Colonial Williamsburg Digital Library

Natural Drainage

Some of the earliest layers on this site were formed by natural events rather than domestic or industrial activity. As mentioned earlier, a large ravine was located along the western boundary of the property. Identified during the 1974-75 excavation, numerous gullies stemmed off of this ravine, acting as drainage run-offs. The soil associated with this natural activity was characterized as wash, consisting of a loose soil that was silty and sandy (Figure 12). All were directly on top of subsoil, which explain their similar appearance. Only a few domestic artifacts were found in association with these layers, such as a tobacco pipe, slate, a nail, wine bottle glass and brick. None were diagnostic. Although these artifacts were not indicative of domestic activity within this specific area, they do attest to domestic activities nearby. The excavated layers relating to drainage were explored primarily in the mid-eastern section of the excavation area and appeared patchy. Although it is probable that these layers continued across the site in a westerly direction, they, as well as other layers and features, were left intact for future preservation along and contiguous to the western sidewall of the site.

Light Olive Brown Sand (10GE-287, -290), found in units 1095N/1098-1099E, was a highly mottled light olive brown (2.5Y5/4) loamy sand. It ranged from 2-10 cm in depth forming two moderately deep depressions in the subsoil. The second, Grey Mottled Sandy Loam (10GE-312, -316, -319) was located in units 1097N/1097-1099E. This soil consisted of mottled light olive brown (2.5Y5/3) and light yellowish brown (2.5Y6/4) sand that was 7-12 cm deep. The last of these washes, Grey Sandy Linear Feature (10GE-301, -302) was also found in unit 1097N/1098E. Due to its linear appearance it was originally treated as a feature, but has since been interpreted as a shallow drainage gully. The soil, measuring 5-8 cm in depth, was a mottled light olive brown (2.5Y5/3) and a very dark grey (2.5Y3/1) sandy clay.

Another feature sitting on top of subsoil was a small ovate posthole (10GE-292-294) whose function is unknown (Figure 13). Found in unit 1095N/1097-1098E, it measured 39 cm N-S, 25 cm E-W, and 10-13 cm deep. The soil was a yellowish brown (10YR5/6) sandy clay with very dark grayish brown (2.5Y3/2) and pale yellow (2.5Y7/4) mottling. Brick inclusions were present in this soil as well as in the square post mold which was located in its northern section. The postmold measured 20 Figure 11. Features discovered in 2000-01 excavation. 21 Figure 12. Erosion gullies on southern part of 1974-75 excavation (from Foss 1977). Figure 13. Plan and profile of 10GE-292-294. 22 12 by 10 cm and was 12 cm deep. The only artifact found in this feature was an unidentifiable, undiagnostic nail fragment.

Ravine Fill

Orange Clay with Brick Inclusions (10GE-M17) (also called "Clay Associated with Drainage Feature") was a layer that was identified during the 1974-75 excavation (Figure 14). It was described as a "spread of deposited mixed clay with some pink/orange brick chips" and labeled as 1967L in the 1974-75 Excavation Register. This layer was also referred to as "dumped yellow clay" in the 1975 archaeological report (Foss 1977: 68). This layer sealed most of the previously discussed layers and features, thus postdating them. The TPQ specified for this layer during the current excavation was 1720. However, according to Foss, their evidence pushed this date back to after 1750. In addition to artifacts, this TPQ was substantiated by the fact that this layer sealed features that dated between 1735 and 1740. Either way, this layer is still representative of early activities that took place on this lot.

Concentrated in the northern section of the excavation unit, this layer's southernmost extent was along the 1094N line. The soil was predominantly a yellowish brown (10YR5/8) clay mottled with a dark grayish brown (2.5Y4/2) loam. It contained a significant amount of medium- to large-sized brick inclusions, including glazed brick. There were an insignificant number of artifacts, consisting of ceramics, tobacco pipes, glass, faunal remains and nails. Of particular interest was a bone tool (10GE-310).

This layer denotes the first attempts by the lot owners to alter the landscape in order to make the land usable. The clay was used to fill in the depressions associated with the drainage gullies and ravine in order to level the land. Despite these efforts, the clay was absent along the western property line in this area. As a result, the slope of the ravine was still visible toward the western property boundary starting on the eastern side of the fence along the 1095E line. According to the 1974-75 archaeological report, this layer did extend further west in other areas of the site (Foss 1977: 68-69).

After the deposition of this clay layer, the ravine began to be filled. A thin layer called Grey Mottled Silty Clay (10GE-M16) represented these activities but did not suggest that the area was heavily utilized during this early period (Figure 15). Measuring only 5 cm deep along the 1096E line, this layer became deeper as it approached the western boundary line eventually measuring 30 cm in depth. The soil was light yellowish brown (2.5Y6/4) mottled with olive brown (2.5Y4/3) and contained brick inclusions. Patches of whitish sandy loam and compact sand mixed with the soil suggest that at least part of this layer was created by natural wash. Despite its natural elements, the layer contained a few artifacts suggesting that domestic activities were taking place nearby. Few in number, artifacts included table, wine, case bottle, and pharmaceutical glass, ceramics, faunal remains and nails. Again, according to current research, this layer postdates 1720. Yet, when information from previous excavations is taken into consideration, the date of this layer is pushed back to postdate 1755 (Foss 1977: 68-69).

Figure 14. Orange Clay with Brick Inclusions.

Figure 15. South wall of 1974-75 excavation area, showing location of Gray Mottled Silty Clay.

Figure 16. Plan and profile of 10GE-295, -309, -296, -313, -315, -322

Figure 17. Photo of 10GE-295, -309, -296, -313, -315, -322.

Western Property Boundary-Fence Line

Cutting this layer in unit 1096N/1094E was a Large Square Fence Line Posthole (10GE-295, -309, -296, -313, -315, -322) containing a repair episode (Figures 16 and 17). This posthole is evidence of an early fence line delineating the western property boundaries. According to the 1974-75 excavation, this feature was part of Fence Line B, described as a series of rectangular holes at eight foot intervals. Most cut circular postholes which represent the first fence line on the lot. This second fence line represents the beginning of more intense cultural activity on the site and was suggested to have been built at the same time the first kitchen was erected on the property (Foss 1977: 66-67).

This particular posthole followed the same pattern identified by Foss and cut a circular posthole located along the eastern edge of the feature. Partially excavated during the 1974-75 excavation, the square posthole measured 62 by 60 cm and was 87 cm deep. Both the postmold and the hole itself were comprised of a number of fills. The first fill in the postmold (10GE-295) consisted of brown sand (10YR5/3) with light yellowish brown (2.5Y6/4) diagonal banding that was 63 cm deep. This banding effect appeared to have been created through natural wash, suggesting that at one 25 time, the post was pulled out of the ground and the posthole remained open and was filled in naturally over time. This post would have been square in shape with a flat bottom and appears to have been put into the ground at a slight angle leaning toward the west. It is possible that this angle was created to loosen the post for removal. There were only a few artifacts in this fill, including colonoware, a fire-cracked rock, and five flattened rods whose purpose remain unidentified.

At the very bottom of this fill was a thin layer of yellow clay separating the sand from a second mold fill (10GE-309), which consisted of a brown sandy clay loam with heavy charcoal inclusions. This fill actually outlined the sandy fill to the west and south and consisted of a brown (10YR4/3) sandy clay loam with heavy charcoal inclusions. It had a rounded bottom that was centered 6 cm below the square bottom of the sandy fill. Due to the distinct differences between the two mold fills and shapes, this fill represents the original post with a rounded bottom that was eventually replaced with one that had a square bottom. In order to execute the repair, the top of the post was pushed at an angle toward the north to loosen it for removal since part of this fill slightly undercuts the posthole fill in the south. Unfortunately, the northern extent of this feature was dug in 1974-75 and few notes were recorded about its appearance. However, it is assumed that the postmold fill followed this trajectory. Artifacts were characteristically different than those in the sandy fill as they were larger in size and included a number of delftware fragments, glass and faunal remains, a large number of wrought nails, and a piece of worked bone.

