. The grid number, size, transect interval, and direction are detailed in Table 1. In a GPR survey, an electromagnetic pulse is generated from the radar and transmitted into the ground. The pulse then travels until it comes in contact with a subsurface material that interferes with the signal. At that point, the pulse is then reflected back to the surface and is recovered by a receiver and an antenna that are attached to the device carrying the GPR. The time it takes for this round trip is measured in nanoseconds (ns). The smaller the ns number, the closer to the surface the disturbance should be (Conyers and 221 TABLE 1. GRID NUMBER, SIZE, TRANSECT INTERVAL AND DIRECTION. Grid # N–S Axis (m) E–W Transect Axis (m) Interval Transect Direction 1 15 15 .5 bidirectional 2 22 10 .5 unidirectional 3 26 10 .5 bidirectional 4 52.3 10 .5 bidirectional 5 10 83.5 .5 bidirectional 6 10 27.2 .5 bidirectional 7 25 13 .5 unidirectional 4 52.3 10 .5 bidirectional 5 10 83.5 .5 bidirectional 6 10 27.2 .5 bidirectional 7 25 13 .5 unidirectional Goodman 1999; Davenport 2001; Whittaker 2009b). The authors’ original plan was to conduct two small test excavations to ground-truth the data and be able to convert the ns scale to actual depths on site. Due to time constraints and cold weather, this invasive method was abandoned. The authors hope to return at a future date and conduct these test excavations. During the survey, conditions were less than ideal for GPR work. The week before, it had rained for several days, and during the first workday, it rained lightly on and off. Both days were unseasonably cold and windy. It was thought that since the cultural deposits were shallow, they would be only minimally affected by the moisture, which attenuates or diminishes the radar signal at depth (Conyers 2004). Figure 7 shows the location of the seven GPR surveys conducted at Fort Crawford. Grid 1. Grid 1 was located at the northeast corner of the fort and targeted a section that the authors thought was not excavated by Cooper. It was considered one of the more likely locations for the latrines, possibly because the soldiers did not want to leave the fort to relieve themselves, as the fort was located in what was perceived to be hostile territory at that time. The southwest and northeast corners of the fort did not have bastions. The southwest corner was eliminated because of a port gate (Figure 5). In the northeast corner, the east and north rows of barracks did not physically connect, and the corner gap of the fort was enclosed by a stockade set at a 90-degree angle (Figure 5). This created an isolated area of the fort that was ideal for a large latrine or series of latrines. This area was approximately 50 cm higher than the present-day level of the excavated rooms, 222 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin Joyce and Van Beckum Figure 7. Ground-penetrating radar grid surveys of Fort Crawford. The black arrow on the lower right points to the powder magazine, which the authors hoped to locate with Survey 7. indicating that Cooper may not have excavated it. There is no record of excavations here in Cooper’s accession books. Today, the area is park-like and mostly level except for the excavated areas. Plan views of the grid data are shown in Figure 8. In the upper left corners of Figure 8, the 15 x 15 m grid is superimposed in red over a portion of the sketch map, which can be seen in its entirety in Figure 5. The excavated rooms in the north and east rows were overlapped slightly by the grid, as were the stockade line and cement corner pad in the upper left. The upper center image shows the features seen near the surface and on the map to the left. These features are highlighted in black (Figure 8). In the lower left image, two possible latrines are indicated by black arrows (Figure 8). The red arrow through each image (Figure 8), from left to right, transects the suspected latrines and indicates the location of the reflection profile slice in Figure 9. The latrines or sinks can be seen in the center of the profile at 6–7.3 and 8.5–9.5 m. A modern disturbance and foundation wall can be seen to the right of the latrines. The object on the left was highly reflective (probably a large piece of metal), which created multiple reflections. The question mark to the left may be a much larger pit feature. This grid indicates intact, unexcavated features in this area and bodes well for future work at the fort. Grid 2. This survey took place