For the reasons summarized at the end of the last chapter, the artifacts from the fort cannot be separated stratigraphically into two components, and must be considered to date between 1758 and 1781 except where datable artifacts narrow this time range. Examples include Type 2a earthenware pottery (creamware) which was introduced after 1762 (Noel Hume, 1970, p. 125) and American uniform buttons which date post-1775. The vast majority of the artifacts were manufactured during a period that spanned the entire time range of the occupation of the fort. The few pieces that can be identified positively as of English manufacture mean little because captured military goods were redistributed to the American army, or sold at public auction (Willet, 8/31/77).
Another method that can be used to identify the artifacts associated with each component of a long term occupation is comparison with other short-term occupations or single component sites in the same time period. The only comparable site for which we found published data was Fort Ligonier (Grimm, 1970). Other sites either have not been published on or were occupied longer than Fort Stanwix. General references were of limited use for the same reason as that cited by Stanley South concerning buttons at Brunswick Town and Fort Fisher ". . . none gave the type of detailed information which would prove of value in a typological analysis such as would be most useful to the archeologist . . ." (South, 1964, p. 113).
The first step in analyzing the artifacts from a site is a description of the material. Because the Fort Stanwix collection numbered nearly 42,000 specimens (not counting the post-1781 material) serving a wide range of functions, we found it necessary to resort to a classification system in order to handle the data. In this report we have used a system based on the function and form of the artifact. We have used a graduated breakdown into series, type and variety levels adapted from L. Stone (1970). Series are divided on the basis of function and/or place of origin. Types are defined on the basis of observable manufacturing techniques or major variations in form. Varieties are generally divided on the basis of decoration or minor variation in form. Like all classifications, this one does not encompass the full range of artifacts recovered. Although it is flexible to a degree, we have abandoned it when the need dictated so as not to obscure what we consider significant at tributes by too rigid an adherence to the system. It should be stressed that this classification was adopted solely for the purpose of grouping similar artifacts in order to facilitate their description. This description may seem unnecessarily complex to some and non-productive in terms of generating new data, but we feel that a less detailed description along strictly functional lines might not meet the needs of future researchers. Although we cannot predict those needs, it is logical to assume that when a larger corpus of data is available, patterns will show up which are not now apparent. We hope, in fact, that this report will lead others to look for such patterns.
To some extent our approach implies levels of relationship between varieties within types and types within series, as opposed to other classifications such as that by South (1964) which imply no relationship between types. Anyone familiar with the data will recognize that neither of these implications are wholly true or false and the two methods are both valid approaches depending upon the aims of the classifier. Again, our aim here is to describe the artifacts found at Fort Stanwix with as little repetition as possible.
One problem encountered in the description and classification of artifacts involves the ravages of time. Everyone knows that wood decomposes and iron rusts, but few consider how much of an artifact inventory is perishable. Thus, under normal conditions, my desk would be reduced in time to eight drawer pulls, a lock, a few screws and a handful of paper clips, provided it was not disturbed and some items carried off. Someone excavating my desk would be hard pressed to identify it except by the presence of paper clips. An accurate reconstruction would be impossible if there existed no drawings or identical desks to work from. This is the problem we faced at Fort Stanwix, and the reader should know that we have occasionally been reduced to making educated guesses preceded by the words "probably" and "possibly" or have put certain artifacts into "categories" because we could not type them due to poor preservation and breakage.
