FORT STANWIX Casemates and Cannonballs Archeological Investigations at Fort Stanwix National Monument

In this section we will detail the 18th-century archeological features uncovered at Fort Stanwix (fig. 10). The major structural elements of the fort are presented in glossary form. Definitions of terms are taken from 18th-century sources, and the spelling of terms generally follows 18th-century usage. The principal elements of the fort are identified on the reconstruction drawings depicted as figures 8 and 9

Figure 8. Architectural reconstruction (three-dimensional projection view) of Fort Stanwix, view to east (after a drawing by Orville W. Carroll).

Figure 9. Architectural reconstruction (side elevation) of Fort Stanwix. These sections are based on archeological evidence, historic plans, contemporary descriptions and comparable forts (after drawings by Orville W. Carroll).

Figure 10. Plan of the 18th-century remains for Fort Stanwix located by excavation. See figure 9 for sections along the A-B and C-D lines. (click on image for a PDF version)

All structural measurements are in the English system as used in laying out the fort in 1758. All elevations are taken from Rome City Benchmark No. 128, 745 feet south-southwest of the fort. The parade ground averaged 451 feet above mean sea level with a variation of less than 6 inches between high and low readings. All depths are given in relation to a datum of 451 feet. All measurements of structures given are exterior dimensions unless otherwise noted. Approximate dimensions are those based on projections of existing walls and angles and, in some cases, on assumed building symmetry when the actual dimensions could not be measured. A measurement relative to the wall of a structure was taken from the exterior of the wall to the nearest point or side of the feature being described. See Appendix C for a list of features, their locations, and approximate dates of origin.

Wood samples were identified by the College of Environmental Science and Forestry, State University of New York, Syracuse.

Bakehouse

The Bakers who are Daily employ'd in Baking bread for this Garrison shall under no pretence Whatsover, take more than One Shilling, for a Loaf of Bread which shall weigh 6 lb. (M. Willett, 12/4/77)

The bakehouse for Fort Stanwix was located inside the southeast bastion (fig. 11). Although we are not certain when the bakehouse was built, one was certainly used during the Revolution (M. Willett, 1777; Lauber, 1932, pp. 541, 572, 577). We know also that it was constructed after the 1758 powder magazine was abandoned since the oven and one corner of the bakehouse were built over the hole created by the collapsed roof of the magazine after that hole had been filled with soil and leveled off. The 1758 magazine was still shown on a 1764 plan (Crown Collection CXXI, 103), the last known plan from the British occupation. Since the Americans built new magazines in 1777 (W. Willett, 1831, p. 49), it had presumably collapsed by that time.

Figure 11. View of the bakehouse excavation, looking east-northeast. Note the oven to the right, and the 19th-century foundation intruding the north corner of the structure. The bakehouse measures 20.3 feet long along the near wall, and 18.4 feet across.

The original excavation of the bakehouse was undertaken in 1965 by J. Duncan Campbell (1965). His report was available but we decided to reexcavate the structure to relate it to our plans, and to find its relationship to the 1758 powder magazine. The foot-wide sill beams were charred and the corners of the bakehouse were sufficiently destroyed so that it was not possible to determine the method of joining, except that no nails were used. The east wall sill beam was made from eastern white pine (Pinus strobus).

The building was not quite square, being approximately 21 feet long on the east side, 20.3 feet long on the west, approximately 19.5 feet long on the north, and 18.4 feet long on the south side. The northeast corner was destroyed by a post-1840 stone-lined cellar. Campbell reported finding clapboards 1.1 feet wide and 1 inch thick along the east wall which he thought were an exterior facing of the building. These should have required a heavy timber framing. We recovered no evidence of the roof, but it was probably sloped for drainage and made from heavy timbers like those found on the bombproofs under the other bastions.

Campbell reported the finding of window glass. However, a window would not have been in the walls since this structure was buried in the southeast bastion, and it is highly unlikely that there would have been a skylight. There may have been a window in the door.

The floor consisted of hard-packed earth. In several areas, especially the three remaining corners of the building, the earth was fire reddened and charcoal was abundant. This probably resulted from the fire that destroyed the structure rather than a deliberate burning of the floor to harden it. The floor was at an elevation of 450.05 feet or 1 foot below the parade ground level.

Entrance to the bakehouse was made down a flight of steps constructed of hewn logs resting on sterile soil (fig. 12). Except where burned, the wood had decomposed. The treads were approximately 1 foot deep, the risers were 5 inches, 10 inches and 3 inches high respectively from bottom to top, although the top step, or landing, was truncated by modern landscaping activities and may once have been higher. The steps were 5 feet wide while Campbell reported the doorway to be 3.75 feet wide flanked by vertical jambs on the bottom step which was actually the sill beam for the wall. One pintle for the door was located in the east jamb; another was lying on the floor. These had been removed prior to our excavation and are presently in the collection of the Rome Historical Society. The steps were 6.7 feet from the west wall and approximately 7.8 feet from the east wall.

Figure 12. View of the charred steps of the bakehouse, looking west. The bottom step is five feet wide.

A brick fireplace was situated opposite the doorway, fully recessed in the south wall 5.7 feet from the west wall and 6.3 feet wide (fig. 13). It was 2 feet deep with a back wall and jambs .6 foot thick. The back narrowed to 5.5 feet in width. The hearth floor was paved with bricks laid parallel to the back wall of the fireplace and extended 1.5 feet out in front of it. The fireplace stood 1.6 feet high but the top had been removed. There was a rodent burrow or intrusive pit (Feature 16) through the hearth and part of the back wall.

Figure 13. View of the fireplace and oven of the bakehouse, looking south-southeast. Slumping of the back wall was caused by a rodent burrow (Feature 16), roots or previous excavation. The rear wall of the fireplace measures 5.5 feet wide.

Behind the fireplace at an elevation of 451.80 feet stood an egg-shaped, arched oven approximately 12 feet long and 10 feet wide. Parts of this were destroyed by a tree and a gas line. Campbell reported a brick paving under the oven which we were not able to find, although we did observe a wood and mortar base under the walls of the oven that probably extended beneath such a floor. The gap in the oven walls to the hearth was 2.3 feet in width, but the actual opening into the hearth had been destroyed and may have been somewhat narrower. The oven was probably heated by wood embers, the coals scraped out into the fireplace and the bread inserted, after which the opening was sealed. The oven could have accommodated about 20 6-pound loaves at one time. Because the oven was underground and well insulated, several batches could have been baked before the oven had to be reheated. At its peak, the garrison would have required only 200 loaves a day at full rations which was probably rare. This was well within the capacity of this oven and most of the time they probably did not require over 60 loaves a day.

Campbell's excavations removed all in situ artifacts and building debris. Level I was his backfill and extended to the floor level. The few artifacts we recovered in this fill that compares to our documented, 18th-century specimens may have come from elsewhere in his excavations.

Banquette

. . . a kind of step made on the rampart of a work near the parapet, for the troops to stand upon, in order to fire over the parapet: it is generally 3 feet high when double, and 1-1/2 when single and about 3 feet broad, and 4-1/2 feet lower than the parapet. (G. Smith, 1779, p. 99)

The only evidence of a banquette at Fort Stanwix was found in the redout of the sally port. Here, we uncovered the base of an earthen banquette, probably revetted with sod, at an elevation of 443 feet. It was approximately 3.5 feet broad at the base. A 1764 plan of the fort (fig. 5) showed an overall width of 3.5 feet with a vertical height of 1.5 feet. The top of the banquette was shown as 2 feet wide with .5 foot taken up by a row of pickets. This produced a 45-degree slope to the face. The banquette on the rampart, which was probably wooden, was gone and no evidence of a banquette was found on the covered way.

Barracks

. . . are places erected for both officers and men to lodge in; they are built different ways, according to their different situations. When there is sufficient room to make a large square, surrounded with buildings, they are very convenient . . . (G. Smith, 1779, p. 20)

By November of 1764 two barracks were erected inside the fort (Crown Collection CXXI, 103). These, or their replacements, burned down in the fall of 1774 (Duncan, 1969). Americans occupied the fort from 1776 to 1777 and rebuilt the barracks. The west barracks were partly dismantled in 1780 when the guardhouse burned down (Van Dyck, 1780) and then rebuilt. All the barracks in the fort were destroyed in a general fire in May 1781 (Lauber, 1932, p. 581). Plans and contemporary powder horn engravings indicate that these were gabled roofed frame structures. (Carroll, 1973).

East Barracks

This structure measured 20 feet wide by at least 110 feet long, based upon the remaining timbers (figs. 14, 15). The southern end was disturbed, but assuming that it extended as far south as the west barracks, it was 120 feet in total length. The remnants of the foundations consisted of 1-foot-wide beams that lay on sterile soil. Two samples were identified as eastern white pine (Pinus strobus). They had collapsed from decay and were charred on their upper surfaces. The beam at the north end was 18 feet long with the side beams flanking the ends at the corners. At the northwest corner was a .5 foot wide beam, 3 feet from the corner, tying together the end and side sleepers. A soil change at the northeast corner indicated a similar structural detail. The only trace of a cross sleeper beneath the floor was approximately 16.5 feet from the south wall. This structure was 2 degrees out of line with the axis of the fort, the only structure so situated. A number of burned or corroded wrought iron nails were found on the sleepers and within the perimeter of the barracks.

Figure 14. Ground plans of the east and west barracks. See figure 15 for sections along the A-B and C-D lines. (click on image for a PDF version)

Figure 15. Sections through the east and west barracks. (click on image for a PDF version)

No fireplaces were found. This cannot be explained except by disturbance, since all 18th-century plans and sketches show chimneys on this building. Brick rubble found in cellar holes probably represents the remains of fireplaces. The floor was probably wooden planks running lengthwise to the building and nailed to sleepers resting on sterile soil. This inference was based on the presence of cellar holes, the absence of a packed dirt floor and sleepers in the west barracks.

Level I was 1 foot thick at the north end of the barracks and 2 feet thick over the middle and southern two-thirds. It contained post-1781 artifacts in a gray loam. Level II was the surface of the undisturbed sterile subsoil but also had post-1781 artifacts on it.

Four cellar holes were found under this barracks. The dates for these are discussed following a description of the west barracks. The artifacts listed for each cellar do not represent the total inventory for each, rather only those types which might give some indication of the age of the cellars. Feature 52 was a 6- by 9-foot hole, 2 feet from the west wall and 3 feet from the north wall. It was 5 feet deep and was probably wood-lined but no trace of the lining remained. This is inferred from the observation that the sand walls were nearly vertical. Intruding this was a 6- by 8-foot hole against the west wall and 5 feet from the north wall. This was 2.5 feet deep. The former hole was designated Level IV and the latter Level III. Intruding Level III was an oval pit, Feature 53, which contained, among other things, a modern soft drink bottle in a fill of gray loam and brickbats. The fill of Level III in Feature 52 consisted almost entirely of brickbats and mortar with some charcoal and almost no soil. The artifacts in this level included a 1734 British halfpenny, two Type 3A axe fragments, three Dutch gunflints, a mortar bomb fragment, two Type A-1a buttons, a Type B-1 button, a Type F-2 button, six Type 1 a porcelain potsherds, 11 Type 1b porcelain potsherds, six Type la stoneware potsherds, two Type 1b stoneware potsherds, a Type 1a earthenware potsherd, three Type 1b earthenware potsherds, four Type 2a earthenware potsherds and a Type 4 earthenware potsherd. Level IV was almost entirely sand with a few brickbats and some charcoal. As indicated, this was the original hole with Level III intruded into it. The artifacts included a Dutch gunflint, a mortar bomb fragment, a Type D-1 button, ten Type 1a porcelain potsherds, 21 Type 1b porcelain potsherds, a Type 1c porcelain potsherd, a Type 2 porcelain potsherd, eight Type 1a stoneware potsherds, two Type 1b stoneware potsherds, two Type 1b earthenware potsherds and a Type 4 earthenware potsherd. The Type 2 porcelain potsherd and the Type 1b stoneware potsherd cross-mended with fragments from Level II above and just south of Feature 52.