The first fill in the posthole itself (10GE-296) appears to have coordinated with the posthole repair. In order to pull it out of the ground, a squarish hole was dug to a depth of 25-32 cm around the post in order to loosen it. Then the round-bottomed post was replaced with the square-bottomed one and the hole was filled back in around the new post. The fill consisted of an olive brown (2.5Y4/3) silty clay loam with brick inclusions. The few artifacts recovered were reminiscent of those found in the sandy postmold fill and were characteristic of a typical eighteenth-century assemblage. The TPQ for this fill, and thus the repair episode, is 1759.

The second fill in the posthole (10GE-313) consisted of an olive brown (10YR4/3) sandy loam mottled with a light olive brown (10YR5/6) sandy clay measuring 26 cm in depth. There was a heavy amount of inclusions in the soil matrix consisting of charcoal, brick, mortar, and shell. Artifacts tended to be small in size and reflected a typical eighteenth-century assemblage. A piece of flint debitage was of special note. The last fill (10GE-315) looked very similar to subsoil except that it was mottled with a grayish hue. The soil consisted of light olive brown (10YR5/6) and strong brown (7.5YR5/8) sandy clays mottled with very dark grayish brown (2.5Y3/2) and light olive brown loams with brick inclusions. As the deepest fill, it measured 58 cm in depth and cut deeply into subsoil. Very few artifacts were recovered but included tobacco pipe fragments, nails and faunal remains. One piece of a sand tempered Native American ceramic was found. When taken into consideration with the fire cracked rock in this feature and others (10GE-42, 46, 169) as well as other Native American ceramics found on the site (10GE-103, 105, 155, 180, 221), these artifacts attest to the presence of Native Americans in this area at one time.

Domestic Activity

The last pre-Anderson context identified, Light Olive Brown Silty Loam (10GE-M15) (2.5Y5/3), was representative of a domestic sheet refuse layer postdating 1762. This layer was highly organic containing medium-sized brick inclusions and charcoal flecking, but lacked industrial waste. It started along the 1096E line and represents filling activities within the ravine to the west. Artifacts in this layer were characteristically large in size with faunal remains being the most represented. These bones tended to be large fragments with visible butchery marks. Additional well-represented artifacts included wrought nails and long tobacco pipes stems, one of which had the maker's mark "T/W" molded onto it (10GE-291). Interestingly, of the many pieces of window glass, two fragments (10GE-283) had writing etched onto them. Unfortunately the only discernible script was the letter "t," but it is obvious that this letter was part of a larger word which continued onto other areas of the window pane. This window glass could be possibly linked to that found in context 10GE-28, which had the letters "Puelli..." etched onto it (Figure 18).

Wine bottle glass was found in addition to a copper wire wine bottle closure (10GE-280). Pharmaceutical glass and table glass were also represented as well as various ceramics such as delftware, porcelain, coarsewares and stonewares. Also of interest was lead alloy casting waste (10GE-283, 286).

Domestic and Industrial Refuse

Luckily a very distinct difference was able to be made between Pre-Anderson and Anderson contexts, which helped in the interpretation of the building's function. In Figure 18. Inscribed glass (Artifact No. 10GE-00028AZ). 27 general, Anderson-related artifacts, such as ceramics, glass, tobacco pipes and faunal remains, were uniformly small in size compared to the size of artifacts found in earlier layers. However, there were some exceptions, such as thickly potted ceramic storage vessels and non-fragile objects like flint and iron. The amount of industrial activity on the site can account for this site formation process.

One large soil lens and two layers, all postdating 1775, are the only non-feature contexts that can be attributed to the Anderson period. The lens, called White Mottled with Brown (10GE-M12), was located in the middle of the site in units 1095N/1095-1097E and 1096N/1095-1097E. Reminiscent of a triangle in shape, its north-south leg measured up to 1.53 m and the east-west leg measured from 60 cm to 127 cm, with its hypotenuse oriented on a north-east line. Fairly shallow, it measured between 4 and 12 cm deep. It was cut by nineteenth-century features (10GE-191, -199) to the north and south. This lens was a light yellowish brown (2.5Y6/3) silty clay loam mottled with olive brown (2.5Y4/3). Its appearance was that of a white subsoil, but it contained brick, shell and charcoal flecking inclusions. There were few artifacts consisting primarily of glass, ceramics and tobacco pipe fragments. The glass objects were by far the most representative group, with window and wine glass being the most frequent. The most interesting artifact, however, was an unfinished key (10GE-243) that had been turned on a lathe with an unaltered bit. A similar one was found within the privy feature (10GE-176), to be discussed later.

This lens, as well as a significant number of features, cut through Dark Greyish Brown Loam (10GE-M13). This layer covered the majority of the site and tapered out along the 1098E line. The soil was a dark grey (2.5Y4/1) sandy loam mottled with light olive brown (2.5Y5/4) measuring 4-8 cm in depth. It contained brick inclusions whose concentrations ranged from light to heavy in an inconsistent manner. Again, there was not a significant number of artifacts, but they reflected a typical eighteenth-century assemblage. Given that this area was located underneath the building, this is not surprising. Artifacts of note were tin and colonoware (10GE-227) and a worked bone-cutlery handle (10GE-247) (Figure 19).

Figure 19. Bone handle (Artifact No. 10GE-00247AM).

The second layer, Yellow Sandy Loam, began along the 1097E line and was sealed by Dark Greyish Brown Loam. The interface of the two layers in profile formed a wedge-shape and in plan view formed a distinct line that occurred in units along the1098E line. Yellow Sandy Loam continued east underneath the sidewalls of the excavation unit into Mrs. Ryland's Garden. 28 The soil was a light olive brown silty-sandy loam with brick flecking. Of these layers, those closest toward Mrs. Ryland's Garden had more artifacts than those toward the west. In general, these layers had a higher concentration of artifacts than those previously discussed. Again, the artifacts reflected a typical eighteenth-century assemblage. Of the ceramics, delftware and white salt-glazed stoneware were the most represented wares, with coarsewares and stonewares close behind. When the artifacts from eighteenth-century contexts were plotted, general concentrations of ceramics and glass were noted in the southeast corner of the site (1092N/1099E) (Figure 20). Colonoware and container glass were some specific types of artifacts concentrated here (Figures 21 and 22). This concentration continued in a southerly and easterly direction out of the excavation area into Mrs. Ryland's Garden. Artifacts of note include flint debitage and a window lead (10GE-167), an iron harness buckle (10GE-169), tin (10GE-172, 188), and jeweler's rouge (10GE-239), used to polish metals. With a high number of domestic rather than industrial-type artifacts, it is possible that this building had a domestic element to it. Perhaps there was a loft where some of Anderson workforce lived.

The main focus of this excavation centered on identifying the industrial activities that took place in the third southernmost industrial building shown on the Frenchman's Map. Soil associated with Anderson's tenure was additionally characterized, not only by small artifact size, but also by a high concentration of coal, coke, clinker, slag and charcoal. In general, many of these terms are used interchangeably by archaeologists and have been and will be referred to as industrial waste in this report. However for the purpose of defining specific activity areas at Anderson, distinctions were noted as certain forms of industrial waste represent highly specific activities (Figures 23-25).

Following is a list of definitions applied by the archaeologists working on the site. Coal was the most efficient fuel used by blacksmiths. Since it was mined from the ground, shale was frequently found in association and was recovered from the site. Historical documentation indicates that Richmond was the primary source of coal at the time (Ken Schwarz, personal communication). Partially burnt coal was defined as coke. Highly porous and light, it was recycled and reused until completely burnt. Clinker, a bubbly dense mass, was the final by-product representing the impurities in the coal combined with ash. Any fire, whether in the home or on an industrial site produced these products. However the quantity of these materials distinguishes between an industrial versus a domestic site.