immediately south of the northwest bastion and covered the northernmost rooms of the west barracks, as well as the stockade fences connecting the north and west barracks to the bastion and to each other (Figure 5 and 7). The grid was bounded at the south by a small rise that was created when the pool was built, and may indicate the extent of Cooper’s excavations because the city pool was constructed over parts of the west row of the fort. The grid area was 22 x 10 m. Figure 10 is a horizontal slice map of the survey area. The southeast corner of the northwest bastion is the upper right black rectangle. The northernmost room of the west barracks can be seen on the left of the figure. The olive green arrows indicate the corners of the northernmost room of the west barracks. A large east-west (top to bottom in the image) disturbance can be seen going through the middle of the room (yellow arrows). All of the cultural deposits were very shallow in this area. The rooms of the west barracks probably extend under the topographic rise to the south or left. This is borne out in Grid 3. The northwest bastion was connected to the west row of rooms by a stockade that appears as a very faint line at the upper green arrow. This stockade line is faint because it paralleled a transect line. It would have been more visible if the grid were laid out at an angle to the building lines. Vol. 96, No. 2 The Wisconsin Archeologist 223 Figure 8. Horizontal slice map of Grid 1. Building foundation outlines are superimposed in black. Figure 9. Grid 1. Reflection profile slice bisecting foundations, disturbance and possible latrines. The lower green line indicates a stockade line from the inner corner of the west row of barracks to the inner corner of the north row of barracks. This survey did not indicate any depth to features in either the rooms or the stockade lines. The features were all close to the surface. Cooper excavated this room and outlined the foundations. It appears from this survey that any subsurface features were removed during excavation. The area between the stockades and the northwest bastion is higher in elevation than the adjacent north and west row of rooms (Figure 11). This suggests that this area was not excavated and needs to be carefully surveyed to determine whether there are any intact features. Even though it appears that Cooper excavated a small portion of the west barracks, there were no accession numbers in his ledger or artifacts from the west row. Figure 10: Horizontal slice map of grid 2 at 6 nanosecond depth. The corner of the northwest bastion is in the upper right. The northernmost room of the west row is at left. Olive green arrows indicate the corners of the room. An H-shaped hearth can be seen at the red arrow. A major disturbance running east–west (up and down) is marked by yellow arrows. The green arrows indicate stockade lines connecting the bastion to the west row. The top (west) stockade line is very faint because it parallels a transect line. Grid 3. Grid 2 ended at the bottom of the rise, and Grid 3 began at the top of the same rise (Figure 7). The west sally port was located in the center of the west row and marked the southern end of the grid. The grid area was 26 x 10 m. In Figure 12, the area to the right of the red line indicates disturbed soils. The red arrow indicates a subsurface wall remnant on the inner or parade ground 224 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin Joyce and Van Beckum Figure 11: The extent of Cooper’s excavations in the northwest corner of the fort. The area between the stockade fences and the northwest bastion/west row/north row is higher in elevation, as it is today, showing that it has not been excavated. The west row excavations consist of just over one room. The extent of the pool’s intrusion into the west row of barracks can be seen here (Wisconsin Historical Society, Museum Accession File 1997.62, Cooper Correspondence Photo P-98A). Figure 12. Horizontal slice map at 4.5 nanoseconds shows an extant inner wall of the west barracks (red arrow) north–south. The area to the right of the red line is the extent of disturbance by construction and eventual demolition of the swimming pool. It appears as if the area to the left of the line and bounded on the east by the existing wall (red arrow) is intact room deposits (yellow arrow). No features can be seen on the surface other than the modern sidewalk. side of the west row of barracks. The area to the left of