Feature 69 was a clay-lined cellar hole 41.5 feet from the north wall and against the east wall (fig. 16). It measured 15.5 feet north-south and 13.5 feet across and was slightly assymetrical. It was 3.5 feet deep and the clay lining was .6 to 2 feet thick. There were three levels. Level I was the fill adjacent to an intrusive stone and concrete foundation built between 1839 and 1851 (Waite, 1972, p. 61) (city map). Level II was a layer of brickbats, mortar and charcoal extending down to the floor. Artifacts found on the clay floor were designated Level III. Artifacts in Level II included two Type 4a axes, a Type 4c axe, a Type 3 spade, a Type 1 file, 1742 Dutch gunflints, 1002 French gunflints, a complete 4.5-inch mortar bomb, 56 mortar bomb fragments, a kettle fragment, three tin can fragments, two Type A-1a buttons, a Type B-2 button, a Type C-1 button, a Type D-1 button, a Type D-2 button, a Type E-1 button, a Type F-2 button, a Type H-1 button, seven Type 1b porcelain potsherds, a Type 1d porcelain potsherd, six Type 1a stoneware potsherds, a Type 1b stoneware potsherd, a Type 1d stoneware potsherd, two Type 1a earthenware potsherds, seven Type 1b earthenware potsherds, ten Type 2a earthenware potsherds and a Type 4 earthenware potsherd. Level III contained four Dutch gunflints, two Type A-la buttons, a Type A-1b button, a Type E-1 button, a Type F-2 button, a Type 1a porcelain potsherd, three Type 1b porcelain potsherds, four Type 1a stoneware potsherds, a Type 1b stoneware potsherd, a Type 1d stoneware potsherd, four Type 1a earthenware potsherds, five Type 1b earthenware potsherds, a Type 1d earthenware potsherd and four Type 2a earthenware potsherds.

Figure 16. View of a clay-lined cellar hole (Feature 69) in the east barracks, looking southeast. The concrete floor at left was broken from an intrusive garage foundation. Wide of the cellar is 15.5 feet.

Feature 72 was another clay-lined cellar hole 65 feet from the north wall and 1.5 feet from the east wall. It measured 15 feet north-south, 14 feet across and 4 feet deep. The clay lining had largely eroded away or been destroyed by modern intrusions, including the ca. 1839-1851 foundation noted under Feature 69 and an oil storage tank. Level I was largely disturbed by intrusions and contained brickbats, charcoal, mortar and loam. Level II lacked the loam and did not appear to have been disturbed by post-1781 intrusions. Level III was confined to artifacts on the clay floor of the cellar. Artifacts in Level II included a French(?) jetton dated 1700, a 1752 British half penny, five Dutch gunflints, two French gunflints, a 4.5-inch mortar bomb, a mortar bomb fragment, a Type A-1a button, a Type A-4 button, two Type B-1 buttons, a Type D-1 button, five Type 1a porcelain potsherds, three Type 1b porcelain potsherds, ten Type 1a stoneware potsherds, three Type 1b stoneware potsherds, four Type 1a earthenware potsherds, four Type 1b earthenware potsherds, a Type 2a earthenware potsherd, a Type 2e earthenware potsherd and a Type 4 earthenware potsherd. Level III contained three Dutch gunflints, a Type A-1a button, a Type B-1 button, eight Type 1a stoneware potsherds, two Type 1b stoneware potsherds and a Type 1b earthenware potsherd.

West Barracks

This structure measured 120 feet long and 20 feet wide (figs. 14, 15). Only fragments of the wall sleepers remained. Two samples were identified as eastern white pine (Pinus strobus). These were 1 foot wide, charred and resting on sterile soil. There were fragments of charred sleepers across the structure at distances of 5, 8, 10.5, 14, 19, 30, 33.5 and 36.5 feet from the south end. The best preserved of these butted against the east wall sleeper and were 1 foot wide. A plank floor running lengthwise to the structure was probably nailed to the sleepers.

A brick chimney base (without a hearth) was found in the south wall, 11.5 feet from the east wall. It was 3 feet wide and 1.5 feet thick. Only a couple of courses of badly deteriorated brick, bonded with a sandy gray lime mortar were found. A fragment of a mortar bed was found in Level II of Feature 56, just north of this brick.

Level I was a gray loam over the whole area containing post-1781 artifacts. There were six cellar holes beneath this barracks. Three were lined with 1-foot-wide vertical wood planks and three were lined with clay. The contents of these cellars, brick rubble, charcoal and burned artifacts, indicated deliberate filling after one or more fires (fig. 17).

Figure 17. Profile, looking north, of the rubble filling a clay-lined cellar (Feature 57) in the west barracks. Width of the cellar is 11 feet.

Feature 60 was a wood-lined cellar 1 foot from the north wall and 13 feet from the east wall. It measured 10 feet from north to south, approximately 7 feet across and 4.5 feet deep. Level I was a sand and cobblestone layer 2 feet deep which contained 19th-century ceramics. Below this was a layer of brickbats, mortar, a lime deposit and charcoal (Level II). In the lime deposit were 2,145 wrought iron nails 1 to 3 inches long. Level III was those artifacts resting on the wood plank floor in what appeared to be burned straw. The 1-foot-wide plank flooring ran north-south with 6-inch-wide plank sleepers under it. The sides of the cellar were shored up with vertical planks 1 foot wide but the framing which must have held these upright was not found. The south end of this cellar was destroyed during demolition work on the site in 1970. There was a trench 3 feet wide along the west side that extended 4 feet beyond the north wall and terminated at a circular wood-lined hole 3 feet in diameter and 2.5 feet deep (not a buried barrel). This may have been an intrusive 19th-century feature, but this could not be determined stratigraphically in the field and there were no artifacts in it. Level II contained a tool kit consisting of five files, a punch, an iron bar and a nail wrapped in coarse cloth and tied with a string, two Dutch gunflints, eight Type B-1 buttons, a Type D-1 button, two Type E-1 buttons, a Type 1a porcelain potsherd, five Type 1b porcelain potsherds two Type 1d porcelain potsherds, three Type 1a stoneware potsherds, two Type 1b stoneware potsherds, four Type 1a earthenware potsherds, three Type 1b earthenware potsherds, eight Type 2a earthenware potsherds, a Type 2d earthenware potsherd and a Type 2e earthenware potsherd. Artifacts in Level III included a Type 5a axe, two inkwells, three Type B-1 buttons, two Type B-3 buttons, a Type 1c stoneware potsherd and a Type 1a earthenware potsherd.

Feature 3, a clay-lined cellar, lay approximately 46 feet from the north wall and 7 feet from the east wall. It was approximately 13 feet long north-south, 10 feet across, and 4 feet deep with a 1-foot-thick clay lining. It was badly disturbed by demolition in 1970 and by two pipelines. Level II was a disturbed layer of gray loam and brick rubble with post-1781 artifacts in it. Level III included the disturbed fill over the two 6-inch pipelines below the top of the clay lining. Level IV was largely brickbats, mortar and charcoal but also contained some post-1781 artifacts. Level V was confined to those pre-1781 artifacts on the remaining portions of the clay floor. An in situ fragment of a pine beam (Pinus strobus) was collected from the top of a remaining section of the clay lining in Level II. In the fill of Level II was a 1766 Spanish 2 reales coin, two French gunflints, a Type 1a porcelain potsherd, four Type 1b porcelain potsherds, 14 Type 1a stoneware potsherds, two Type 1a earthenware potsherds, two Type 1b earthenware potsherds, a Type 1d earthenware potsherd, three Type 2a earthenware potsherds and a Type 2e earthenware potsherd. Level IV contained two kettle fragments, a Dutch gunflint, two Type A-la buttons, a Type B-3 button and two Type D-3 buttons. Artifacts in Level V included three Dutch gunflints, a French gunflint, a Type 1a porcelain potsherd, a Type 1b porcelain potsherd, four Type 1a stoneware potsherds and two Type 1a earthenware potsherds.

Feature 63, a wood-lined cellar, lay 42 feet from the south wall and 1.5 feet from the east wall. It was 8 feet north-south, approximately 7 feet across and 4 feet deep. There were vertical planks 1 foot wide along the east side but preservation was poor on the other sides. Feature 64, a clay-lined cellar, had been dug through the west side of this feature, destroying a part of it. Level II contained brickbats, mortar and charcoal but had been disturbed by a 1910 foundation which made it impossible to see the outline of Feature 64 until we reached a depth of 6 inches below the foundation so that this level contained artifacts from both features as well as later material. Level III was undisturbed and contained brickbats mortar and charcoal. Level IV was restricted to those artifacts on the floor. Artifacts in Level III included a balance beam from a scale, a Dutch gunflint, two Type 1b stoneware potsherds, a Type 1a earthenware potsherd, two Type 1b earthenware potsherds and seven Type 2a earthenware potsherds. Level IV contained a Type 1a stoneware potsherd and a Type 2a earthenware potsherd.

Feature 64, a clay-lined cellar hole was 37 feet from the south wall and 9 feet from the east wall. It was 13 feet north-south, 10.5 feet across and 3.5 feet deep. As noted, it cut into the lining and fill of Feature 63 and, as a result of an intrusive 1910 wall, Level II was included with Level II of Feature 63. Level III contained brick bats, mortar and charcoal, and Level IV was restricted to the floor of the cellar. Level III contained a Type 3b file, six French gunflints, a Type 1a stoneware potsherd, a Type 1b stoneware pot sherd and 12 Type 2a earthenware potsherds. Level IV produced a Dutch gunflint, a French gun flint, a Type 1b stoneware potsherd and nine Type 2a earthenware potsherds.

Feature 57 was a clay-lined cellar hole 12.5 feet from the south wall and 7 feet from the east wall (fig. 17). It measured 14.5 feet north-south, 11 feet across and 4 feet deep. The clay lining was .7 to 2.0 feet thick but broken on the east side for a distance of 3.5 feet (entry?). The south end of this cellar intruded the north end of Feature 56. Level II contained brickbats, mortar and charcoal but had been disturbed by a 1910 foundation. Level III also contained brickbats, mortar and charcoal but was undisturbed. Level IV was restricted to artifacts on the floor. Artifacts in Level III included a George I halfpenny (date illegible), a 1722 British halfpenny, two Dutch gunflints, nine French gunflints, a Type A-la button, a Type B-1 button, two Type 1b porcelain potsherds, two Type 1d porcelain potsherds, six Type 1a stoneware potsherds, two Type 1a earthenware potsherds, five Type 1b earthenware potsherds and seven Type 2a earthenware potsherds. Level IV contained a kettle fragment, a brass pot lid, a Type 3 spade, three Dutch gunflints, a French gunflint, two Type A-1a buttons, two Type B-3 buttons, a Type D-1 button, four Type E-1 buttons, a Type F-3 button, a Type 1a porcelain potsherd, six Type 1b porcelain potsherds, three Type 1c porcelain potsherds, a Type 1b stoneware potsherd, three Type 1a earthenware potsherds, five Type 1b earthenware potsherds, three Type 2a earthenware potsherds, three Type 2d earthenware potsherds and a Type 2e earthenware potsherd.