For this study, clinker was saved, weighed and charted. Patterns show the major concentrations of clinker located toward the edges of the site on the exterior of the structure. In comparison to the massive amounts of clinker recovered archaeologically from the interior of the reconstructed forge shops, the lack of clinker argues that this particular building did not have a forge associated with it. In addition, no structural remains of a forge were located. A concentration of clinker in features outside of the southwest corner of the building show a deliberate disposal pattern of this material away from the source of its production. This is not surprising as clinker must be removed frequently from the forges in order to sustain the firing operation. Another factor affecting the clinker's displacement was the slope of the land toward the ravine at the western edge of the property and water run-off patterns created by it.

Figure 20. Distribution map of concentrations of general ceramics.

Figure 21. Distribution map of concentrations of colonoware.

Figure 22. Distributions map of concentrations of container glass.

Figure 23. Distribution map showing concentrations of industrial waste.

Figure 24. Distribution map showing concentrations of tinned iron.

Figure 25. Distribution map showing concentrations of lead objects.

Slag, another industrial waste product, is a very dense, glassy material formed during the metal smelting process (Figure 26). It represents the impurities within the metal being melted. Although slag and two crucible fragments were recovered, they were mixed in twentieth-century contexts (Noël Hume Excavation [10GE-M3] and Loam with Shell and Brick [10GE-M5]). In order to cast metal successfully, the fire/melting area needed to be a short distance from the actual molds. Otherwise, the metal hardened too quickly in the crucibles and could not be poured. Some metals, such as brass, bronze and silver, have a high melting point and require structures that would allow for a forced air draft to fuel the flames. This would require bellows and tuyeres. From the evidence already presented, the possibility of this type of metal smelting in this area can be ruled out.

The Building and Evidence for a Tinsmithing Operation

In order to gather information about industrial activities associated with the building, the structure first needed to be located. This particular building was identified in the 1940s through cross-trenching activities. According to the 1943 report, only the southwest corner of the building and part of its east wall were found (Knight 1943: 3) (Figure 27). In 1974-75, archaeology exposed the remains of a possible robber's trench in both the north and south ends of unit ER 1943/1967, which was located directly north of the current site. This feature was described as being 1 foot thick with orange brick rubble whose westerly edge lined up with the property boundary. The artifacts date it to the "colonial" period (Noël Hume 1974-75). Consequently, Knight listed the measurements of the bricks which he uncovered in 1943 as 8½ inches long and 4 inches wide and commented that they were laid in English bond. Thus, if the dimensions of one header and one footer are measured, the resulting width of the wall would be approximately 1 foot wide, coinciding with the robber's trench found in 1974-75 (Figure 28). Additionally, the orange brick from the robber's trench would match the "buff" brick that Knight described in his report (Knight 1943: 3). Therefore, it can be assumed that the robber's trench identified in 1974-75 was part of the west wall originally identified by Knight.

The 2000-01 excavation revealed that the only extant remains of the building in the excavation area was its southwest corner, measuring two courses wide and at least two courses deep (Figure 29). Evidence from the 1940s trenching activities was still visible and thus additional information was able to be gathered about the building. The remains of the south wall trench dug by Knight abruptly ended four meters (approximately 13 feet) to the east in unit 1093N/1098E. Knight's crew dug to sterile soil along this line exposing the downward slope of the landscape toward the ravine in the west. The elevation of subsoil at the eastern end of the south wall trench was at the same height, if not higher than the elevation at the top of the remaining brick foundation. 33 Figure 27. Southwest corner of building and east wall (from Knight 1943). Figure 28. 1974-75 sketch map showing robber's trench (from Excavation Register [E.R.] notes). Note north is at bottom. 34 Figure 29. Brick foundations. The lack of a builder's trench indicates that the brick foundation was laid directly onto or very shallowly into the ground. If the brick foundation continued for the entire length of the south wall, the southwest corner, at one time, had to be significantly higher than its current measurement and would have gradually tapered toward the east in order to accommodate the slope of the land.

A linear feature (10GE-178, -179), found in unit 1093N/1099E, extended off of the west end of Jimmy Knight's south wall trench. This feature measured approximately 1.10 meters (3.6 feet) long and continued underneath the sidewalls of the excavation unit into Mrs. Ryland's garden. It was approximately 34 cm (1.1 feet) wide and 7 cm deep. A short perpendicular extension ran off of its north edge ending abruptly after 20 cm. Small brick inclusions were found in the fill along with artifacts. Domestic debris such as ceramics, table and wine glass, faunal remains and nails date the feature to post-1775.

It is possible that this feature represents the continuation of the south wall of the building. Having approximately the same width, it was almost in line with the projection of the south wall based on the corner foundation. The single brick found adjacent to this feature was suggested by Jimmy Knight in his map to have been the remains of the eastern wall of the building. However, this brick was not and did not appear to have been in course. Additionally, it was found in a late nineteenth-century layer that sealed this feature. It is possible that this feature represented a continuation of the Jimmy Knight trench, although its soil characteristics, depth and shape are drastically different from the Jimmy Knight trench that cut it to the east. If this is a continuation of the building, its shallow depth and lack of significant brick suggest that a wooden sill, rather than a brick foundation, might have been laid directly onto the ground in this area. Without uncovering the full extent of this feature, its interpretation remains inconclusive.

Unfortunately, the nature of the 1974-75 archaeological excavation destroyed the west wall of the building in the northern section of the site and cut Jimmy Knight's west wall trench. The remains of the robber's trench in the north wall of the Noël Hume unit were not located during the current excavation. However, as mentioned before, excavation notes and site maps taken in the field suggested that the building ran to the northern end of the Noël Hume unit, measuring 10 feet north to south. Therefore, the distance between the southwest corner of the building and the northern edge of the Noël Hume unit is at least 22 feet.

It must be remembered that only one corner of the building has been positively identified and that the exact dimensions of the building have yet to be determined. According to Colonial Williamsburg Foundation architectural historians, industrial buildings of this nature did not usually conform to standard sizes. In general, though, a building such as this would more than likely be under 30 feet in length due to the tremendous cost of successfully framing a building that exceeded this measurement (Carl Lounsbury, personal communication). Additionally, the distance between the southwest corner of the building and the reconstructed shop directly to the north of it is approximately 28.7 feet, limiting its size. The Frenchman's Map indicates that the these buildings were not connected, unlike the first and second industrial buildings shown; therefore the maximum length of this building had to be under 28.7 feet in length.

No evidence of a level, hard-packed earthen floor or brick/tile floor within the building was found. Combined with archaeological evidence to be soon discussed, this suggests that the work being done here was primarily a bench trade. In fact, the presence of two sets of parallel drain features (10GE-262, -266, -265, -271, -323- 36 Figure 30. Plan and profile of drainage features. Figure 31. Drainage features. 326) implies that the floor was raised (Figures 30 and 31). These features were underneath a series of layers that were formed during Anderson's tenure on the property suggesting that they were created at an early date. Although these drainage features were not as elaborate as the brick drains or the clay and wooden drain found in the northern shops, they do show Anderson's concern for drainage across the site. All extend underneath the sidewall of the excavation area into Mrs. Ryland's Garden. Characteristically, they averaged 20 cm in width and contained anywhere from 1 to 14 37 cm of silt loam fill. Each set of drains was approximately 18 cm from its partner and neither contained a significant amount of artifacts.

The north set of drains were significantly longer and more well defined than those in the south. These drains partially cut through subsoil and a patchy layer of yellow clay that was deposited to fill in the ravine. This clay was not found lining the bottom of either of these features and there was no evidence of their being lined with wood or brick. In general, the soil was a consistent brown silt loam (10YR4/3) forming well defined walls that formed a flat-based U in profile. Feature 10GE-265 measured 2 meters in length. Feature 10GE-262 was the longest of the two measuring 2.68 meters with a 35 cm extension at its east end. Forming an L-shape, it is reminiscent of the drainage feature found in Shop A-2 during the 1974-75 excavation (Foss 1967: 29, Figure 3) (Figure 32).