the “disturbance” red line, and bounded on the east by the existing wall (red arrow), appeared to be made up of intact room deposits indicated by the yellow arrow. This appeared to be as much as two partial rooms and one full room that have not been excavated and forms one of the few areas of the fort known to be unexcavated. No features could be seen on the surface other than the modern sidewalk. The potential for future data recovery in this area of the fort is great. Vol. 96, No. 2 The Wisconsin Archeologist Grid 4. Grid 4 started at the west barracks sally port and proceeded south beyond the southwest corner of the fort (Figures 5 and 7). The grid was 52.5 x 10 m. Figure 13 shows an extant buried wall. According to the 1930s map of the fort, the west barracks ended north of here (approximately 12 m) and the south row of barracks extended all the way to the west wall. Another gate, or stockade separated the west row and barracks from the south row and barracks. As seen in Figure 5, the “port gate” marked the western side of the last room in the south row of barracks. Though this room was excavated by Cooper in 1938 (Figure 5), a buried portion of this wall remains. The green arrow (Figure 13) indicates the location of the cement obelisk that marked the southwest corner of the fort. 225 Two swimming pools were known to have been dug in the area of the fort: one pre-1936 and one post1936. The post-1936 pool is referred to as the “New City Swimming Pool.” In 1936, a trench for a water main was cut through the west barracks of the fort for the pool, as indicated by the two red arrows in Figure 14 and seen in Figure 5. To the left (west) of the red line were highly disturbed sediments from construction and subsequent removal of the pre-1936 swimming pool. This line was only an approximation, and there may be some in situ deposits to the left (west). An intact partial room may be located to the north of the pool trench and a full room may be intact to the south of the trench. Figure 15 is a profile slice at the 2-meter line going south. It and all of the other reflection profile slices in Figure 13. Horizontal slice map of the southern half of the west barracks at 6.5 nanoseconds. The red arrow points to an extant buried wall. The green arrow points to the southwest corner of the fort where there is a small cement obelisk at the surface. Figure 14. Horizontal slice map of the southern half of the west barracks at 18 nanoseconds. The red arrow indicates a trench that was created for the pre-1936 swimming pool. To the left of the red line are massively disturbed sediments from the construction and subsequent destruction of the 1936 swimming pool. There may be some intact areas to the north, or left, of this line. Figure 15. Profile slice at the two-meter line. The trench is indicated by the red arrows, and the pipe is indicated by the yellow arrow. The pipe lay 18 to 18.5 meters south of the cement obelisk at the southwest corner of the fort. 226 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin this grid clearly show a trench for the pre-1936 swimming pool. Two red arrows indicate the trench, and a yellow arrow indicates a pipe. The pipe for the pool lay 18.5 m north of the cement obelisk at the southwest corner of the fort. These two pools, and their impact on the remains of Fort Crawford, can be seen in Figure 5. Grid 5. Grid 5 started at the west end of the south barracks and overlapped the south end of Grid 4. Grid 5 started at the southwest corner of the fort at the obelisk and proceeded east 83.5 m, partially up the mound at Villa Louis (Figure 7). The grid area was 83.5 x 10 m. The GPR showed increasing disturbance approaching the mound at Villa Louis. This area of disturbance overlapped Grid 7 and confirms data presented there. These disturbances may relate to trenching and testing by Cooper in his efforts to find Fort Shelby. There are hints of possible walls and features in this profile, but none should be considered definitive until a better survey is completed. This was the last survey completed during fieldwork and in hindsight should have been planned and collected more carefully to find the remains of Fort Shelby. Grid 6. Grid 6 was located in the west part of the north row. It is 10 x 50 m, as shown in Figures 5, 7, and 16 and covered four rooms and a sally port. This survey was undertaken to see whether any of the features outlined above ground still existed beneath the surface. Given Cooper’s total