Because there was post-1781 occupations of the site, one of our first problems was to separate the later material from the fort artifacts. The bulk of the fort-related artifacts was found in the undisturbed strata described in the previous chapter. We also included those artifacts in the upper disturbed zones which could be clearly identified as pre-1781 by comparison with artifacts found in pre-1781 context or reference to published descriptions. For example, a sad iron found on the surface was identical to two found in pre-1781 contexts, and was included with them for descriptive purposes because none of this type were found in post-1781 features such as privies and cellars. On the other hand, pearlware, perfected by 1779 (Noel Hume, 1970, p. 128) was not found in a pre-1781 context, but was common in early 19th-century privies. This led us to omit pearlware from this report on the basis that an association with Fort Stanwix could not be established even though it was temporally possible. This task was made easier by the 20-year hiatus in the occupation after the destruction of the fort, during which there were modifications in artifacts. The association of most of the post-1781 artifacts were with privies, house foundations and utility lines. Where a separation is not possible, it is so noted. In general, the only instances where we allude to post-1781 artifacts is to make a comparison with the fort-related material.
The spatial distribution of artifacts was handled on two levels, within the site and with other sites. With most types of artifacts, we compared the relative amounts in the various buildings of the fort, the ditch, the parade ground (which included the gateway), the ravelin and the dumps on the east scarp and in the sally port complex (actually, in the communication and redout). As noted, we hoped to achieve a better understanding of the function of each building by its contents. We included the artifacts in the Campbell collection of 1965, material not identified by provenience, although most of it apparently came from the bakehouse and southeast casemate. For purposes of comparison we converted the actual number of specimens found to a standard ratio per 100 square feet of excavated area per structure. This is a crude technique in that it lumps in situ artifacts along with artifacts from the fill, but an examination of the in situ artifacts alone which formed a preliminary study for this report did not generate any testable hypotheses, primarily because there were not many in situ artifacts compared to the total collection. It is assumed that there was little lateral movement in the artifacts, an assumption that is borne out by the relative absence of midden outside the structures except on the east scarp. We were thus able to treat each structure as an entity for quantitative analysis. In some cases, as in the barracks, it was feasible to break down the artifacts by the various cellar holes. By and large, there were too few comparable artifacts to make any meaningful comparisons above the presence or absence level in the various buildings and dumps, but the figures are included for the record in the event that someone else wants to compare their buildings and trash deposits with ours. Comparisons with other sites by us were hampered by a lack of published material, and the inability of the authors to examine comparable collections due to a lack of travel funds and time. Through the courtesy of Mr. Horace Wilcox and staff, we were able to look at some of the material from Valley Forge and Dr. Lyle Stone permitted us to use his Ph.D. dissertation on Fort Michilimackinac. The quality of the published data varied considerably. Thus, it was possible to make definitive statements about the pipes and gunflints, but we found ourselves in virgin territory with such items as window glass and shovels.
An attempt to cross-mend potsherds from the site showed that a significant number of vessels (6.7 percent) from the north casemate had sherds in the dumps on the east scarp and in the sally port. Material from the west end of the north casemate (Levels II and III) was dumped in the sally port (Level II) and material from the east end of the north casemate (Level I) was dumped on the east scarp (Level II). There was one cross-mend between the sally port (Level II) and the south end of the east barracks (Level I) and another between Level II of the west end of the southwest casemate and Level I of the south end of the west barracks.
The artifact descriptions that follow are grouped under general headings so that functionally similar objects are described together. There are a few items at the end which could not be fitted in and the best way to find these is to refer to the index. A summary of the artifacts is incorporated in the last chapter.
The structures at Fort Stanwix were of wood, either of log or frame construction. The casemates and northwest bombproof were constructed of logs; the buildings on the parade ground, and probably the northeast and southwest bombproofs and the bakehouse, were frame buildings. All the frame buildings were quite flimsy as this report of a fire in the guardhouse at midnight on April 13, 1780, suggests:
It is likely that the buildings were probably quite simple with wooden walls, floor and roof and interior partitions to set off rooms. Judging from documents, each had a door and at least one window and fireplace. Building hardware was kept to a minimum: nails, rivets and staples for joining wood; pintles and hinges for hanging doors and shutters; locks and latches for securing the doors and gates; and hooks and bricks for the fireplaces. Bricks and nails are the most common artifacts of this class. The general absence of hardware may be attributed to the collection of iron that was made just prior to the fort's abandonment in 1781 (Lauber 1932 p. 582).