Feature 56 was a wood-lined cellar hole 2 feet from the south wall and 10.5 feet from the east wall. It was 10.5 feet long, north-south, 7 feet wide and 5 feet deep. There was an earth bench 2 feet wide and 1.5 feet high along the east and north sides with two planks laid on this on the east side. The upper half of the northwest corner had been destroyed by the clay lining of Feature 57. The walls were lined with charred vertical planks, 1 foot wide, with vertical posts in the corners 6 inches in diameter. Level II contained brickbats, a fragment of a mortar bed, slate and burned earth. It was separated from Level III by a layer of wood planks ranging from .5 to 1.5 feet wide. The mortar bed probably came from a hearth associated with the chimney base found at the south end of the west barracks and the whole level represented material in a depression caused by settling of the lower fill. Level III was a layer of brickbats, mortar and charcoal down to the level of the bench. Level IV was largely sand and gravel with some brickbats to a depth of 1.5 feet below the bench surface. If it were not for the artifacts and a slightly darker color this level would have looked like the sterile soil underlying the fort. Level II contained a 1768 Spanish 1 real coin, a Type 4 spade, two Dutch gunflints, a Type B-1 button, a Type D-6 button, a Type G-1b button, a Type I-1 button, a Type 1a porcelain potsherd, eight Type 1b porcelain potsherds, ten Type 1a stoneware potsherds, two Type 1b stoneware potsherds, 17 Type 1a earthenware potsherds, 11 Type 1b earthenware potsherds, and four Type 2a earthenware potsherds. The Type G-1b button is evidence of a post-1776 date for this level. Level III included two kettle fragments, two wooden buckets, three sacks of wheat and oats, a Type 4a axe, two Type E-1 buttons, a Type 1a porcelain potsherd, a Type 1b porcelain potsherd, nine Type 1a stoneware potsherds, three Type 1b stoneware potsherds, two Type 1d stoneware potsherds, ten Type 1a earthenware potsherds, eight Type 1b earthenware potsherds and a Type 2a earthenware potsherd. Level IV contained a glass beaker, a Type 1a stoneware potsherd and a Type 1a earthenware potsherd.

We know that the barracks burned twice, in 1774 (Duncan, 1969) and in May 1781 (Lauber, 1932, p. 581). Either of these fires could have produced the rubble in the fill of these cellars. The problem to be resolved is whether all the cellars were filled after one of these fires, and if so which fire, or whether some were filled after one fire and some after the other. Stratigraphically, Feature 57 intruded Feature 56, destroying part of the latter's wood lining and Feature 64 intruded Feature 63 destroying the west side of Feature 63. It is impossible, therefore, for Features 56 and 63 to be contemporaneous with Features 57 and 64 and they must be earlier in time. It is logical to assume that Features 56 and 63 were filled ca. 1774 and Features 57 and 64 filled ca. 1781. As the former pair were wood-lined and the latter clay-lined, we sought evidence from the fill of these cellars which would corroborate these dates and allow us to state that all the wood-lined cellars were filled ca. 1774 and all the clay-lined cellars ca. 1781. To do this we resorted to a null hypothesis that there was no temporal difference between the cellars. If this were true they should have the same types of artifacts in the same proportions. No coins were found dating after 1774 in these cellars but there were two post-1776 American buttons, one in Level II over Feature 56 and the other in Level III of Feature 69. Gunflints, buttons, ceramics and pipestems were deemed the most useful artifacts for comparison because work has been done on providing dates for the various types. Lyle Stone (1972, p. 47) has suggested that French gunflints were introduced after 1740. Table 8 illustrates that the frequency of French gunflints relative to Dutch gunflints increased with time. No French gunflints were found in the wood-lined cellars but half of the 44 specimens in the clay-lined cellars (excluding the concentration of 3,412 in Level II of Feature 69) were French. This high a percentage (if we use the Valley Forge sample for comparison) should indicate a post-1775 date. Unfortunately, the sample for the clay-lined cellar holes is quite small and not consistent except that all have French gunflints in them.

The buttons are less instructive. All the types found in the wood-lined cellars were also found in clay-lined cellars although nine types, including one Revolutionary War button, were found only in clay-lined cellars, primarily Feature 69. Type A-1 buttons were more prevalent in clay-lined cellars (11 in four as opposed to two in Feature 52) but, again, the sample size is small. The high number of Type B-1 buttons in Feature 60 could have originated from a single garment although they were not found in a cluster.

The ceramics were also scarce. We applied Stanley South's mean ceramic date formula (South, 1972) to what we had with a singular lack of success, even after eliminating the Oriental porcelain (Type 1 porcelain) and Delftware (Type 1 earthenware), because of the small samples. Lumping together the cellars to increase the sample size produced mean dates of 1770.8 ± 13.7 years for the wood-lined cellars (67 sherds) and 1774.1 ± 14.3 years for the clay-lined cellars (124 sherds). This indicated that the clay-lined cellars were filled later than the wood-lined cellars but that the fill contained material nearly as early in time as the latter (1759.8 and 1757.1 respectively).

The pipestem bore diameters were no help because the sample sizes were too small. Lumping the samples together and applying the Harrington principle (Harrington, 1954) indicates that the clay-lined cellars might be somewhat earlier than the wood-lined cellars. However, the clay-lined cellar combined sample (36 specimens) is barely adequate, and the difference of the two means is only .05/64th of an inch, not statistically significant for the size of the sample.

We are left with only the stratigraphy, one button and logic for the statement that the wood-lined cellars were dug ca. 1764 and filled ca. 1774, while the clay-lined cellers were dug ca. 1776 and filled ca. 1781. The fill of the cellars does not vary sufficiently for us to separate them in time. This is probably due to (a) only a 7-year time interval between the two periods of filling in which artifacts did not change sufficiently to allow us to measure that change and (b) the nature of the fill itself. The compactness of the brickbats, etc. in the holes, the absence of brickbats beyond the limits of the holes, and the absence of lensing, indicate deliberate and rapid filling. If there was any pre-1776 debris left on the site after the fort was rebuilt it could have been shoveled into the clay-lined cellars along with any post-1776 artifacts. The finding of similar button and ceramic types in the two kinds of cellars supports the null hypothesis. That French gunflints and some button types occur only in clay-lined cellars could be accidental, and are not sufficient by themselves to disprove the hypothesis. As we cannot distinguish between the artifacts in the two kinds of cellar holes then we must conclude that the contents date from 1758 to 1781 since they could contain material which was on the parade ground before the barracks were built and the earliest cellars dug (1764).

Bastions

. . . a part of the inner inclosure of a fortification, making an angle towards the field, and consists of 2 faces, 2 flanks, and an opening towards the centre of the place, called the gorge: or it is rather a large mass of earth, usually faced with sods . . . (G. Smith, 1779. p. 99)

All sources agree that Fort Stanwix had four bastions located at the corners of the square fort. The archeological record was not too helpful; the information derived from the excavations had to be augmented with what was supposed to have been built. The northwest bastion was the most intact and yielded the most information. We were able to outline the right flank, part of the right face and part of the left flank. This enabled us to determine the following dimensions and angles at parade ground level: flank length, 35 feet; face length, 108 feet; reentry angle, 111 degrees; flank angle, 125 degrees; salient angle, 62 degrees. The angles and measurements were probably the same for all the bastions and checked out when we projected a plan of the fort (fig. 10).

In the original 1758 fort, there was a powder magazine in the southeast bastion (fig. 4), and the bastions are shown as earth-filled structures revetted with logs on the exterior. The magazine will be described under the section to follow entitled "Bombproofs."

When the Americans occupied the fort, they rebuilt the bastions and incorporated in them a bakehouse and three bombproofs. These will be described separately. The passage leading into these greatly constricted the gorge, or throat, of each bastion, thus hampering movement of troops or cannons.

Berm

. . . a little space, or path of 6 or 8 feet broad, between the ditch and talus of the parapet; it is to prevent the earth from rolling into the ditch, and serves likewise to pass and repass. As it is in some degree advantageous to the enemy, in getting footing, most of the modern engineers reject it. (G. Smith, 1779, p. 99)

The only evidence of a berm at Fort Stanwix came from the north and east curtains. Along the north curtain a beam .5 foot wide was found lying parallel to the scarp of the ditch, 5.5 feet from the edge. We believe this beam was part of the rampart, although much smaller in diameter than expected. It may have served as a brace to keep the rampart logs from slipping. On the east curtain were a pair of large posts flanking the sally port, 7.5 feet from the edge of the scarp. These two posts formed part of the revetment where the sally port cut through the curtain wall. Based upon this, the berm on the east side was 7.5 feet wide. A 1764 plan (fig. 5) shows a 6-foot berm around three sides and a 5-foot berm on the south side. The latter was probably constructed to increase the thickness of the south curtain wall. Another 1764 plan (Crown Collection CXXI, 103) shows pickets on the berm on the east side, but no archeological evidence for these was found.

Bombproofs

MAGAZINE, a place in which stores are kept, or arms, ammunition, provisions, &c. . . . Powder-Magazine . . . ought to be . . . bombproof. In fortifications they are frequently placed in the rampart . . . (G. Smith, 1779, p. 155)

Four bombproofs were located at Fort Stanwix, one inside each bastion. The one beneath the southeast bastion was built in 1758, but the others were probably in use during the Revolution. The southwest bombproof also served as a hospital and repository for valuables during the siege in 1777 (Reid, 1905, p. 99).

Northeast Bombproof

Although we have no written reference to this structure, the symmetry of the fort is completed with a bombproof under each bastion. The main chamber of the structure was approximately 16 feet square, although much of it was destroyed by tree roots and a cistern (figs. 18, 19). A single layer of thick roof planks, 1 to 1.5 feet wide, that collapsed after fire destroyed the structure, ran east-west. No evidence of a floor was found. Presumably, it was a dirt floor at an elevation of about 448 feet, 3 feet below the parade ground. Most of the floor was fill, apparently used to build out the hill on which Fort Stanwix stood, and this made it impossible to define since it did not differ from the surrounding soil.