removal of foundation wall stone, it was thought that he might have removed stones associated with smaller features, such as hearths, as well. This area of the north row has the walls and several features outlined by stone on the ground surface today (Figure 16). This is the only area where the authors have several of Cooper’s feature maps to compare with the GPR results. The excavated area, like all of the excavated and stone-outlined areas, was well below the adjoining surface. Features in Cooper’s maps can be seen at the top of Figure 17. They include hearths, a forge area, a sally port, and several unspecified features. The green arrows link like features between the GPR plan views and Cooper’s feature maps. They clearly indicate that there are subsurface materials within the features, which are only outlined in stone on the surface. This was especially true in the third room from the left. A large, roughly rectangular feature with an opening at the bottom can be clearly seen in the 2 ns slice and on Cooper’s feature map (red arrow). Grid 7. Grid 7 was located on the west side of Villa Louis approximately where the powder magazine was located (Figures 5 and 7) and was 25 x 13 m. It Joyce and Van Beckum included the cobblestone walkway on the west side of Villa Louis and was designed to see whether foundations for the powder magazine were still intact. The magazine is indicated by a black arrow on the 1821 map on the right in Figure 7. The east end of Grid 5 overlapped the southwest end of Grid 7. Results indicated that this area was highly disturbed, not surprising considering its proximity to the house. Past excavations adjacent to the west side of the house for stabilization purposes showed that this overlapping area was “largely made up of debris from the construction and removal of the first house as well as debris from the construction of the second house” (Broihahn 2009). Depending on the depth of this disturbed ground, future non-invasive and/ or invasive surveys might uncover the remains of the fort’s powder magazine. Figure 18 shows a paved walkway that surrounds Villa Louis. This walkway was made of very thick, cut stone that shows some signs that its present use was secondary in nature (Michael Douglass, Director of Villa Louis, personal communication 2009). The stone pathway runs adjacent to the west side of the house and was surveyed in Grids 5 and 7. These stones are cut and have a flat, prepared surface. Due to their flat surface, coupling was maintained during the GPR surveys. Perhaps when Hercules Dousman dismantled the fort in 1843 and built his house on the mound where the southeast bastion was, he also dismantled the adjacent 32 x 20 foot powder magazine with its four-foot thick walls, and used the material as pavers around the new house. Conclusions These surveys have yielded clues and some clarification as to the history of the fort and the WPA excavations. Those excavations, lacking field notes and a final report, have left us with more questions than answers. The unfortunate result is that the state has been denied a part of its history. Fortunately, WPA photo documentation has allowed us to, in a limited manner, reconstruct Cooper’s excavations. Locating unexcavated areas of the fort was unexpected. The northeast corner of the fort is the most promising for future excavation. The area needs to be carefully surveyed again to better understand numerous anomalies. Reviewing survey results and the architecture of other forts leads us to believe that this is the only possible area for company latrines. Intact rooms were located in the north end of the west barracks and potentially at the southernmost end of the west barracks. Immediately southeast of the Vol. 96, No. 2 The Wisconsin Archeologist 227 Figure 16. Grid 6 outline can be seen via the white measuring tape lying on the ground. The old stable, once converted into a museum (red building), can be seen in the background. Figure 17. Comparison of Grid 6 horizontal slice maps at 2 nanosecond depths and Cooper’s feature maps. 228 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin Joyce and Van Beckum Figure 18. Grid 7, west side of Villa Louis. reconstructed bastion was another pentagonal unexcavated area between the stockades, the northwest bastion, and the north and west rows of buildings. In the north barracks, the survey indicated that features internal to the rooms were still extant. A shortcoming of these surveys was that the grids were