If any generalization can be made about the building hardware it is that there was a lack of uniformity. This may be interpreted as a result of the low level of technology in America, no one manufactory having been capable of supplying all the needs for a particular item, even in such a small area as the Mohawk Valley. As a consequence, hardware was supplied by a number of small forges and may even have been scavenged from existing buildings in the Valley. Colonial furnaces produced 14 percent of the world's iron supply at the outbreak of the American Revolution (Ransom, 1966, p. 6). Most of this iron went to England to be made into finished products, but there were American forges capable of producing goods. Bridenbaugh (1950, p. 61) lists eight iron furnaces, one steel furnace, 42 forges, one plating mill and one slitting mill for New Jersey alone in 1767. We know that the Type 4 axes discussed under the section entitled "Tools" were invented and made in America. Prior to the war most nail rods were imported to the Colonies with the finished nails turned out as a sort of cottage industry (Clark, 1929, p. 222). The war forced Americans to produce their own finished products although conscription and the unsettled conditions greatly lowered the productivity of the furnaces (Ransom, 1966, p. 6). Robert Erskine was probably the best known American iron manufacturer in the Revolution (Ransom, 1966, p. 36). Except for cast items such as cannonballs, shot and shells, he was shipping most of his pig iron to the American army where it was apparently being forged into tools and hardware as needed (Clinton, 1900, Vol. I, pp. 225-226, p. 666).
Five series of metal fasteners were used at the fort for jointing wood and metal to other woods and/or metals. These were: nails, staples, screws, rivets and bolts. Nails were the most common because they were the easiest to manufacture and use. Essentially, rivets were nails with a large head at both ends of the shank; they were apparently used in place of a clinched nail.
Bolts and nuts were sometimes used when clinched nails or rivets were not strong enough (Mercer, 1960, p. 248). The advantage of the bolt and nut was that adjustments could easily be made by tightening or loosening as the need arose. More initial work had to be done when using a bolt because a hole had to be drilled. Many blacksmiths did not have the dies to make bolts and nuts and, although various inventions in the 17th and 18th centuries increased the production of these items, they were not common (Mercer, 1960, pp. 250-251). The same limitations applied to screws.
Approximately 24,600 wrought iron nails were found in pre-1781 context. Of these, 8,005 were measurable and the other 16,600± were fragmentary. The military occupation of the site as a whole ended in 1781, nearly a decade before cut nails were introduced (Nelson, 1968; Noel Hume, 1970, p. 252). Therefore, only hand wrought nails will be discussed in this report, although cut nails were found in post-1781 contexts. Of the 8,005 measurable wrought nails, 871 may not have been associated with the fort since they were found in association with 19th century material or in disturbed areas.
A few of the wrought nails were well preserved because they had been burned, but most were badly corroded. Because of the great number of specimens, only a small number were treated for preservation. These were not randomly selected; we only attempted to clean fairly well preserved or unusual specimens. This precluded any attempts to make any percentage studies of the functional types of nails based upon head, shank and point shape.
The lengths of a large sample of nails were measured to the nearest 1/8 inch and plotted on a graph. They did not fit into the modern penny sizes, but clustered into the following inch-range groups: 01, 11-1/2 1-1/23, 33-1/2, 3-1/24, 44-1/2, 4-1/25, 56, 67, 78, 89 and 910. The 1-1/23 inches category spanned the 4d-10d sizes, but there were as many specimens falling between the pence sizes as in them, and the curve was a normal distribution (fig. 37). Few specimens were found above 6 inches in length and the upper divisions were arbitrary.
The barracks and guardhouse were frame structures with board siding, plank floors and shingle roofs. These buildings should have had a larger number of nails than the casemates which were constructed of logs with plank floors. Floor planks were probably the thickest pieces, about 1 to 2 inches thick. Most of the framing was probably mortised or dovetailed in some fashion, which explains why 85 percent of all measured nails were from 1/2 to 4 inches long (table 1).