Figure 18. Ground plan of the northeast bombproof. See figure 19 for section along the A-B line. (click on image for a PDF version)

Figure 19. Section through the northeast bombproof and passageway. (click on image for a PDF version)

The entrance to the bombproof was a passage, 6 feet wide on the west side with an interior width of 4.5 feet. The passage ran for 12.5 feet, then turned 35 degrees north and continued for another 12.5 feet. It increased in exterior width to 7 feet at the turn. The walls of the passage were revetted with horizontal planks .7 foot wide, held in place by randomly spaced vertical posts. In one area nine of these planks were crushed together giving the passageway a height of at least 6 feet. Trenches for two sleepers .8 foot high were spaced along the passage. Three layers of planks .6 to 1.2 feet wide and 1 inch thick were found on the floor of the passage. A few of these may have fallen from the walls but some of them were definitely floor planks under which lines of musket balls and canister shot were found where they had fallen through cracks. Level I consisted of a gray loam above the roof planks in the passageway and main chamber. It contained some post-1781 artifacts. Level II was a brown sandy loam beneath the roof planks and on the floor. It was undisturbed. One large deposit of 45 cannister shot found under the floor at the widest point in the passage may have fallen through a substantial gap along the wall. All the artifacts ascribed in the next chapter to this structure were found in Level II of the passageway except a Dutch gunflint, two French gunflints, a scabbard hook, both buckles, a .69 cal. musket ball, a Type 1a porcelain potsherd, three Type 1b porcelain potsherds, seven Type 1a stoneware potsherds, two Type 1b stoneware potsherds and two Type 1a earthenware potsherds. The dirt beneath the floor boards was at an elevation of 449.24 feet.

At the west end of the passageway two steps led to an elevation of 451.88 feet. The steps of hewn beams had .6-foot risers, .8-foot treads and were 4 feet wide. The top step, above the parade ground elevation, probably served as a curb to keep water out of the passageway. Two iron pintles were found on the steps and a door was probably at the top of the steps on the north side. The passage was roofed with planks .5 to 1.2 feet wide which lay across the passageway and were supported by the walls.

Northwest Bombproof

. . . the engineer had neglected to build a magazine, though he knew there was no secure place for the ammunition. The garrison, in order to remedy this difficulty, took the seven spare feet which were left of the pickets,. and having framed them, so as to form a square inclosure, the whole was placed within the body of one of the bastions, and being covered with earth, formed a safe deposit for the powder. (W. Willett, 1831, p. 49)

This trapezoidal structure met the description of the magazine constructed in 1777 (figs. 20, 21). The walls of the main chamber and the passageway were tightly spaced vertical logs .4 to 1.2 feet in diameter with .5 feet being the most common size. Seven logs were identified as eastern hemlock (Tsuga canadensis). There was some evidence for a frame to keep the log walls upright. Lengths of the walls in feet were: north, 16; south, 13; west, 21 and east, 21 (including the doorway). The roof was made of a single layer of .5-foot-thick planks running east-west, .6 to 1.5 feet wide. The floor had eight hewn .5-foot square sleepers running east-west, resting on sterile soil with another beam across the west wall and short round beams connecting the hewn sleepers down the midline. Nailed to these were ten 1-inch-thick planks 1.2 to 1.5 feet wide. Both the sleepers and floor boards were identified as eastern white pine (Pinus strobus). The floor was at an elevation of 449.84 feet while the sterile soil below was 449.22 feet. The highest point on the bombproof was 452.04 feet, but if we use Willett's account, the ceiling was probably 7 feet above the sterile soil at an elevation of 456 feet. The thickness of the roof would add .5 foot leaving 4.5 feet of dirt on top of the roof if the bastion were raised 10 feet above the parade ground. Although the interior was burned, the top of the roof was not, indicating that it was covered with earth.

Figure 20. Roof and floor plans of the northwest bombproof. See figure 21 for section along the A-B line. See figure 10 for excavation limits. (click on image for a PDF version)

Figure 21. Section through the northwest bombproof and passageway. (click on image for a PDF version)

The passageway ran to the east 12.5 feet and then made a 43 degree turn to the south, continuing for another 45 feet. It was 6 feet wide with an interior width of 4.5 feet, widening 1 foot at the turn. Vertical hemlock posts about .7 foot in diameter lined both sides of the passageway. Sleepers were located along the base of the interior of the walls with perpendicular sleepers irregularly spaced along the passageway. To these were nailed eastern white pine planks 1.5 feet wide and three abreast.

The roof was formed of planks .6 to 1.2 feet wide lying across the passageway and resting on the walls. The door was located at the outer end of the passageway where two large strap hinges and pintles were found (fig. 40c). It probably opened out toward the east.

Level I was a gray loam above the roofs of the bombproof and passageway. Level II was burned brown sandy loam between the partially collapsed roofs and the floor. This material washed in after the roofs collapsed judging by the lensing of the deposit and hollow spaces beneath some roof planks. Level III was confined to objects on the floor or beneath it. All of the material listed in the chapter on artifacts for this structure came from the passageway except the musket balls, a Type 1c strap hinge, two Type 2b pintles, two bayonet fragments, six iron spear tips, a Type 1a earthenware potsherd, a Type 1b earthenware potsherd and a Type 2a earthenware potsherd. The last three were in Level II and the remainder in Level III. All the musket balls listed in the chapter for the northwest bombproof came from the floor (Level III) of the main chamber just in side the entrance from the passageway.

Southwest Bombproof

. . . This Day the Col. Ordered . . . all the public papers and money in the Hands of Mr Hanson & the papers in the Hands of Mr Van Veghten belonging to the Paymaster to be lodged in the Bombproof in the S W Bastion . . . (Reid, 1905, p. 99).

This structure was definitely in use during the siege of Fort Stanwix.

The main chamber was nearly square, approximately 20 feet long east-west and 19.5 feet across (figs. 22, 23). Like all the other structures, this was destroyed by fire. The walls were greatly disturbed by modern intrusions; accurate measurements were not possible. The walls rested on narrow pine sleepers (Pinus strobus) laid on sterile soil. There was a single sleeper 9 feet long, 1 foot wide and .5 foot thick, below the floor. This was located 9.5 feet west of the entrance, running north-south. The floor was composed of pine planks resting on sterile soil. These were 1.1 to 1.5 feet thick, but may have been thicker before burning. They ran east-west except along the west wall where there were four planks running north-south, some of which lay under the ends of east-west planks. These were laid loose and not nailed down. The floor was at an elevation of 451.20 feet. Only a few fragments of the roof were found. These measured 1.1 to 1.3 feet wide and .4 to .85 feet thick. The planks appeared to run north-south and probably rested on the walls. There was a 4.7-foot wide doorway in the east wall, 7.2 feet from the northeast corner. Vertical door jambs, .8 foot square, were morticed into a threshold .8 foot wide. A passageway 6.5 feet wide with an interior width of 5.5 feet extended east 5 feet where it made a 60 degree turn to the north, then continued for at least 17 feet more before it was obliterated by intrusions. Beyond the turn it narrowed to 6 feet in exterior width. The walls were not found, but they were probably horizontal planks held up by framing since there were several planks in the fill which could not be related to the floor. The elevation of the floor of the passage was 450.77 feet above mean sea level. Three sleepers, 1 foot wide, were found across the passage at intervals of 4.5 feet from the door threshold and 5.5 feet and 7 feet from each other. An undetermined number of planks were laid lengthwise over these for a floor.

Figure 22. Plan of the southwest bombproof. See figure 23 for section along the A-B line. (click on image for a PDF version)

Figure 23. Section through the southwest bombproof. (click on image for a PDF version)

Level I was disturbed fill over the bombproof including trenches cut through the floor of the bombproof. Level II was undisturbed charcoal (after 1781) and burned earth fill over the floor. Level III was those artifacts directly on the plank floor and in the sleeper trench while Level IV was those artifacts beneath the floor. All of the artifacts listed in the next chapter for the southwest bombproof were in the main chamber except for a French gunflint and two 3-pounder cannonballs. Level II contained two 3-pounder cannon balls, a mortar bomb fragment, a cannister shot, a Type 1b buckle fragment, a door latch, a door pull, a steel axe bit, six straight pins, three round-cornered staples, a Type 1b porcelain potsherd, four Type 1a stoneware potsherds, two Type 1a earthenware potsherds, three Type 1b earthenware potsherds, two Type 2a earthenware potsherds, two Type 2a earthenware potsherds and a Type 5c earthenware potsherd. Level III contained a Dutch gunflint, a screw, a 3-pounder cannonball, two musketballs and three miscast musketballs. Level IV contained three cannister shot, 23 musket balls, three miscast musketballs, lead waste, a round-cornered staple and a Type E-1 button.

1758 Magazine

This magazine was shown in the southeast bastion on all the British plans up to 1764 but apparently had collapsed by the time of the Revolution since it was never referred to in documents relating to the rebuilding of the fort. The bakehouse was built over the fill of the hole in which the magazine was placed. Because of intrusions we were able to locate only a part of the structure. Finding no artifacts in the part we excavated to the floor level, we did not complete the excavation because of its low priority. The magazine measured 16 feet in width—narrower than shown on the 1758 elevation (fig. 4). Only 25 feet of the original length was uncovered, the ends having been destroyed by an 1842 cellar (Waite, 1972, p. 40) and a street. The floor was at an elevation of 442.76 feet, with the highest remaining wood from the roof at an elevation of 448.63 feet. Two samples of this roof wood were identified as eastern white pine (Pinus strobus). Fill below the roof shows lensing indicating abandonment and gradual decay.

Bridge

. . . the outside Gate and the Draw Bridge are to be shut at Retreat Beating . . . (Lauber, 1932, p. 541).

Information on this drawbridge was scanty, except that it was built ca. 1764 (Crown Collection Map CXXI, 103). The method of raising and lowering the bridge is unknown and our excavations did not shed much light on the subject. A large pintle and a matching hinge (fig. 41) which showed much wear, were found in the southeast casemate (Level II) and at the north end of the bridge (Level XIII) respectively. Along with the footings, these artifacts are interpreted as the only solid evidence for a drawbridge.

Level I was a gray loam layer about 2 feet thick over the entire area. Levels II-X were 19th-century strata above the bridge (judging by the presence of post-1781 ceramics). Level XI encompassed the bridge decking at the north end of the bridge. Two dark gray clay lenses at the south end were labeled Levels XII and XIV. These appeared to be buried sod layers washed down the counterscarp and separated by graded sterile sand and gravel. At the north end Level XIII was a burned sandy clay beneath the bridge decking. Level XV was two layers of clay separated by sterile wash similar to that found at the south end of the bridge. Artifacts from Levels XI-XV were all secondary deposits and dated ca. 1764-1781. At the north end of the bridge, the rampart had a 12-foot indentation, approximately 13 feet wide. The scarp below this had an angle of 45 degrees from vertical. In the bottom of the ditch at a distance of 12 feet from the berm stood two large vertical pine posts (Pinus strobus) 9 feet apart which ended slightly above the floor of the ditch. Although too badly decayed to measure accurately, these appeared to be at least 1.5 feet square. Fifty-two feet south of these were another pair of posts .5 feet in diameter and 9.5 feet apart set in the counterscarp which had an angle of 37 degrees from horizontal. Allowing for a 4-foot overlap on the counterscarp, the bridge was 74 feet long. On the floor of the ditch (Level XIII) were two .8-foot-square white cedar sleepers (Thuja occidentalis) 8 feet apart. The best preserved specimen was over 39 feet long and exhibited one square mortice hole 28.6 feet from the north end which ended against one of the large vertical posts at the base of the scarp. The space between the sleepers was filled with sand to an elevation of 442.18 feet. A few cross planks were found under the sleeper, probably to shim it up.