laid out in the same cardinal directions as the fort buildings. When this is done, walls that are in the direction of the survey collection lines are less visible than walls crossed at a 90-degree angle. The unexcavated areas need to be carefully surveyed again at a 45-degree angle to the walls. Two surveys of each area with survey lines at 90 degrees to each other can then be merged for maximum clarity. This work indicates that the potential for further surveys and excavations is great. The unexcavated areas have potential for yielding insights into the lives of soldiers on the frontier, with other areas potentially yielding information on the westernmost battle of the War of 1812. Fortunately, areas inside and outside of the fort rooms were not excavated by Cooper. Many questions remain. Where was the fort cemetery? Were there outbuildings associated with the fort? Did Cooper find Fort Shelby/McKay? The artifacts that Cooper associated with 1814 Fort Shelby must be examined further to determine whether he was correct. They may yield information on potential in situ deposits relating to that war and the small battle that occurred there. There is still unexpected potential for deriving more information on frontier military life both during and after the War of 1812 at a fort thought to have been excavated in its entirety in the 1930s. Acknowledgments The authors would like to thank the Wisconsin Archeological Society for its financial support of this project via the Wisconsin Archeological Society Research Award. We would also like to thank the Wisconsin Historical Society for allowing us to conduct our research at the site. Our thanks also go out to our families for allowing us to spend a long, and wet, weekend in Prairie du Chien. The authors would Vol. 96, No. 2 The Wisconsin Archeologist also like to thank Michael Douglass, former Director of Villa Louis, for his hospitality, patience, and assistance while we were conducting the GPR surveys on the grounds of Villa Louis. We look forward to returning to Villa Louis and continuing our research on Fort Crawford. Finally, we would like to thank those who read, reviewed, and offered suggestions for improvement of this article. Their assistance was much appreciated. As always, any remaining errors are the fault of the authors. References Cited Broihahn, John 2009 E-mail correspondence October 7, 2009 quoting 1998 site file update. Office of the State Archaeologist, Department of Historic Preservation, Wisconsin Historical Society. Conyers, Lawrence B. 2004 Ground-Penetrating Radar for Archaeology. AtlaMira Press, Walnut Creek, California. Conyers Lawrence B., and Dean Goodman 1997 Ground-Penetrating Radar: An Introduction for Archaeologists. AtlaMira Press, Walnut Creek, California. 1999 Archaeology Looks to New Depths. Discovering Archaeology 1(1):70–77. Cooper, Leland 1938 Letter to Mr. Thomas M. Pitkin, July 4, 1938. Cooper Correspondence, Museum Accession File 1997.62. Wisconsin Historical Society, Madison. 1954 Letter to Mr. John Jenkins, December 10, 1954. Cooper Correspondence, Museum Accession File 1997.62. Wisconsin Historical Society, Madison. 1955 Letter to Mr. John Jenkins, November 21, 1955. Cooper Correspondence, Museum Accession File 1997.62. Wisconsin Historical Society, Madison. Davenport, G. Clark 2001 Remote Sensing Applications in Forensic Investigations. Historical Archaeology 35(1):87–100. Finley, Robert W. 1970 Geography of Wisconsin: A Content Outline. College Printing & Publishing, Inc., Madison, Wisconsin. Geister, David P., and Thomas G. Shaw 1998 American Troops at Prairie du Chien, 1814. Military Collector and Historian, Journal of the Company of Military Historians 50(2):88–89. Hargrave, Michael L., Lewis E. Somers, Thomas K. Larson, Richard Shields, and John Dendy 2002 The Role of Resistivity Survey in Historic Site Assessment and Management: An Example from Fort Riley, Kansas. Historical Archaeology 36(4):89–110. 