The south set of drains were not as well defined as those in the north. Both contained a light olive brown silt (2.5Y5/3) mottled with a very dark grayish brown loam (2.5Y3/2). Feature 10GE-323 measured 1.07 meters in length and contained charcoal inclusions and artifacts in its soil matrix. It had an undulating bottom. Feature 10GE-325 measured 1.04 meters in length and was devoid of artifacts.

Figure 32. Plan map of L-shaped drainage features.

In general, tinned iron was well represented on the site with over 200 fragments recovered. Historical documentation clearly shows that Anderson employed tinsmiths in his operations (Figure 30). Until this time, however, documentation could not pinpoint the location of this operation. Although limited at first, the arrival of Nathaniel Nuthall, a "Tin Man," on June 25, 1779, increased tin production considerably (Daughters of the American Revolution Daybook 1778-1780). At this time, the account books of Anderson were dominated with orders for various war-related objects. Even after Anderson moved to Richmond, documentation shows that the production of tin objects continued in Williamsburg. These objects included camp kettles, plates, dishes, cups, coffeepots, sauce pans, tumblers, cans, mess kits for officers, lanterns, funnels, quart measures, speaking trumpets, tubes, and canisters for ball shot (DAR Daybook 1778-1780).

Tinned iron was delivered to Anderson's business in the form of thin sheets as is referenced in the Daughters of the American Revolution Daybook. On July 5, 1779 an entry recorded that "1971 of Rusty Small tin Sheets" was delivered to Anderson's business. These sheets were approximately 15 inches square and were then cut and shaped into objects as seen in Diderot's Encyclopedia, Plate 180 (see Figure 33). The unusable and non-recyclable clippings were then discarded primarily on the exterior of the building toward the ravine, in and around the privy. The lack of tinned-iron concentrations within the interior of the building confirms this discard pattern. However, one concentration of tin, presumably within the building, correlated with a concentration of lead in the northeast corner of the site. This is not surprising since lead solder, loosely a fifty percent mix of lead and tin based on weight, was an important aspect of this trade, sealing the seams of the tin objects.

A limited amount of lead from the site was tested for the presence of tin to see if solder could be identified. Unfortunately, none of the tests had a positive result, but this is most likely the result of a small sample size (see Appendix 4). The negative results in no way deter the interpretation of this building as a production site for tin objects. It simply suggests that tinsmiths were careful with the use of solder and that once the objects were completed, they were removed from the property. Thus these materials and objects were eliminated from ever entering the archaeological record. However, evidence of tin production is still highly visible through the large quantity of tinned-iron fragments.

Although the possibility of complicated metal casting has been ruled out, not all metals required extensive operations. Lead, for example, melts easily and quickly over the simplest flame which could have been produced in free-standing braziers as seen on the table in Plate 180 and in the right corner (Plate 190). A number of lead objects were found in layers and features associated with Anderson's tenure on the property. Specifically these objects were lead shot, window leads, a piece of scrap, and unidentified hardware. When considering lead objects found in related and contiguous contexts (more than likely mixed into the matrix by later activities), the amount of lead present on the site was immense. Of interest were casting waste, solder, and shot. One piece of shot in particular had an attached sprue that showed it had been cast in a gang mold.Given the extensive range of Anderson's business during the war in conjunction with historical documentation, specifically the Daughters of the American 39 Figure 33. Tinsmith working images shown in Diderot's Encyclopedia (1763). Top, Plate 180: The Tinsmith's; bottom, Plate 190: The Brazier I. Revolution Daybook, it is safe to assume that lead objects were being cast on this particular property.

Aside from the presence of casting waste and lead shot, other artifacts support this theory. Historical documentation noted that raw materials were scarce during the war. Earlier, it was noted that the cupola on the Capitol had been stripped of its lead during this time as a result (Letter, Col. George Muter to Thomas Jefferson, December 12, 1780). There were a number of unusual artifacts that would not have been present on the Anderson property unless they had been specifically brought to the site. For example, the presence of window leads (which were probably not incorporated into the building fabric of Anderson's shops) and two pieces of printer's type (probably acquired from the print and bindery shop across the street) (10GE-44, -155) 40 suggest that lead was being scavenged and recast by Anderson's workforce. Distribution maps show that lead objects were being discarded in the privy. According to distribution maps, this activity was concentrated in the northeast corner of the site and continued into an area underneath Mrs. Ryland's garden.

An entry in the Daughters of the American Revolution Daybook, dated August 18, 1779, lists the delivery of "960 wt of lead [to Anderson's business] to make Ball." Less than a month later on September 4, 1779, Anderson "Deliverd Leut Milbanks 1161 lb of Muskett Ball" (DAR Daybook). Given the extent of lead production undertaken by Anderson, as is evidenced by these two entries, it is not probable that this particular building housed the major lead production center. The concentration of lead in the northeast corner might be an indication that this type of lead smelting operation was nearby. It is also possible that this type of work was not performed on this property at all. More than likely, the lead found in association with this building was that used to complete the tin work.

Privy

Located outside of the southwest corner of the building was a privy (10GE-175, -?176, -177) measuring 1.07 meters (3.5 feet) by 71 cm (2.32 feet) and 61 cm (2 feet) deep (Figure 34). This is an extremely important feature in that it has aided in the interpretation of this building and offers insights into the health and activities of Anderson's workers. The privy ran parallel to the south wall of the building (length-wise) and was slightly offset from the corner. Whether this building was a lean-to or free-standing structure is unknown. No postholes or foundations were found that might answer this question.

The privy feature consisted of two fills. The top layer (10GE-175), which postdates 1805, represents efforts to fill the structure once it fell into disuse after Anderson's death. Measuring between 30 and 40 cm deep, this fill had a noticeable line of grey loam around the outer edges, suggesting that the privy was once lined with wood.

This layer sealed a highly loamy, organic layer (10GE-176) postdating 1775. Measuring between 20 and 32 cm deep, the soil consisted of an olive brown silt loam layer (2.5Y4/3) mottled with dark yellowish brown. Charcoal flecking, brick bats, shell mortar and oyster shells were present throughout. Industrial waste, although weighing 421.8 gm total, did not seem to be heavily represented as compared to other Figure 34. Plan and profile of privy feature. 41 inclusions and artifacts. Most surprisingly, lime, an agent used to control privy waste and odor, was barely present. On an industrial site already subject to the pungent odor of coal fires, smell might not have been a concern for Anderson and his crew. Aside from the standard mix of eighteenth-century artifacts, copper straight pins, lead shot, flint debitage, a window lead, and fragments of tinned iron were included in the matrix. The most interesting artifacts were the unfinished iron pieces, definitively attributing the deposition of this layer to the Anderson period. These included a possible toasting fork, grape shot, nail rod, and a washer. Additionally an unfinished key which matched that found in the lens called White Mottled with Brown (10GE-243) was recovered. Sandstone, which would have been utilized in the form of a grinding stone, was also found in this context, as well as seven others across the site.

A large quantity of faunal remains were collected from this context, although typical of Anderson collections, these remains were characteristically small in size. In addition, soil samples were floated in the hopes of recovering additional faunal materials and botanical remains. Although there was a large number of miniscule bone fragments found in the flotation samples, very few botanical remains were recovered. Preliminary analysis identified a grape seed and a few other non-food related plants (Steve Archer, personal communication). In an undisturbed context such as this, the lack of information can be attributed to either non-deposition or to acidic soil conditions which do not preserve ethnobotanical remains well.

As seen with the concentration of unfinished iron work, this feature, aside from its intended purpose, was also utilized for trash disposal, as is common with most privies. When artifacts were plotted across the site, this feature consistently showed concentrations of all the various artifact classes such as ceramics, glass, tobacco pipes, and wrought nails. However, there were some exception of which to make note. For example, the concentration of faunal remains was not influenced by the presence of oyster shells. The concentration of ceramics, lacked colonoware, despite its presence on the site. The concentration of glass objects lacked container glass, but contained one of the highest concentrations of pharmaceutical glass on the site. This suggests that health and hygiene was of some concern to Anderson and his workforce.