Little of the superstructure was found. At the north end, extending 5 to 19 feet from the north end of the bridge, were planks at a 90-degree angle to the axis of the bridge (Level XI) which may have been decking. They ranged in width from .7 to 1 foot wide; none were more than 7.5 feet long. Under these at a distance of 3 to 10 feet from the north end were the large hinge and 97 large spikes.

No details on how the drawbridge was raised were found but the indentation in the scarp could have served as a well for counterweights.

Casemates

. . . a work made under the rampart, like a cellar or cave with loopholes, to place guns in it. (Muller, 1746, p. 214).

Common usage of the term "casemate" on military plans in late 18th-century North America meant structures under the ramparts which offered shelter from artillery fire. They may have been used for troops' quarters as well. Loopholes were not feasible in earth-and-log forts. All of the casemates at Fort Stanwix were log structures built into the ramparts with the roofs serving as the terrepleine of the curtain walls. Casemates were a consistent feature of the fort and, although they may have been rebuilt several times, they retained the same general shape and size on the plans.

North Casemate

This structure was approximately 144 feet long on the back wall, approximately 120 feet in length on the front wall with the west end wall 25 feet long (figs. 24, 25). The east end was not found and length was projected from the midline of the fort. The angle formed by the back and west end walls was 60 degrees, producing a total depth to the structure of 22 feet. The sill beams were 1 foot wide and rested on sterile soil. One of these specimens was identified as eastern white pine (Pinus strobus). Two possible doorway areas were uncovered in the form of cobblestone paved areas centered 10 feet and 90 feet from the southeast corner along the front wall. Two stone foundations, Features 15 and 19 intruded the north casemate along with a drain pipe, sump and water line associated with the former. Feature 15 was actually two foundations, one dating from ca. 1873 (city map) and the other from ca. 1878 (city map). Feature 19 dated from ca. 1878 and represented the moving of a small structure from the 1873 foundation. The east end of the casemate for about 30 feet was disturbed by gardening activities; plow scars and bush holes intruded the floor of the structure.

Figure 24. Plan of the north casemate. See figure 25 for section along the A-B line. See figure 10 for excavation limits. (click on image for a PDF version)

Figure 26. Section across the north casemate. (click on image for a PDF version)

There were six single-hearth fireplaces (features 10, 12, 13, 17, 30 and 42) along the rear wall which measured 7.5 feet wide except the westernmost (fig. 26) which was 7.8 feet wide. They were 3 feet deep, with 1-foot-thick jambs at an angle of about 85 degrees from the back wall. None of the fireplaces were aligned to the rear wall, indicating that they were not a part of the original structure, having been added after the wall was built. The jambs were one and one-half bricks thick and the rear wall one brick length thick in alternating courses of headers and stretchers. The hearths were all composed of sterile soil projecting only a short distance in front of the fireplace. Rubble from the collapsed chimneys extended to the front of the casemate but not beyond. Feature 10 contained 13 wrought iron nails, a .58 cal musket ball, a .69 cal. musket ball, a dark green bottle fragment, a blob of lead weighing 237 grams, a Type 1a porcelain potsherd and some calcined bone. Feature 12 contained 5 wrought iron nails, a round-cornered staple, a Type 3 knife fragment, a Type 1b porcelain potsherd, a Type 1b porcelain potsherd, a Type 1a earthenware potsherd and a Type 2a earthenware potsherd. Feature 13 contained three wrought iron nails, a Dutch gunflint, a .69 cal. musket ball, five case bottle fragments, a Type 1a stoneware potsherd, a Type 1b earthenware potsherd, and a Type 2a earthenware potsherd. Feature 17 contained three wrought iron nails, two Type 1a porcelain potsherds, a Type 1b stoneware potsherd, a Type 1a earthenware potsherd, a Type 1b earthenware potsherd and three Type 2a earthenware potsherds. This fireplace was badly disturbed, probably by a garden. Feature 30 contained a lead game counter, a Type 1a stoneware potsherd, a Type 2a earthenware potsherd, a Type 6b earthenware potsherd and a Type 6g earthenware potsherd. This fireplace was largely destroyed by an 1873 foundation, Feature 15. Feature 42 contained 12 wrought iron nails, a 1745 British halfpenny, two Type 1b shoe buckle fragments, a Type 1c shoe buckle fragment, a scabbard hook, a .69 cal. musket ball, four dark green bottle fragments, a case bottle fragment and a Type 1b stoneware potsherd. All of this material is included in the north casemate artifacts listed in the next chapter.

Figure 26. View of the westernmost fireplace (Feature 10) in the north casemate before excavation, looking north. Note the brick fall from the chimney. Width of the fireplace is 7.8 feet.

Between the fireplaces, along the rear wall, was a bench of sterile soil 4.5 feet wide at an average elevation of 451 feet. In front of this were occasional sleeper trenches at a 90-degree angle to the rear wall and about .5 foot deep. The sterile soil was quite uneven and many of the sleepers may have rested on top of it with no trench. At the west end, floor planks were found running parallel with the back wall. They were too badly decomposed to measure. Over the floor and running perpendicular to the rear wall were planks 1 to 2 feet wide which probably fell from the roof. Level I was a gray loam over the entire area with some post-1781 ceramics and other objects in it. Level II was a series of brickbat concentrations in front of each fireplace. Level III was a brown sandy loam under the brick rubble and over the floor of the casemate.

Starting near the midline of the casemate and reaching a maximum depth of 3.5 feet toward the west wall, where it stopped, was a trench below the floor level. This was designated Feature 34 and was filled with sand but no function for it could be determined except possibly drainage. It was definitely a pre-1781 feature. Two artificial Levels (III and IV) were designated, but no difference was noted between them except that most of the artifacts were in Level IV (more than 2 feet below the floor or 3 feet below datum). Feature 34 contained 42 wrought iron nails, a rivet, a Dutch gun flint, a French gunflint, a piece of buckshot, 11 mortar bomb fragments, a bone handle fragment, five Type 1b porcelain potsherds, a Type 1d porcelain potsherd, eight Type 1a stoneware potsherds, a Type 1b stoneware potsherd, a Type 1d stoneware potsherd, nine Type 1a earthenware potsherds, ten Type 1b earthenware potsherds, 64 Type 2a earthenware potsherds, a Type 2e earthenware potsherd, 14 dark green glass bottle fragments, 31 case bottle fragments, 24 green pharmaceutical bottle fragments, two clear table glass fragments and 288 pipestem fragments with a mean bore diameter of 4.05/64th of an inch.

The even spacing of the fireplaces suggests that the structure was divided into 20-foot-square rooms, probably for use as officers' quarters.

West Casemate

Only the north end of this structure was found (fig. 10). The middle section was destroyed by a swimming pool and utility lines while test trenches dug where the south end should have been failed to find any trace of the building. The footings for the rear and north end walls were 1-foot-wide eastern white pine beams (Pinus strobus) at an angle of 52 degrees, resting on sterile soil. The rear wall sill extended over the end wall, suggesting dovetailing of the joint. The front wall had been removed by a large modern pit with a concentration of late 19th-century bottles. Parallel to the rear wall, 9.5 feet in from it, was a sleeper 1 foot wide set in a trench 1 foot deep. The sterile soil beneath the floor was at an elevation of 451.84 feet. There was a 1-foot-deep trench 8.5 feet long and 2 feet wide in the northwest corner running parallel to the rear wall and 3 feet from it. Level I was a gray loam containing a large lens of cement such as one might find at a construction site, and probably dates from the building of the swimming pool ca. 1930 (Waite 1972, p. 30). There was also a concentration of post-1781 animal bones with saw marks typical of bones found in 19th-century privies on the site. There was some evidence of gardening in the form of bush holes. Level II was a brown sandy loam over the casemate floor and in the trenches. In the test trenches at the south end of this casemate Level I extended down to sterile soil confirming an account that this area had been scraped to level the fort for a house built in 1828 (Wager, 1896, p. 52).

The remains of a plank floor parallel to the rear wall were found in the back half of the casemate. On top of this, confined primarily to the front half, were planks at a 90-degree angle to the rear wall. These were 1 to 1.5 feet in width. On top of one of these was the remains of a brick pillar .8 foot square and at least 6 courses high (approximately 1.5 feet) which had fallen on its side. This may have been a footing for a roof support. It lay 10 feet from the back wall and 10 feet from the north end wall. No fireplaces were found, but there was a burned area in the floor beginning 25 feet from the northwest corner and 5 feet from the rear wall which extended into the area disturbed by the swimming pool. Projected from the midline, the approximate length of the rear wall was 144 feet; front wall, 117 feet; and the north end wall, 28 feet. The angle of the south end wall could not be determined.

Southwest Casemate

This building was one of the better preserved ones (figs. 27, 28) although the southwest corner was destroyed by a 1910 cellar (city directory). The front or north wall was 56 feet long marked by a trench (Feature 48) 1-foot wide and .8-foot deep. Some wood was found in the east end of the trench. The east wall was marked by a .6-foot-wide beam at a 90-degree angle to the front wall. Although most of the beam was missing, it probably measured 22.5 feet long. The rear wall beam was 1.5 feet wide and approximately 69 feet long. It was made of eastern white pine (Pinus strobus). The west wall beam was 1.5 feet wide and set at an angle of 60 degrees to the rear wall. It was approximately 25.5 feet long. These sill beams rested on sterile soil.

Figure 27. Plan of the southwest casemate. See figure 28 for section along the A-B line. (click on image for a PDF version)

Figure 28. Section across the southwest casemate. (click on image for a PDF version)

Level I was a gray loam over the area which included some post-1781 ceramics. Level II was a layer of ash, mortar, brickbats and brown sandy loam over the floor of the casemate. The floor (Level III) was composed of a mixture of hard-packed earth, ash and mortar. At the ends of the casemate and along the rear wall there seemed to be no prepared floor, only sterile soil. Running the length of the structure, 9.5 feet from the rear wall, was a wood sleeper 1 foot wide set in a trench 1 foot deep. This was interrupted in the center by the foundation of a fireplace (Feature 50). The fireplace was set on a yellow sand layer 8.5 feet long and 7.5 feet wide located 6 feet from the rear wall and 23.5 feet from the east wall. Only the mortar bed of the fireplace remained. It had back-to-back hearths that faced the ends of the building. The rubble from the chimney lay over the center of the structure. In the fill above the fireplace there was a cluster of four Type 4 spades. The roof consisted of 1 to 1.2-foot-wide planks at a 90-degree angle to the rear wall. The central sleeper may have served as a foundation for posts to support the roof.

Buried to a depth of .7 foot in the floor, 6 feet from the northwest corner, was a wood-lined box 6 feet long and 3 feet wide (Feature 51) (figs. 29, 30, 31). This was floored with three .9-foot-wide planks. The walls were vertical planks 1.1 to 1.5 feet wide. There were two rows of six staples each driven into the floor .7 foot from each end and spaced .4 foot apart. Their purpose is not known. The box was filled with brickbats to a height of 1.6 feet and contained 385 wrought iron nails, 24 brass tacks, a cannon primer, two 3-pounder cannonballs, a .56 cal. musketball, a .64 cal. musketball, four .69 cal. musketballs, a sad iron, coarse cloth, a Type F-2 button, a Type 2 bead, a Type 1e buckle, a Type 5a buckle, 9 brass straight pins, a brass pot, an axle hub, a Type 6 knife, a Type 6b axe, a claw hammer, a Type 1 ruler fragment, a protractor fragment and a plumb bob. We strongly suspect that it was a carpenter's tool box.