229 Joyce, Dan J. 1997 The Outline of the Fort Crawford Excavations as They Relate to the Current Buildings at Villa Louis. Kenosha Museum, Kenosha, Wisconsin. Joyce, Dan J., and Jonathan T. Van Beckum 2010 Ground-Penetrating Radar Survey of Selected Portions of First Fort Crawford (47CR249) Prairie du Chien, Wisconsin. Report of Investigations, Number 1. Kenosha Public Museums, Kenosha, Wisconsin. Prepared for the Wisconsin Historical Society, Office of the State Archaeologist. Kane, Lucile M., June D. Holmquist, and Carolyn Gilman 1978 The Northern Expeditions of Stephen H. Long: The Journals of 1817 and 1823 and Related Documents. Minnesota Historical Society Press. Kaufmann, Kira E., and William F. Kean 2002 Archaeological Spatial Analysis at Aztalan State Park (47JE1): Using Geophysical Techniques to Interpret Prehistorical Landscape Features. The Wisconsin Archeologist 83(2):111–122. Kellogg, Louise P. 1935 The British Regime in Wisconsin and the Northwest. State Historical Society of Wisconsin, Madison. Mahan, Bruce E. 2000 Fort Crawford and the Frontier. Prairie du Chien Historical Society, Howe Printing Company, Inc., Prairie du Chien, Wisconsin. Martin, Lawrence 1965 The Physical Geography of Wisconsin. University of Wisconsin Press, Madison. Parrington, Michael 1979 Geophysical and Aerial Prospecting Techniques at Valley Forge National Historical Park, Pennsylvania. Journal of Field Archaeology 6:193–201. Prucha, Francis P. 1953 Broadax and Bayonet: The Role of the United States Army in the Development of the Northwest 1815-1860. State Historical Society of Wisconsin, Madison. Scanlan, Peter L. 1998 Prairie du Chien: French-British-American. Prairie du Chien Historical Society, Inc., Prairie du Chien, Wisconsin. Slota, Robert W., and Glenn D. Garvey 1961 Soil Survey of Crawford County, Wisconsin. United States Department of Agriculture, Soil Conservation Service. Series 1958, No. 18. U.S. Government Printing Office, Washington D.C. Smith, Alice E. 1973 The History of Wisconsin, Volume 1: From Exploration to Statehood. State Historical Society of Wisconsin, Madison. 230 GPR Survey First Fort Crawford, Prairie du Chien, Wisconsin Twinde-Javner, Vicki 2005 Rediscovery of Second Fort Crawford (47Cr247), a Nineteenth Century Military Post on the Frontier of the Mississippi River 1829 to 1856. Report of Investigations No. 570. Mississippi Valley Archaeology Center, University of Wisconsin-La Crosse, 2009a Fort Shelby, Fort McKay, and the First Fort Crawford, 1814–1831. In Frontier Forts of Iowa: Indians, Traders, and Soldiers, 1682–1862, edited by William E. Whittaker, pp. 75–84. University of Iowa Press, Iowa City. 2009b The Second Fort Crawford, 1829-1856. In Frontier Forts of Iowa: Indians, Traders, and Soldiers, 1682–1862, edited by William E. Whittaker, pp. 121-132. University of Iowa Press, Iowa City. Van Beckum, Jonathan T. 2006 The “Rediscovery” of the First Fort Crawford, a Study Using Archaeological and Archival Data to Interpret and Date an Early 19th Century American Fortification. Master’s thesis, Department of Anthropology, University of Wisconsin–Milwaukee. Wesley, Edgar B. 1935 Guarding the Frontier: A Study of Frontier Defense from 1815-1825. University of Minnesota Press, Minneapolis. Whittaker, William E. 2005 Ground-Penetrating Radar Survey of Historic Fort Atkinson (13WH57) and a Possible Fort Atkinson Cemetery (13WH210), Winneshiek County, Iowa. Contract Completion Report 1315. Office of the State Archaeologist of Iowa, Iowa City. 2006a Extensive Ground-Penetrating Radar Survey of Historic Fort Atkinson (13WH57), Winneshiek County, Iowa. Contract Completion Report 1392. Office of the State Archaeologist of Iowa, Iowa City. Joyce and Van Beckum 2006b Ground-Penetrating Radar Survey of a Portion of Second Fort Crawford (47Cr247), Prairie du Chien, Wisconsin. Contract Completion Report 1413. Office of the State Archaeologist of Iowa, Iowa City. 2007 Ground-Penetrating Radar Survey of Portions of the Fort Des Moines No. 2 Site, 13PK61, Southeast Connector Project, Des Moines, Iowa. Contract Completion Report 1480. Office of the State Archaeologist of Iowa, Iowa City. 2009a Ground-Penetrating Radar Survey of a Portion of the Fort Johnson Site, 1HA957, Warsaw, Hancock County, Illinois. Contract Completion Report 1694. Office of the State Archaeologist of Iowa, Iowa City. 2009b Testing the Effectiveness of Ground-Penetrating Radar at Three Dragoon Forts in Iowa and Wisconsin. Historical Archaeology 43(1):56–74. 2012 Notes on the 1993 Excavation of a Trempealeau Phase Middle Woodland Burial at the Mill Coulee Site, 47CR100. Manuscript on file, Wisconsin Historical Society, Madison. Wisconsin Historical Society 2015 History of Villa Louis. Electronic document, http:// villalouis.wisconsinhistory.org/About/History.aspx, accessed on September 15, 2015. Wisconsin State Cartographers Office 2004 Maps Available from the SCO. Electronic document, http://www.sco.wisc.edu/find-maps/ wisconsin-outline-maps.html, accessed January 12, 2011.