Almost all individuals living in the eighteenth century were inflicted with parasites. It was not until the twentieth century that efforts to eradicate these problems were made. Intestinal roundworms (Ascaris lumbricoides), whipworm (Trichuris trichiura), hookworm, and tapeworm (T. lacnioid) were the most common parasites afflicting populations within the New World. These parasites are commonly identified in eighteenth-century privies and vary in frequency according to urbanization, occupational differences and the social-economic class of the individuals under investigation. In general, everyone was afflicted to some degree with worms, but "insufficiencies of poorer peoples tend to coincide with heavier worm burdens" (Reinhard et al. 1986). The spread of these parasites is primarily attributed to the consumption of vegetables grown in soils fertilized with human and/or animal fecal material.

The daybooks of Dr. John M. Galt, an apothecary working in Williamsburg, confirm that Anderson and his family suffered from this affliction (see Appendix 5). On December 26, 1782, Dr. Galt administered "Pulv. Cath. Merc." and on September 25, 1783 administered "Pulv. Cath. & sal. Cath., Pulv. Vermifug dos. viii" (Dr. John 42 M. Galt Daybook, December 17, 1782-September 30, 1783, M-55.1). These records document Dr. Galt administering worming powder (vermifuge) and various pulverized, saline and mercurial purges (cathartics) to Anderson and his family. These cathartics caused the bowels to empty, hopefully expelling the pesky parasites in the process.

Parasite analysis was conducted on soil samples taken from context 10GE-176, the lower of the two privy fills, in order to try and identify the presence of these parasites within archaeological contexts related to Anderson. Unfortunately archaeoparasitology did not identify parasite ova or cysts. According to Leslie Driscoll of the Andrew Fiske Memorial Center for Archaeological Research at the University of Massachusetts Boston, who performed the analysis, "while truly abundant ova would indicate presence of fecal material in considerable amounts, the absence of the eggs does not prove that the deposit was not fecal in nature." It is possible that parasite eggs washed or percolated away over time (see Appendix 6). Parasite analysis conducted on an early eighteenth-century privy from the Brush-Everard property in Colonial Williamsburg's Historic Area positively identified intestinal roundworms and human whipworm in Williamsburg (Samford and Moodey 1989: 20). It is easy to assume then, that these were some of the same worms for which Anderson and his family were treated.

A study on a blacksmith's privy located in Newport, Rhode Island dating to the third quarter of the eighteenth century provides an excellent comparison to that at the Anderson site and offers a possible explanation as to why no parasites were identified in the Anderson samples. Although high levels of intestinal roundworms and human whipworm were identified in this privy's soil, it was noted that the fecal deposits were diffuse, indicating that the privy was not used extensively for the disposal of human waste. Ceramics from the privy, however, were similar to those found in associated sheet refuse, suggesting that the privy was used more often as a trash pit than as a necessary (Reinhard et al. 1986). The privy at Anderson followed this same pattern. Ceramics found within the privy match those found in associated contexts. From the quantity of artifacts found in conjunction with disposal patterns identified by distribution maps, it is clear that this feature was used as a trash pit more than as a privy. This could explain the reason for both the lack of lime and the lack of parasites within the soil. Perhaps chamber pots were used more often and emptied within the yard or over the fence into the adjacent ravine.

Other health concerns in eighteenth-century Williamsburg were recorded year by year in the accounts of Dr. John de Sequeyra, a physician working in Williamsburg. Malaria, measles, diphtheria, typhoid, pneumonia, whooping cough, and measles were examples of diseases that he treated throughout the years (Gill 1972: 99-115). Whether or not Anderson and his workforce were afflicted with and/or treated for any of these diseases is presently unknown. There are only a few references to Anderson's workers getting ill. In the DAR Daybook, one entry, dated July 25, 1778, lists Anderson paying "To Cash by TL when sick 12" (DAR Daybook 1778-1780). Aside from viral and bacterial infections, those working for Anderson would have been particularly susceptible to ailments caused by occupational hazards. Arthritis, tendonitis and eye problems were common as well as burns and puncture wounds. Numerous accounts from Dr. Galt's daybook lists Anderson providing medical treatment for his workers. 43 For example, on December 24, 1782, Anderson had Galt dress an apprentice's leg and on a few occasions had Galt dress a Negro woman's finger. On February 12 and July 7, 1783 Anderson had Galt administer pulverized cathartics to an apprentice (Dr. John M. Galt Daybook, December 17, 1782-September 30, 1783, M-55.1). Although the exact nature of these wounds or how they were inflicted is not indicated, puncture wounds, burns, and worms often required such medical attention (Shell Browder, personal communication).

Fence Line

As has been evidenced in previous archaeological studies, a well-maintained fence line along the western boundary of the property existed (Figure 35). Aside from marking lot lines, this fence line also provided protection from the ravine immediately adjacent to the property. Two sets of postholes, whose centers were approximately eight feet apart, were uncovered from this excavation. The first posthole was located contiguous to the southwest corner of the building (10GE-216, -211, -212) and contained a post mold. This mold was located in the northwest corner of the feature with rounded edges, except where it abutted the edges of the corner. Its diameter measured approximately 26 cm and was 36 cm deep. The mold was characterized by a very dark grayish brown clay loam with a significant amount of inclusions such as clinker, mortar, shell and marl. Of particular note were the brick bats and whole bricks mixed into the matrix. Artifacts reflect a typical eighteenth-century assemblage. The post hole itself, measured 64 cm N-S by 58 cm E-W, and was 36 cm deep. The fill was a brown loam with orange clay mottling containing the same inclusions as those Figure 35. Plan and profile of posthole series. 44 found in the postmold. Of note were the large pieces of clinker and brick bats in the fill. The presence of clinker suggests that this post was erected after Anderson had began working on the property. Artifacts were typically eighteenth-century. Of note was a tobacco pipe bowl with a foliate design and a dressed slate fragment with either an "E" or "F" written upon one side and markings (possibly a "v" or "a") on the other. Both fills had TPQs of 1775.

The second posthole (10GE-223, -224), located in unit 1095N/1094E, was cut by a nineteenth-century posthole erected during Nancy Camp's ownership of the property. The remaining south and east edges measured approximately 55 cm N-S by 40 cm E-W and the hole was 44 cm deep. At approximately 15 cm from the top, the side formed a step that angled downward. There was probably a postmold at one point in time, but this was more than likely destroyed by the fence line repair. The soil consisted of a dark grayish brown loam with a significant amount of clinker and typical eighteenth-century artifacts including faunal remains, tin and nails. Its TPQ was 1787, reconfirming that Anderson erected these postholes after work began on the property. A heavy amount of brick inclusions and brick bats, similar to the first post hole, were also present. Although, the brick bats in this feature seemed purposefully organized, lining the outer edges toward the top of the feature. The deliberate placement and number of brick bats within both of these postholes suggest that Anderson was using brick to shore up the post. In comparison to the pre-Anderson posthole whose depth was 87 cm, perhaps the shallow depth of around 40 cm was not enough to stabilize it.

This second posthole in turn cut a small square feature (10GE-229, -230) that was located along its eastern edge. This small feature was filled with brick bats, brick inclusions, industrial waste, tin and metal. Measuring only 28 cm E-W and 20-24 cm deep (a N-S measurement was not possible since that information was destroyed by the cutting posthole), this feature was certainly not large enough to have been for a post. Of what little soil there was, it consisted of a very dark grayish brown loam mottled with a light olive brown silt. It was created somewhere between 1775 and 1787.