Figure 29. View of the brick rubble on top of Feature 51, a wooden box in the floor of the southwest casemate, looking south-southwest. See also figures 27, 30, and 31.

Figure 30. Plan of the box in the floor of the southwest casemate (Feature 51). See figure 31 for section along the A-B line.

Figure 31. Section through wooden box (Feature 51) in the southwest casemate.

Southeast Casemate

Most of this structure was destroyed prior to our excavations. In 1965 Campbell found a 1-foot-wide sill beam located where the east end wall should have been. We uncovered a 1-foot-wide white cedar beam (Thuja occidentalis) which was part of the front wall at the northwest corner, and a fragment of a 1-foot-wide beam 12 feet south of this down the center of the floor. Projecting dimensions from these, we arrived at the following approximations: front wall, 60 feet long; rear wall, 68 feet long; west end wall, 22.5 feet long and east end wall, 23.5 feet long. The west wall was at a 90-degree angle and the east wall at a 70 degree angle to the rear wall. Beyond this, there was no information on the structure. It probably was a mirror image of the southwest casemate except that it was longer. Level I was a gray loam with post-1781 ceramics in it over the entire area which extended down to sterile soil over the eastern three-fourths of the structure. Only on the west end could we find evidence of undisturbed Level II. This was a brown sandy loam. Features 14, 24, 25, 26 and 27 were small post-1900 rubbish pits dug into the sterile subsoil.

East Casemate

We know the least about this structure. All we found was a fragment of a 1-foot-wide beam for the rear wall and a large trench 30 feet long, 5 feet wide and 3 feet deep which was 10 feet from the rear wall in the northern half of the structure. Running through the approximate center of the casemate on an angle of 6 degrees south of the rear wall was a drainage trench which is described under the section to follow entitled "Drain." This ended where the front wall of the casemate should have been, 22 feet from the rear wall, the only dimension we could determine for this structure. Level I was a gray loam with post-1781 ceramics in it over the entire area extending down to sterile soil. Two pipelines and a deep stone-lined cellar (ice house?) (Feature 55) built before 1851 (city map) and not closed until the early 20th century accounted for some of the destruction. Level II was a brown sandy loam and was confined to the 30-foot-long trench.

Communication

. . . trenches that unite one work to another, so that men may pass between them without being exposed in the enemy's fire . . . (G. Smith, 1779, p. 149)

See section to follow entitled "Sally Port."

Covered Way

. . . is a space of 6 fathom broad (36 feet), going quite around the works, and is adjoined to the counterscarp of the ditches, covered by a parapet 7-1/4 feet high, including a banquette of 2-1/2 feet . . . (G. Smith, 1779, p. 101)

The covered way at Fort Stanwix was never this broad except at the corners of the ditch where the salient angles of the glacis went beyond the rounded corners of the ditch. At the only point where the covered way was located it was 10 feet broad. This was opposite the west face of the southwest bastion. Level I was a gray loam with post-1781 ceramics. At the point where we located the covered way we found a burned hearth area 1.7 feet in diameter (Feature 47) with two wrought iron nails, a strike-a-light flint made from a Dutch gunflint and a pipe stem. This probably represented a temporary guardpost from the 1777 period when a quarter Guard was mounted on the covered way (M. Willett, 9/10/77).

. . . the quarter-guard . . . is to maintain good order in the camp. This guard now fronts outwards, though it was formerly the custom to front inwards . . . and as this guard is merely to maintain good order in the camp the present alteration seems improper. The quarter-guards of all Foreign troops front inwards . . . (G. Smith, 1779, p. 311)

The purpose of the quarter guard at Fort Stanwix was probably to prevent desertions; hence their post on the covered way enabled them to observe anyone climbing over the walls.

At the base of the glacis on the covered way was a trench 2 feet wide and 2.5 feet deep (Level III). There were traces of postmolds in the trench about .5 to .7 feet in diameter and set only a few inches apart. We know that in 1777, a line of pickets was placed on the covered way and that the posts were 10 feet long (W. Willett, 1831, p. 44). This would mean that they portruded 7.5 feet above the covered way. It appeared to us that these posts had been pulled out at some time in the past since some of the postholes had collapsed and others were filled with debris from Level I including a blue transfer-printed pearlware potsherd matching specimens from Level I. The pickets were still standing after the fire of 1781 (G. Clinton, 1900, Vol. XI, p. 878) and some were still visible as late as 1828 (Wager, 1896, p. 52).

Curtain

. . . is that part of the body of the place, which joins the flank of one bastion to that of another. (G. Smith, 1779, p. 101)

The curtains at Fort Stanwix were earth-filled log structures (Carroll, 1973, pp. 103-106). On the inner side they were revetted up to the level of the terrepleine (7.5 feet) as the rear walls of the casemates. The thickness of the curtain walls was determined by measuring the distance from the rear of the casemates to the edge of the ditch and then subtracting the probable width of the berm. We arrived at the following dimensions: north curtain thickness, 16 feet; west curtain, 19 feet; south curtain 12 feet and east curtain, 16.5 feet. It should be noted that the thickest walls face the most probable direction of enemy attack. Wood that lay on the scarp of the northwest bastion in Level X has been identified as northern white cedar (Thuja occidentalis) and was probably from the rampart wall.

Counterscarp

. . . is the outside of a ditch, opposite to the parapet of the work behind the ditch . . . (Muller, 1746, p. 215)

Two measurements on the sterile soil of the counterscarp produced angles of 43 degrees and 37 degrees from horizontal. An average of 40 degrees would mean that the foot of the counterscarp would have been 12 feet out from the top at a depth of 10 feet.

Ditch

In 1758, the British attempted to put a ditch completely around the fort (fig. 4). The ditch on the east side was abandoned and, by 1764, the ends of the ditch were closed at the southeast and northeast corners of the fort (Crown Collection Map CVXXI, 103). At the same time, a causeway before the main gate was replaced by a wooden bridge. The bottom of the ditch was at an elevation of 441.08 feet at the north face of the northwest bastion, 440.78 feet at the west face of the southwest bastion and 440.80 feet below the bridge on the south side of the fort. We projected the angles of the scarp and counterscarp and found the following widths of the ditch: top of north (fig. 32) and south ditches at the center of the curtain, 59 feet; at the bastions, 42 feet; bottom of north and south ditches, 35 feet and 18 feet, respectively; top of west ditch at the center of the curtain, 54 feet; at the bastions, 39 feet; bottom of west ditch, 30 feet and 15 feet respectively. The bottom of the ditch was dug into sterile graded sand. For this reason it drains very well and was probably dry most of the time. Levels I-X in the ditch contained post-1781 ceramics and constituted the fill put in the ditch to landscape the area. Some of this fill was undoubtedly the walls of the fort (Wager, 1896, p. 52). Level XI at the bottom of the ditch was fort related and dated ca. 1758-1781. In the bridge area the complexity of the stratigraphy led us to divide Level XI into Levels XI-XV.

Figure 32. Cross section of the north ditch showing natural stratigraphy. (click on image for a PDF version)

Drain

A wooden drain 1 foot wide ran through the east casemate and beneath the sally port at an angle of 6 degrees to the long axis at the east casemate (figs. 33, 34). This drain was a wood lined box type judging by the nails found. It was set in a trench and covered with sterile soil. It apparently served to keep water from the east barracks roof from running through the sally port and ruining the footing. We lost track of it at the edge of the berm but a trench through the glacis of the communication at an angle of 64 degrees north of the communication to the redout, and 15 feet out from the edge of the berm, probably was a continuation of the drain. This indicated that the drain angled over to the north side of the communication below the steps and continued out under a wooden banquette. It probably emptied into the creek east of the redout. The drain began at the front of the east casemate. No grill was found but it may have been simply a barrel top with holes drilled through it. The elevations of the bottom of the drain are 447.74 feet at the beginning, 446.75 feet at the berm and 443.72 feet where it left the communication.

Figure 33. Plan of the sally port, drain, communication and redout. See figure 34 for section along the A-B line. (click on image for a PDF version)

Figure 34. Section through the sally port, drain, communication and redout. (click on image for a PDF version)

Drawbridge

See section entitled "Bridge."

Gateway

The main entrance to the fort lay behind the drawbridge between the southwest and southeast casemates (fig. 35). The distance between the casemate walls was 13.5 feet. Inside of this was a large wooden frame which probably supported the gate and the lifting mechanism of the drawbridge. The sills of the frame were 1.5 feet wide and rested on sterile soil. They were 18 feet long and projected 5 feet south of the casemates onto the edge of the indentation of the scarp below the bridge.

Figure 35. View of the main gate excavation area, looking north-northeast. Note the brick drain on the right side. Rubble fill in the foreground is from two 19th-century outhouses disturbed by Campbell's 1965 test trench. Width of the gateway is 10.5 feet.

Along the east side of the gateway was a brick drain, one brick length wide with a line of bricks on edge along the western side. The drain dropped from an elevation of 450.60 feet above mean sea level to 450.36 feet about 4 feet from the edge of the scarp where it ended. The northern end had been destroyed by a test trench dug in 1965 (Campbell, 1965). Level I was a gray loam with post-1781 ceramics and Level II was a brown sandy loam layer over sterile soil.

No remains of the main gate itself were found. Spikes found in Level XIII of the ditch under the bridge probably were used to stud the gate. If so, the gate was 7 inches thick, since the spikes were clinched at 7 inches. Considering the width of the frame, it was probably a double-hung gate. It probably hung at the rear of the frame and swung inward into niches created by the frame and the casemates to leave the gateway unimpeded.

Glacis

. . . is the part beyond the covert-way, to which it serves as a parapet and terminates toward the Field in an easy slope at about 20 fathom distance [120 feet]. (G. Smith, 1779, p. 102)

The glacis at Fort Stanwix appears never to have been 120 feet broad. No two historic plans agree as to its actual width but none show it wider than 75 feet. We were unable to get an angle of the slope from which to project its width. We did uncover the base of part of the parapet of the glacis opposite the west face of the southwest bastion. Its slope measured 40 degrees. Projecting a height of 6 feet (see "Parapet") and allowing a slope of 5 degrees for the face of the glacis, the width of the glacis would be 75 feet. At the southeast and northeast ends it should have been 10 feet narrower (Crown Collection CXXI, 102 & 103). A slope of 5 degrees was selected inasmuch as this would project the face of the glacis to the sole of the embrasures on the parapet, giving the maximum field of fire to the fort's defenders.

Laboratory

. . . signifies that place where all sorts of fire-works are prepared, both for actual service, and for pleasure, vic. quick-matches, fuzes, portfires, grape-shot, case shot, carcasses, hand-grenades, cartridges, shells filled, and fuzes fixed, wads &c. &c. (G. Smith, 1779, p. 135)

In a plan prepared ca. 1778, Defleury identified the west barracks as a laboratory (W. Stone, 1838, p. 230). Nothing found in the excavation confirmed that identification. The material found in Feature 69 suggests that if there were a laboratory, it was in the east barracks.

Magazine

See section entitled "Bombproofs."