Located on the western side of this fence line was evidence for the ravine (Figure 36). One layer and a deep gully were located in units 1092N/1094-1095E and 1093N/1094E adjacent to the southwest corner of the building. The layer, which was labeled Mottled Grey-Brown with Heavy Inclusions (Ravine Fill I) (10GE-203, ?204, -215), was characteristic of layers created by natural drainage activity and gently sloped toward the west. Approximately 10 cm deep, the soil was a dark grayish brown (10YR4/2) silty clay loam that contained a tremendous number of inclusions such as coal, clinker, brick, mortar and shell. It also contained a large number of artifacts which is not surprising given the disposal patterns and drainage patterns identified on the site. Aside from the typical eighteenth-century artifacts, tin, two bone buttons, and sandstone were present. Mottled Grey-Brown with Heavy Inclusions (Ravine Fill II) (10GE-209, -210) was exactly the same as the layer above except that it was more than 63 cm deep. It is possible that Ravine Fill I and II are the same, but they were separated once the fill began to dive. This gully was oriented on a north-south line and continued underneath the sidewall of the excavation unit toward the west. Both postdate 1775.

Figure 36. Circular postholes in relation to one another.

Figure 37. Circular postholes in relation to one another.

A series of shallow circular postholes were found in different areas of the site, but it is not evident how or if they relate to one another (Figure 37). Since they were all similar in character, they will be described as if they were related. Although too small to be significant postholes, perhaps they represent evidence of small driven stakes. The first (10GE-213, 214), was found at the western edge of unit 1096N/1094E and continued underneath the sidewall of the excavation unit. It had a diameter of 26 cm and a depth of 20 cm, with a profile shape of a flat-based U. The soil consisted of a 46 loose, very dark grayish brown (10YR3/2) loam with shell, brick and coal inclusions. Its TPQ was 1775. The second (10GE-225, 226) was located in unit 1093N/1099E and averaged 26 cm in diameter. It too was a flat-based U in profile filled with a dark grayish brown (10YR4/2) sandy loam. Charcoal and brick inclusions in addition to artifacts with a TPQ of 1775 were present in the matrix. Of particular interest was a one-piece copper alloy button. The third posthole (10GE-263, 264) was approximately seven feet from the second, although it varied slightly in character. Located in unit 1095N/1099E, it measured only 17 cm in diameter and 23 in depth. In profile, it formed a U-shaped bottom. Its soil was a black (10YR2/1) sandy-silty clay loam containing brick inclusions. The few artifacts found in the feature gave it a TPQ of 1762.

There were three other features that seemed to be independent in relation to others identified on the site. Although their TPQs or lack there of suggest that they might actually be pre-Anderson features, the small size of the artifacts in each context suggests that they were created during Anderson's tenure. There was a large posthole (10GE-305-307) located in units 1097N/1097-1098E, which was cut in the west and south by twentieth-century features. What was visible was squarish in shape measuring 54 cm N-S, 46 cm E-W and 20 cm deep. The fill was a brown (10YR4/3) sandy loam mottled with yellowish brown (10YR5/6) clay and contained brick and shell inclusions. A post mold was located in the northeast corner, measuring 31 cm N-S, 27 cm E-W and 20 cm deep. Its fill was a brown (10YR4/3) sandy clay loam which also contained brick and shell inclusions. Artifacts were similar for both elements, except the post mold contained a significant number of tin artifacts and clinker. Although the TPQ for this feature is 1650, based on stratigraphy it postdates 1762.

Posthole with Shell (10GE-273, -269, -270) was a small rectangular posthole that was characterized by the significant number of brick inclusions in its fill. This feature was sealed by a twentieth-century feature and measured 36 cm E-W, 26 cm N-S, and 36 cm deep. It had a rounded base and contained a dark yellowish brown (10YR4/4) silt loam fill. A small, round post mold, distinguished by a concentration of whole oyster shells, was located in the southwest corner measuring 15 cm in diameter and 28 cm deep. This fill was a brown (10YR4/3) silt loam and also had a rounded base. Both fills contained few artifacts and lacked industrial waste. Given this as well as the fact that the post hole fill had a TPQ of 1650 it is possible that this is a pre-Anderson feature, but it definitively postdates 1762.

The last feature was a Square Posthole (10GE-267, -268) that was cut to the north by a twentieth-century feature. What was visible measured 26 cm E-W, 10-15 cm N-S, and 10-17 cm deep with a flat bottom. No post mold was identified. The fill consisted of a brown (10YR4/3) sandy loam with brick inclusions and a few pieces of coal. The few artifacts that were recovered were undiagnostic, but the feature postdates 1762. As with all three of these features, none seemed to be related to any other features that were identified in the excavation area. Thus, their functions were difficult to interpret.

Transformation of Property

There is little sheet refuse associated with nineteenth-century activities except for an area located along the western edge of the excavation unit. Extending out approximately 30 cm from the baulk, this layer, Dark Loam with Heavy Inclusions, consisted of a dark grayish brown loam (10YR4/2) mottled with dark brown soil (10YR3/3) and contained a high amount of inclusions. The inclusions were characteristically small in size and consisted of brick, mortar, charcoal, shell and industrial waste. It is probable that at one time this layer extended further eastward. However, it was cut to the east by Jimmy Knight trenches and to the north by the Noël Hume excavation. If it extended past these features, the layer was destroyed by twentieth-century activities.

Due to the high number of brick and mortar inclusions, it is probable that this layer reflects the decay and removal of the industrial buildings present on the property after Anderson's death. In general, the artifacts recovered primarily contain eighteenth-century materials reflecting Anderson's work mixed with nineteenth-century domestic artifacts. In total, 862.7 grams of clinker, 142 iron objects comprised of hardware and nails, 27 tinned-iron fragments, coal, shale and slate were incorporated into the matrix. Of particular interest is lead shot (Figure 38) and a bone toothbrush fragment. Artifacts from the Anderson collection suggest that both of these items were manufactured on the property. There are a high number of artifacts, aside from the toothbrush, which reflect domestic activity, such as wine and pharmaceutical bottle glass, ceramics, tobacco pipe fragments, and faunal remains. These artifacts are typically found in both the Anderson-related contexts and nineteenth-century assemblages.

Figure 38. Piece of lead shot (Artifact No. 10GE-00186AZ).

There is strong evidence that Nancy Camp began to alter these buildings shortly after her acquisition of the property in 1806. Of the two soil layers found within the privy, the top layer provides the evidence. Distinct patches of yellow-orange clay (10YR6/6) and whitish sand (2.5Y6/4) were interspersed with patches of grey loam (2.5Y5/2). This patchiness suggests a rapid filling episode. This sandy loam was practically devoid of industrial waste and contained a significant number of faunal remains and long tobacco pipe stems. Artifacts from this layer have a TPQ of 1805, indicating that the privy was filled after this date. There are examples, such as the Grissell Hay property in Colonial Williamsburg's Historic Area, of privies in particular being filled once the property changed ownership (Meredith Poole, personal communication). However, this does not imply that Camp removed the extant industrial structures on the property at this time.

A Mutual Assurance Policy record dating to April 1823 states that the buildings fronting Duke of Gloucester Street are contiguous to "three brick buildings covered with wood, and five other wood buildings" (Mutual Assurance Policy #5012) (Figure 39). The three brick and wood buildings were more than likely buildings found on abutting properties belonging to Mary Stith and Thomas Sands. Of the five wood buildings, one was the smoke house that was located between the kitchen and the main house. The remaining four wooden buildings were probably extant buildings from Anderson's business. An insurance policy written seven years later, dated May 1830, only references one wooden building, suggesting the four remaining wooden industrial structures had been removed previous to this date (Mutual Assurance Policy #7578) (Kevin Kelly, personal communication) (Figure 40).

Archaeological evidence also confirms the presence of Anderson's work buildings on the property into the early nineteenth century. In general, soil layers associated with the rapid destruction of buildings tend to be thick and filled with large-sized architectural inclusions. Evidence for the slow decay of the buildings is seen in the fact that Dark Loam with Heavy Inclusions was not unusually thick, measuring approximately 10-15 cm in depth, and contained very small-sized architectural inclusions. Additionally, the lack of a burn layer associated with the 1842 fire that destroyed all of the buildings on this block also suggests that the buildings were not standing at this time. Therefore, it can be inferred that the buildings slowly fell into disrepair and the above-ground remains were eventually removed from the property. In general, abandoned and decrepit buildings within town limits were removed, rather than allowed to decay in place for an extended amount of time (Andrew Edwards, personal communication). This type of activity would leave virtually no artifactual evidence.