Main Gate

See section entitled "Gateway."

Necessary

A necessary was a privy or outhouse. On the 1758 plan of Fort Stanwix (fig. 4) and the Gansevoort Plan (ca. 1777, fig. 7), large wooden necessaries were depicted off the east face of the southeast bastion, standing on stilts over the small creek that ran along the east side of the fort. Passage to the privy was accomplished by a catwalk from the bastion. The orderly book for the Third New York Regiment (M. Willett, 1777-1778) mentioned building new necessaries from time to time in the ditch and elsewhere. The site of the necessaries off the southeast bastion was occupied by modern buildings and could not be excavated.

We found three pits on the parade ground, two of which we interpreted as necessaries (fig. 10). Feature 73 was 14 feet north of the center of the fort along the midline, and consisted of a rectangular pit 9 feet long, 5 feet wide and 4 feet deep. A northern white cedar post (Thuja occidentalis) was found in the upper part of the fill, 18 feet from the center of the fort on the midline (See "Whipping Post"). This contained sand to a depth of 4 feet with a layer of gray loam on the bottom of the pit. Artifacts included a Type B-3 button, 20 mortar bomb fragments, seven pieces of birdshot, a .53 cal. musketball, a .58 cal. musketball, a .63 cal. musketball, three .69 cal. musket balls, two brass pins, a Type 4a pipe bowl, four Type 1b porcelain potsherds, two Type 1d porcelain potsherds, seven Type 1a stoneware potsherds, two Type 1b stoneware potsherds, two Type 1a earthenware potsherds, two Type 1b earthenware potsherds and two Type 2a earthenware potsherds.

Feature 58 was 5.5 feet in diameter and 4.5 feet deep located 24 feet north of the southwest casemate and 27 feet east of the west barracks. There was some charred wood around it which may have been related to the guard house. The fill was sand except for a thin layer of gray loam on the bottom of the pit. It contained a Type D-3 button, eight mortar bomb fragments, two .63 cal. musketballs, a .69 cal. musketball, a brass pin, a Type 1d stoneware potsherd, a Type 1a earthenware potsherd and a Type 2a earthenware potsherd.

Feature 70 was only 1 foot east of Feature 58 and was approximately 8.5 feet in diameter and 2.5 feet deep. It was filled with sand and there was only eleven badly corroded nail fragments and a Type 5c earthenware potsherd.

Parade Ground

. . . the place where troops assemble, before they go on duty. (G. Smith, 1779, p. 310)

The parade at Fort Stanwix was only 90 feet long, 85 feet wide and located in the center of the fortifications. The parade ground elevation was established as the top of the sterile soil zone where there was a faint trace of a living surface (Level II) in the form of bits of charcoal, bone splinters and pipe stems. Level I was a gray loam with post-1781 ceramics. Elevations taken around the site resulted in an average elevation of 451 feet to the parade ground with a variation of less than 6 inches above or below this reading.

Parapet

. . . an elevation of earth, designed for covering the soldiers from the enemy's cannon, or small shot . . . its height 6 (feet) on the inside . . . of the covert-way, is what covers that way from the sight of the enemy . . . (G. Smith, 1779, p. 199)

No evidence of the parapet was found on the rampart. For the parapet on the glacis see the section entitled "Glacis."

Palisades

. . . are a kind of stakes made of strong split wood of about 9 feet long, fixed 3 feet deep in the ground, in rows about 6 inches asunder: they are placed in the covert-way, at 3 feet from, and parallel to the parapet of the glacis . . . (G. Smith, 1779, p. 103)

. . . stakes made of strong split wood, about 9 feet long. 6 or 7 inches square, 3 feet deep in the ground, in rows about 2-1/2 or 3 inches asunder, placed in the covert-way . . . (G. Smith, 1779, p. 198)

When Fort Stanwix was initially constructed, a palisade was placed around the fort in the ditch (fig. 4). The posts were about 11 feet long and sunk 3 feet in the ground. We found part of this palisade line in the north ditch at the base of the northwest bastion (fig. 10). The posts were badly decomposed and there were gaps in the line where some had apparently rotted away without a trace. The sections most intact were posts spaced 1 foot apart, center to center, and were about .5 to .7 foot in diameter.

In 1777, the palisade line was moved to the covered way (W. Willett, 1831, pp. 44, 49). Here we found a trench 2 feet wide and 2.5 feet deep along the base of the glacis into which the posts had been set. Inasmuch as these posts were 10 feet long (W. Willett, 1831, p. 44) they must have projected 7.5 feet above the covered way and 1.5 feet above the glacis. It was difficult to discern postmolds, but a few were found about .5 to .7 feet in diameter. When the fort burned in 1781 the palisade apparently remained intact (G. Clinton, 1900, Vol. XI, p. 878). At some later date, these posts were probably pulled out and used for other purposes as the fill in the postmolds differed from the trench in which they were set. They contained more man-made debris including one blue transfer-printed pearlware potsherd, matching sherds in Level I. Some pickets in this area were still visible in 1828 (Wager, 1896, p. 52). If the northwest bombproof walls were made from the remains of these pickets as Willett reported (1831, p. 49), the posts were probably eastern hemlock (Tsuga canadensis) since this was the type of wood used to construct the northwest bombproof.

Rampart

. . . an elevation of earth raised along the faces of any work, of 10 or 15 feet high, to cover the inner part of that work against the fire of an enemy . . . (G. Smith, 1779, p. 103)

See section entitled "Curtain" for the thickness of the ramparts at Fort Stanwix.

Ravelin

. . . works raised on the counterscarp before the curtain of the place, and serve to cover the gates of a Town, and the bridges. They consist of two faces forming a saliant [sic.] angle, and are defended by the faces of the neighboring bastions . . . (G. Smith, 1779, p. 221)

There was a ravelin before the bridge in 1764, (fig. 5); repaired to some extent by the Americans (W. Willett, 1831, p. 45). Archeologically, the only evidence was a trench (Feature 76) 33+ feet long, 8 feet wide and 5 deep deep across the east face of the ravelin and a disturbed deposit of trash in the area where the outer gate should have passed through the ravelin. Level I was a gray loam containing post-1781 ceramics. Level II was the trash deposit. The purpose of the trench was not ascertained, although it was probably defensive. It would have necessitated a bridge before the outer gate and we suspect it was used only a short time. It was steep sided, the walls having slumped from erosion. The original width may have been closer to 6 feet. The trench fill contained a 1732 British halfpenny, four Type A-1a buttons, five Type B-2 buttons, a Type B-3 button, two Type D-1 buttons, eight Dutch gunflints, a French gunflint, a Type 1 gun cock, two bayonet fragments, a frow, a Type 1b axe, a Type 3b axe, a Type 4d axe, a Type 4 spade, three pairs of scissors, a Type 1a key, four Type 1b porcelain potsherds, 42 Type 1a stoneware potsheds, 14 Type 1b stoneware potsherds, a Type 1d stoneware potsherd, 47 Type 1a earthenware potsherds, 40 Type 1b earthenware potsherds, 14 Type 1d earthenware potsherds, three Type 2a earthenware potsherds, a Type 2d earthenware potsherd, and a Type 2e earthenware potsherd. We found nothing in the trench to indicate it was in use during the Revolution, and it was probably filled prior to that date.

Assuming that the trench was parallel to the face of the ravelin, the faces of the ravelin would have been at an angle of 55 degrees to the midline of the fort, and at least 60 feet long. From the 1764 plan (fig. 5) they would have been 66 feet long but inasmuch as the salient angle we found was more obtuse than that depicted, the faces would have had to have been 77 feet to extend as far into the covered way as they did on the 1764 plan, and give maximum protection to the main gate. On the 1764 plan, the center of the outer gateway was 30 feet from the salient angle while our trash deposit was 36 feet from the salient angle.

The deposit included a Type A-1b button, a Type A-1j button, two Type D-3 buttons, two Type F-2 buttons, a Type F-4 button, four Dutch gunflints, three French gunflints, a Type 1 pick, a Type 3 key, two Type 1a porcelain potsherds, seven Type 1b porcelain potsherds, five Type 1a stoneware potsherds, six Type 1b stoneware potsherds, a Type 2a stoneware potsherd, eight Type 1a earthenware potsherds, two Type 1b earthenware potsherds, a Type 1c earthenware potsherd, a Type 1d earthenware potsherd, four Type 2a earthenware potsherds, a Type 2c earthenware potsherd and five Type 2d earthenware potsherds.

The elevation of the works was 448 feet. This made it possible for the ravelin to contain a 10-foot-high gateway and still be low enough for defenders on the curtain wall behind it to fire over the top. This elevation also made it necessary for a ramp to be built at the south end of the bridge to compensate for the 3 feet difference in elevation.

Redout

. . . a kind of work placed beyond the glacis, and is of various forms . . . (G. Smith, 1779, p. 103)

The one redout at Fort Stanwix covered the sally port entrance on the east side of the fort (figs. 33, 34). It was pentagonal and the body of the work stood at an elevation of 443 feet above mean sea level. It was fortified with a row of posts surrounded by an earthen parapet or glacis. Those found were .4 to .5 foot in diameter and set .7 to 1.0 foot apart, center to center. Only 11 of these postmolds of the redout were found. There was a banquette 3.5 feet wide and 1.5 feet high around the inside of the posts and an earthen parapet outside the palisade line. The angle of the one line of posts was 43 degrees from the midline, while the banquette base on the other side was 40 degrees from the midline. The approximate interior lengths: east flank, 5.5 feet; east face, 14 feet; south face, 18 feet; south flank, 4.5 feet and northwest face, 21 feet. The northwest face was joined to the sally port by a covered communication 9.5 feet wide. The southwest corner extended 2.5 feet beyond the corner of the communication with a gate approximately 3 feet wide on this side for access to the small creek on the east side of the fort.

Level I was a gray loam with post-1781 ceramics. Level II was a deposit of garbage described under the section "Sally Port," which was disturbed by tree roots along a retaining wall on the east side. Level III was a brown sandy loam on the floor of the redout. It contained two Dutch gunflints, a cannon primer, a wheel hub, a Type A-1a button, a Type A-1i button, a Type B-1 button, a Type B-2 button, a Type 1a porcelain potsherd, a Type 1b porcelain potsherd, five Type 1a stoneware potsherds, four Type 1b stoneware potsherds, a Type 2b stoneware potsherd, seven Type 1b earthenware potsherds, four Type 1d earthenware potsherds, two Type 2a earthenware potsherds and a Type 4 earthenware potsherd.

Revetment

. . . a strong wall built on the outside of the rampart and parapet, to support the earth, and prevent its rolling into the ditch . . . (G. Smith, 1779, p. 103)

The only revetment found at Fort Stanwix was on the scarp of the north ditch which had been sodded to keep the earth in place. The sod was apparently laid on flat, rather than cut and stacked like bricks. To hold it in place, wooden stakes .5 to 1.5 feet long were driven into the scarp at 1-foot intervals in rows .7 foot apart. The rows tended to drop slightly in elevation from east to west.