A drainage gully also hints at the presence of Anderson's shops on the property into the nineteenth-century. This feature was amorphous in shape despite its linear orientation on a slight diagonal from east to west measuring 270 cm in length. Toward the east end (1095N/1099E) it was fairly rounded and tapered into a narrow channel toward the west (1095N/1095E, 1094N/1095E). In this area, it was cut by a Jimmy Knight trench. The shallow fill (approximately 5-11 cm deep) was a Dark Greyish Brown silt loam (2.5Y4/2) that contained few artifacts. Again, the artifacts reflect the work that Anderson was doing on the property as well as nineteenth-century domestic refuse.

A naturally formed erosion gully such as this says a lot about Camp's attitude toward this building. Numerous examples of eighteenth-century drainage features show a deliberate effort on Anderson's part to control the flow of water through his shops toward the ravine. Without it, the buildings and their contents would have suffered from water damage. The fact that Camp was not insuring these structures, as well as the lack of evidence that she was attempting to control the flow of water, suggests that these buildings were abandoned and in disrepair.

Domestic Activity

Other evidence for nineteenth-century activities on the property is primarily in the form of features which date to the mid-nineteenth century and later. Two in particular were located in the area of Mrs. Ryland's Garden (Figure 41). The first (10GE-157, -158), 49 Figure 39. April 1823 Mutual Assurance policy. Top, image of original; bottom, transcription. Figure 40. Transcription of May 1830 Mutual Assurance policy. 50 Figure 41. Drainage gully, amorphous features in Mrs. Ryland's Garden. which was located in the northeast corner of the site (1097N/1099E), was an L-shaped feature in which both legs continued under the sidewalls of the unit. In profile, it was a flat-based U measuring 5-11 cm deep. The soil was a very dark grayish brown sandy loam (10YR3/2) with a high number of inclusions. Artifacts included nineteenth-century domestic trash and Anderson-related materials such as tinned iron, 79 grams of clinker and 15 nails, 13 of which were wrought.

Unfortunately, only a part of this feature was exposed and there was a significant amount of root disturbance in this area making it difficult to interpret. Due to the nails and the shape of the feature, it was originally thought to be an outbuilding. The high number of wrought nails in particular suggests that it might even be an eighteenth- 51 century feature, but, due to the presence of nineteenth-century artifacts in the fill, it has been interpreted to the latter time period and given a TPQ of 1850. Presumably, if this represented an outbuilding, there would have been a difference in both soil and artifacts on the outside and inside of the structure. However, the feature cut through a layer, Yellow Sandy Loam, which was found on both sides of the fill. Another possible interpretation of this feature is that it was related to gardening due to the soil's dark, loamy quality. Further archaeological investigation would be required to determine its function.

Approximately 200 cm south (1095N/1099E, 1094N/1099E) is the location of a second feature (10GE-200, -201) found in Mrs. Ryland's garden. The feature was linear on the north side, appearing to have been deliberately cut into the ground. The south side was rounded, and had the characteristics of spill, but never became defined. The shape of the feature tapered to form a flat, rather than a pointed end in the west. The soil was a dark grayish brown sandy loam (10YR3/2) measuring between 15-20 cm deep with inclusions some of which were shell and sand mortar. Artifacts included nineteenth-century domestic refuse and Anderson-related materials such as sandstone, shale, slate and clinker, but maintained a TPQ of 1850. This feature was associated with considerable root disturbance, but the presence of sand mortar and nine wire nails strongly suggests a nineteenth- rather than an eighteenth-century date. This feature could represent a shallow trash pit or a garden feature, but further archaeological investigation would be required to determine its function.

The third nineteenth-century feature identified (10GE-105, -199) was a long linear feature running east to west located in the northern half of the site. It measured approximately 300 cm in length and 11-22 cm in depth. In profile, it was a flat-based U. The eastern extent of the feature (1096N/1097E) ended abruptly and was not associated with any other features. At the western end (1096N/1095E) there is some evidence that the feature may have turned south. However it was cut by a Jimmy Knight trench in this area, making it impossible to determine. It did not show up on the other side of the Jimmy Knight trench, so it is possible to say that it did not continue much further west than what was exposed. The exact width of this feature is also unknown as it was cut in the north by the Noël Hume excavation. Noël Hume made no notes, reference or maps regarding this feature, and therefore it is not known whether part of it was excavated in 1975 or not. The soil was a very dark grayish brown sandy loam (10YR3/2) with brownish yellow clay inclusions (10YR6/8) containing bone, 99.5 g of clinker, coal, and brick. There were a significant number of artifacts primarily ceramics, glass and unidentified iron objects. Architectural items of interest include marble, a shutter dog, and 49 nails many of which were identified as cut or wire. Other items of interest are two tobacco pipe stems with the maker's marks "?/W," etched table glass, a fragment of colonoware/Native American ceramic, one worked bone, a bale seal, and a barrel hoop (Figures 42 and 43). The TPQ for this feature is 1864, making it a late nineteenth-century feature associated with post-Camp occupation.

Fence Line

Previous archaeology has determined the presence of a fence line along the western boundary of Lot 18. This fence line predates the Anderson occupation and was maintained 52 Figure 42. Tobacco pipe with maker's mark (Artifact No. 10GE-00105AV). Figure 43. Colono ware or Native American ceramic (Artifact No. 10GE-00105BK). into the nineteenth-century. Two square post holes (10GE-216, ?211, -212; 10GE-223, -224), whose centers were approximately eight feet apart, were located along the western edge of the excavation unit (1093-4N/1094E, 1095N/1094E), situated immediately outside of the building being investigated. Both postholes have TPQs dating to the late eighteenth century (1775 and 1787, respectively) suggesting that they were erected during Anderson's occupation of the property. The northern posthole (1095N/1094E) was cut by another posthole whose postmold had a TPQ of 1805. Further evidence supporting this posthole as a nineteenth-century repair is seen when comparing the amount of clinker and tinned-iron present in the two features. The early posthole contained 253 grams of clinker and one piece of tinned iron versus 883 grams of clinker and 85 pieces of tinned iron found in the repair. These higher levels of industrial waste argue for the presence of a higher amount of accumulated industrial waste associated with the passage of time. Therefore, it is assumed that Camp made at least one fence line repair during her tenure on the property.

Gardening Activities

The majority of twentieth-century activities identified during excavation relate to activities undertaken by the Colonial Williamsburg Foundation (Figure 44). The top three layers-Topsoil, Loam with Shell and Brick, and Brown Sandy Loam with Clay Mottling-had a mixture of material culture related to Anderson-period deposits up through the nineteenth century with twentieth-century inclusions. Because of the inclusion of Anderson-related materials and the lack of a significant number of nineteenth-century deposits, it can be inferred that heavy landscaping activities, primarily undertaken by Colonial Williamsburg, have had adverse effects upon the landscape.

Figure 44. Twentieth-century features.

Mrs. Ryland's Flower Garden, located in the easternmost part of the excavation area, has been maintained by Colonial Williamsburg since they purchased the property. Although only a small portion of this garden was investigated, the plants within the garden had major impact on the entire excavation area. The root systems for the trees in the garden, including an old pecan tree and dogwoods, produced significant intrusions into intact eighteenth- and nineteenth-century layers and features. Additionally, the planting of numerous flower bulbs created a very deep organic layer of topsoil that sealed the site in this area.

Colonial Williamsburg's Restoration Efforts

The majority of activities that impacted the excavation site were actions that Colonial Williamsburg undertook to reconstruct the Anderson property based upon the Frenchman's Map (circa 1782). In 1941, the architectural restoration team, under the leadership of James Knight, cross-trenched this area in order to find the foundations of buildings and re-expose foundations that had been identified in 1935 (Knight 1943) (Figure 45). Two diagonal trenches were identified in the southwestern portion of the site where the restoration team found the corner of a building dating to the eighteenth century. However, additional diagonal trenches, documented in Jimmy Knight's photographs, were not able to be identified. In an effort to establish the dimensions of the building, they then trenched along the projected paths of the walls uncovering the west wall's robber's trench, the southwest corner with brick in course and the limited remains of the south wall. Once information about the building was recorded, the trenches were backfilled with the same soil mixing stratigraphic layers and artifacts related to different time periods together. Many of the artifacts recovered from these fills appear to be related to Anderson activities. Of particular interest were numerous tobacco pipes (one with a maker's mark), pharmaceutical bottles, lead alloy objects, nail rod, 54 Figure 45. Cross-trenching in excavation area. 55 a stirrup, a piece of flint and various late eighteenth-century ceramic and glass fragments.

In 1974-75 the Department of Archaeological Research, under the direction of Ivor Noël Hume, used modern archaeological techniques to gather information regarding the industrial aspect of the Anderson property (Foss 1977). This project was undertaken in preparation for Colonial Williamsburg's reconstruction of Anderson's work areas identified on the Frenchman's map. The southernmost excavation area from this dig (ER 1943 and ER1967) was identified in the northern part of the site. Surprisingly, a large amount of material culture was left in the soil, which like Jimmy Knight's backfill, was a conglomeration of different time periods and occupations. Fewer Anderson-related artifacts were included in this matrix in comparison to Jimmy Knight's activities. In general, the artifacts tended to be small in size and varied in date from the eighteenth up to the twentieth century. Anderson-related artifacts of interest include late eighteenth-century ceramics, colonoware, lead and copper scrap, tinned-iron fragments, a bone button blanks (Figure 46), slag (evidence of metal smelting), French case bottle glass and a French gunflint.

The partially excavated remains of a linear trench (10GE-192-195), located in the southeast corner of the 1974-75 excavation, were identified in the profile of the cut for the Noël Hume unit. In 1975 it was labeled as a CW Trench on a site map, but was not given an ER number. No measurements or information about it were recorded and there was no explanation as to why this function was assigned to it. The extant remains of this trench (1096N/1097E; 1096N/1098E) were fully excavated this past summer.

Running on an east-west axis, the trench measured 2.18 meters (not including the portion excavated by Noël Hume) and was approximately 30 cm wide and 30 cm deep with the profile shape of a flat-based U. It consisted of three different fills, the first of which was a very dark grayish brown (2.5Y3/2) sandy loam speckled with clay. Approximately 20 cm deep, it contained a large amount of brick, coal and charcoal inclusions but few artifacts. The second fill was a light olive brown (2.5Y5/4) silt Figure 46. Bone button blanks (Catalog Nos. 37-10GE-00049BU and 4-10GE-00006BN). 56 loam with few artifacts and inclusions. Concentrated along the north wall of the feature, it measured approximately 66 cm in length tapering out in the eastern end. It was approximately 15 cm wide and 6 cm deep. Underlying this was the third fill consisting of a very dark grey (10YR3/1) sandy loam with yellowish brown (10YR5/6) clay inclusions. Concentrated along the southern wall of the feature, it measured 18-30 cm in width and was approximately seven cm deep. It contained coal, charcoal and brick inclusions as well as a few artifacts.

In general, the artifacts were primarily eighteenth-century ceramics, wine and window glass, wrought nails and faunal remains. This feature cut the long linear feature (10GE-105, -199), previously discussed under "Nineteenth-Century Activity," running along the southern edge of the Noël Hume excavation. Due to the 1864 TPQ of the feature it cuts, the feature in question must also postdate 1864. For the sake of continuity, this feature will be called a CW trench the purpose of which has yet to be unidentified.

In 1986, the D.A.R. investigated the southwestern edge of the property located on the west side of Mrs. Ryland's Flower Garden. Evidence of this excavation was located along the southern edge of this year's excavation unit (10GE-M6, -M7). Again, redeposited artifacts were present in the soil matrix of undefined context. Few can be specifically attributed to the Anderson period, but there is evidence of industrial activity such as late eighteenth-century ceramics, clinker, pharmaceutical bottles, and tinned-iron fragments.

Purpose:

During the excavations conducted at the James Anderson Blacksmith Shop in 2000-01, a number of pieces of iron sheet metal with what appeared to be either bare metal or a secondary white metal on at least one surface were excavated. Discussions with the Historic Area blacksmiths suggested that this material might be associated with tinning activities carried out at the site. Colonial Williamsburg's Archaeological Conservation lab was asked to help develop a plan for testing the material in order to determine whether the secondary metal was in fact tin. The conservation lab was also asked to test pieces of white metal found in related contexts to determine whether they also contained tin and might in fact be remnants of the high tin/lead solder used to adhere the seams of tinned vessels.

In conjunction with the project archaeologist, thirty-four samples were selected from contexts thought to be associated with the activities. Due to the rather large number of samples selected and the cost of elemental analysis such as X-ray fluorescence, it was decided that the initial material characterization of the material would be conducted using spot tests. Spot tests are sensitive and selective tests based on chemical reactions between the artifact and a reagent or test solution in order to establish the composition, identity, or nature of naturally-occurring or artificially-manufactured substances. They are qualitative in nature and are rarely carried out on samples of completely unknown character-generally some previous knowledge of the sample and/or its context is known. In this case we suspected that tin might be present on either the iron sheet metal samples or as a component in the white metal alloys.

Methods:

There are several spot tests for the presence of tin. The "test for tin using cacotheline" (Odegaard et al. 2000: 88) was selected since it was felt it would be the easiest test to interpret when applied to partially corroded materials. In principle small amounts of the metal are dissolved using hydrochloric acid and then react with a solution of cacotheline to form a purple color in the presence of tin. In practice one drop of 6 Molar hydrochloric acid solution is placed on the object for ten minutes. It is then wicked away using a piece of filter paper. A drop of cacotheline solution is then placed on the filter paper. In the presence of tin the filter paper will turn purple; in the absence of tin it will remain a yellowish color.

Results:

Emily Williams, Conservator of Archaeological Materials, and Julie Eklund, Marshall Steel Sr., Postgraduate Intern in Conservation, carried out the tests in April 2001. As well as the 34 samples selected from the Anderson excavations, two test blanks, one of modern tin and the other of copper, were also tested. As expected, the modern tin sample tested positive for tin, and the modern copper sample tested negative for tin.

The results are listed in Table 1. In all 24 pieces of iron sheet metal were treated. A positive result was obtained for fourteen of the samples, indicating that the surfaces of the iron in fact had been tinned. On one sample the first test resulted in a very faint color change that was not sufficiently similar to the results obtained from either the modern test pieces or the other archaeological samples to instill confidence. A second test was performed on the piece and a negative result was obtained. Since spot tests typically only test a very small portion of a sample, it may be that the second test completely missed any areas of surface tinning or that the tin residues on the artifact were so small that they were effectively used up during the first test. Eleven other samples of the iron sheet metal tested negative for tin. Iron sheet metal samples from the same contexts did not always yield the same results, producing both positives and negatives in several cases. None of the 10 samples of white metal tested positive for tin. Further cleaning of one sample revealed that it was a window lead.

Interpretation:

The tests confirm that there was tinned iron sheet metal on the site. The absence of any evidence of soldering, either on the samples themselves or in the associated contexts, may indicate that the tinned iron sheet metal was being stored on the site rather than being actively worked on. It may also indicate a bias in the sampling and testing strategy for the white metals. In order to confirm this result, ideally all the non-artifactual white metals (i.e., those that could not be readily identified as belonging to buttons, weights, bale seals, etc) from the contexts that have produced positive results for tinned iron sheet metal should be tested.