Sally Port

. . . are those under-ground passages, which lead from the inner works to the outward ones . . . When they are made for men to go through only, they are made with steps at the entrance, and going out. They are about 6 feet wide, and 8-1/2 feet high. There is also a gutter or shore made under the sally-ports, which are in the middle of the curtains, for the water which runs down the streets to pass into the ditch: but this can only be done when there are wet ditches. (G. Smith, 1779, p. 312)

The sally port at Fort Stanwix was located in the east curtain wall of the fort (figs. 33, 34). No trace of it was found through the east casemate except the drain that ran through, 6 degrees south of a right angle to the rear wall of the casemate (see section entitled "Drain"). Through the rampart and out onto the berm it was also marked by charred sill beams and sleepers. This indicated a passageway 5 feet wide on the exterior with .5 foot thick walls. The first arm of the sally port, through the casemate, was 22 feet long and the second arm, through the rampart, 21 feet long. The angle between the two arms was 24 degrees.

The sally port at the berm entered a communication 61 feet long which ran down the scarp and out to a redout. This was 9.5 feet wide with .5-foot thick white pine sleepers (Pinus strobus). The remains of three steps were found at the western end of the communication on the scarp. The end of the sally port projected 3 feet into the communication on the berm. The communication was anchored by two 9-inch squared uprights in the exterior of the rampart. At the eastern end there was a sleeper with two nails driven vertically into it 2 feet apart. They probably had something to do with a threshold since the door at the end of the communication must have been wider than this. This sleeper, and another one 10 feet west of it, were 1 foot wide.

Above the floor of the communication and the redout was a deposit of dark gray loam containing numerous artifacts. The presence of a number of Revolutionary War period buttons in this midden (Level II) indicated a date of deposit between 1776 and 1781. Since the deposit would have interferred with access through the communication, this must have been abandoned and torn down sometime after 1776. The upper surface of this deposit was slightly concave, indicating that it formed the base of a trail from the sally port to the redout. Artifacts in this deposit included a 1732 British halfpenny, a signet inset, a Type 4b axe, a Type 4c axe, a Type 2 spade, a saw fragment, two scissor fragments, a bone button bit with unfinished blanks, a .62 cal. two-pronged gun worm, an escutcheon plate marked "17th R^t.", a Type 2 lock plate, 21 Dutch gunflints, 23 French gunflints, a cannon primer, ten mortar bomb fragments, 13 Type A-1a buttons, three Type A-1b buttons, two Type A-ic buttons, a Type A-1d button, a Type A-1e button, a Type A-1g button, four Type A-1i buttons, a Type A-1j button, a Type A-1k button, a Type A-1l button, a Type A-1z button, a Type A-3 button, a Type A-4 button, a Type B-1 button, five Type B-3 buttons, a Type D-2 button, two Type D-3 buttons, a Type D-4 button, eight Type E-1 buttons, two Type F-2 buttons, seven Type 1a porcelain potsherds, 24 Type 1b porcelain potsherds, two Type 1d stoneware potsherds, 53 Type 1a stoneware potsherds, ten Type 1b stoneware potsherds, two Type 1d stoneware potsherds, 32 Type 1a earthenware potsherds, 39 Type 1b earthenware potsherds, a Type 1d earthenware potsherd, 57 Type 2a earthenware potsherds, five Type 2b earthenware potsherds, a Type 2c earthenware potsherd, three Type 2d earthenware potsherds, seven Type 2e earthenware potsherds, two Type 2f earthenware potsherds and a Type 2g earthenware potsherd. Some ceramics from this deposit could be cross-mended with sherds from the west end of the north casemate. On the floor of the communication and the scarp below the sally port (Level III) there was a brown sandy loam which contained three Dutch gunflints, a French gunflint, a Type A-la button, a Type A-1b button, a Type A-1e button, a Type B-3 button, a Type D-2 button, a Type E-1 button, a Type F-2 button, a white oak (Quercus) scrub brush, two Type 1a porcelain potsherds, nine Type 1b porcelain potsherds, four Type 1d porcelain potsherds, thirty-six Type 1a stoneware potsherds, seven Type 1b stoneware potsherds, a Type 1d stoneware potsherd, 22 Type 1a earthenware potsherds, 20 Type 1b earthenware potsherds, four Type 1d earthenware potsherds, four Type 2e earthenware potsherds and a Type 4 earthenware potsherd.

The floor elevation of the sally port was 451.10 feet. The communication dropped from 450.09 feet on the berm to 444.08 feet at the base of the scarp and 442.96 feet at the eastern end. Banked against both sides of the communication was a ramp of earth or glacis. The top was about 4.5 feet above the floor level and sloped at an angle of 10 degrees. This produced a width of 25 feet to the embankment.

Scarp

. . . the interior talus or slope of the ditch next [to] the place, at the foot of the rampart. (G. Smith, 1779, p. 228)

The scarp at Fort Stanwix was measured at several points and averaged 40 degrees in slope. Below the bridge, on the south curtain it was 45 degrees.

On the east side of the fort a deposit of pre-1781 artifacts was found on the scarp. Level I was a gray loam over some lenses of brown loam which were post-1781 deposits used to landscape the area and build out the slope to a ca. 1839 retaining wall along Spring Street. Level II was a deposit of pre-1781 artifacts lying on the slope of the scarp north of the communication at the angle of the glacis of the communication. Artifacts found include a George II halfpenny, a Type 2 axe, a Type 7 axe, a Category 1 hammer, a sad iron, a 4-pounder cannonball, a 6-pounder cannonball with a broad arrow, five mortar bomb fragments, twenty-two Dutch gunflints, nine French gunflints, a "Jesuit" ring, five Type A-1a buttons, a Type A-1b button, a Type A-1f button, a Type A-1i button, a Type A-1k button, a Type A-1z button, a Type A-4 button, six Type B-1 buttons, 15 Type B-2 buttons, three Type B-3 buttons, two Type D-3 buttons, a Type D-4 button, a Type E-1 button, a Type E-2 button, a Type F-2 button, a Type G-1 button, 12 Type 1a porcelain potsherds, 38 Type 1b porcelain potsherds, three Type 1d porcelain potsherds, 55 Type 1a stoneware potsherds, 30 Type 1b stoneware potsherds, a Type 2a stoneware potsherd, a Type 3 stoneware potsherd, 82 Type 1a earthenware potsherds, 67 Type 1b earthenware potsherds, 13 Type 1c earthenware potsherds, ten Type 1d earthenware potsherds, 48 Type 2a earthenware potsherds, two Type 2b earthenware potsherds, seven Type 2e earthenware potsherds and two Type 4 earthenware potsherds.

The presence of Revolutionary War buttons indicated that this deposit was made between 1776 and 1781. Despite the presence of a few military objects this was predominantly a deposit of household trash and cross-mends of a few potsherds indicate that it came from the east end of the north casemate. Higher up on the east scarp just below the berm and 4 feet north of the sally port communication, stood a brick fireplace (Feature 68) (fig. 36). The jambs and back wall were .8-foot thick and were made from alternating courses of headers and stretchers. It measured 4.8 feet wide at the back and was 3 feet deep. The jambs expanded to a thickness of 1.2 feet at the ends and were at an angle of 93 degrees to the back wall. The hearth was paved with bricks laid at right angles to the back wall at an elevation of 449.32 feet. It projected 1.3 feet beyond the front of the fireplace and was approximately 7.8 feet wide. Its position is a mystery, but there may have been some temporary shelter erected in this relatively protected area between the rampart and the communication. There was a "USA" button (Type A-1b) on the hearth indicating it was used during or after 1776 and a broiler made from a folded iron strap (barrel hoop?).

Figure 36. View of the brick hearth on the east scarp, looking west-northwest. Note the barrel hoop broiler fragments on the floor. The front edge has eroded away. The back wall of the fireplace is 4.8 feet wide.

Terrepleine

. . . the top platform, or horizontal surface of the rampart, whereon the cannon are placed, as well as the troops that defend the place . . . (G. Smith, 1779, p. 243)

The terrepleine of the curtains at Fort Stanwix was formed by the roofs of the casemates. The bastions had solid earth-filled terrepleines which sloped down slightly toward the parapet to take up the recoil of the cannon. For this reason the casemate roofs must have been lower on the rampart side.

Whipping Post

The stump of a white cedar post (Thuja occidentalis) was found 18 feet from the center of the fort along the midline. Its position, which would be approximately 28 feet in front of the headquarters building, suggests that it was the remains of a whipping post. The fort is known to have had one capable of holding four men at once (Elmer, 1848, p. 138).

Summary

Fort Stanwix was a highly organized military post with all available space used for quarters or some other purpose. Our excavations were facilitated by this organization, and the resulting symmetry and the existence of a number of British and American plans of the fort (Luzader, 1969) made while it was in use. We took advantage of this information in planning our excavations, and attempted to excavate each structure as an entity in the hope that we would be able to tell from the artifacts recovered just how the buildings were used, draw inferences about the social and economic status of the occupants, and separate British and American components. For a variety of reasons, principally disturbance and the short time frame of the 18th-century components, 1758-1781, the data was not what we had hoped for. Because of the disturbance, few of the buildings were intact enough that they could be thoroughly excavated. With rare exceptions what was excavated in terms of the artifacts was rather homogeneous; approximately the same types of artifacts in nearly the same frequencies showed up in all the structures. Finally, our efforts to separate documented British and American components came to naught. The only structural evidence which supports the documentation is: the presence of clay-lined and wood-lined cellar holes with two of the former intruding two of the latter, six single-hearth fireplaces in the north casemate where the British plans show three double-hearth "H-shaped" fireplaces, and a bakehouse built over a 1758 powder magazine which had collapsed prior to the building of the bakehouse. None of these really gave us any clear-cut distinction between the two components, because we could not document exactly when these changes took place except that they occurred after 1764, the date of the last British plan for the fort, and 1781, when the fort was destroyed and abandoned. This time span overlaps both periods of occupation. Trash deposits on the east scarp, in the sally port communication, and in Feature 69 clearly date after 1776, but we strongly suspect that earlier material is included in the deposits and application of the South mean ceramic date formula confirms this (South, 1972).

The documentation on the fort shows that it was being repaired constantly and, in some cases, modified. The major changes were made in 1763-64 when the east and west barracks and headquarters were constructed. Sometime later, by 1777, a guardhouse and storehouse were added and three bombproofs and a bakehouse were put into the bastions. From the excavated remains, it is hard to escape the conclusion that the Americans either completely eradicated the British foundations when they rebuilt the fort, or found substantial parts of it intact and merely made repairs. The fact that the rebuilding was done in a relatively short time favors the latter conclusion. Although the British barracks burned in 1774 (Duncan, 1969) there was still habitable space occupied by a trader, probably in one of the casemates.

Poor preservation prevented us from making any detailed study of construction methods. It was noted that the six fireplaces in the north casemate were not squared to the rear wall, and that the entire east barracks was slightly out of line (2 degrees) with the rest of the fort. This is just not enough evidence to generalize on the abilities of the workmen who built the fort. Basically, the archeological work enhanced our knowledge of the location, size and shape of Fort Stanwix and its component structures, assisted us in identifying some of the materials used in its construction (Wood, bricks, mortar and hardware) and provided additional information on the distribution of certain artifact types in the period 1758-1781. Figure 10 depicts a plan of Fort Stanwix as projected from the archeological evidence, incorporating documentary evidence to fill in the holes. Figure 8 shows how the fort would look based on archeological evidence, historic plans, contemporary descriptions and comparable forts.