Table 1. Distribution of iron nails at Fort Stanwix by
length in inches. (Ratios given are the numer of nails per 10 square
feet of excavated area withn the structure; U = unmeasurable
specimentes; ST = subtotals; T = totals.)
The barracks were constructed more than once. The British built the first barracks in 1764. These burned in 1774. The Americans rebuilt on the same locations in 1776. In 1780, as noted, the guardhouse burned and part of the west barracks was pulled down to prevent the spread of the fire. Both were rebuilt shortly after the fire, and in 1781 the fort burned again. Since the majority of the nails found were in the fill of the cellars, the nails probably were from the barracks in which the cellars were located. One lens in Feature 60 (Level II) of the west barracks contained 2,145 nails, many of them bent. The debris probably came from an area of about 1,200 square feet, or half the barracks. The same may be true to a lesser degree of all the cellars. After the barracks burned in 1774 and 1781, debris from them may have been used to fill the cellars in order to level the parade ground. An effort was made to plot all the nails found in the excavation that might have been in situ but this proved a futile endeavor, as no pattern could be discerned.
All the common types of hand wrought nails of the period described by Nelson (1968, p. 6) are present (fig. 38). All the types include both sharp and flattened points.
Rose-head (fig. 38a). Rose-head nails appear in all sizes but not all specimens of this type have rounded heads; many were flat with only short slopes near the edges.
T-head (fig. 38b). T-head nails also appear in all sizes. Some were obviously rose-head nails with two sides flattened to form the "T" but the large spikes have flat heads and are not reshaped rose-heads.
L-head (fig. 38c). The L-head nails appear in all sizes. However, most are 3 inches or shorter in length as they were used primarily as finishing nails (Nelson, 1968, p. 6).
There are no records of nails being received or requested by the garrison. A blacksmith was listed as one of the artificers during the American occupation, but he probably could not have made enough to keep up with the demand. Therefore, most of the nails found probably came from the lower Mohawk Valley area. It is known that pig iron was being shipped from American furnaces to the army on the Hudson where it was presumably forged into tools and hardware (G. Clinton, 1900, Vol. I, pp. 225-226). Prior to the Revolution most nails were probably imported from England (Nelson, 1968, p. 2).
Forty-five metal staples (fig. 38) of various lengths and thicknesses were found scattered widely around the site with no large clusters in any particular area (table 2). Most of the staples were probably made from nail rods, as 21 are square in cross section, 15 are rectangular and the remaining nine are round. All the staples, whether made from a round or four-sided rod, were pounded either to a tapered point or to a chisel point. The staples varied in length from 5/8 to 6-5/8 inches, with the majority ranging from 1-1/2 to 3-1/4 inches, the widths falling primarily in the 3/4 to 1-1/2-inch range. The largest was 1/2 inch in diameter and the smallest was 3/16 by 1/4 inch while most were in the 1/4 to 3/8 range.
Table 2. Distribution of staples, screws, rivets, nuts and bolts, and washers.
Square cornered (fig. 38e). There are nine of these made from four-sided rods.
Round cornered (fig. 38d). U-shaped staples were made from both round and four-sided rods. This was contrary to the findings and conclusions at Fort Ligonier (Grimm, 1970, p. 147) where all U-shaped staples were round in cross section. Some staples show wear in the interior of the "U" but not on the sides, suggesting that they were driven into the ceiling and objects were suspended from them.
Six round-topped staples have both ends clinched, some with the ends in the same direction and some in opposite directions. None of these could have been driven into a plank more than 1-1/2 inches thick if they were to be used to suspend objects.
On the east scarp (Level II) there was found a U-shaped bracket with out-turned flat ends pierced for attachment to a wall. Hanging from this are two chain links, 2 inches long and 1/2 inch wide, and a 3-1/2 inch long wire with bent ends made to attach it to the chain and provide a hook end. The function of the object is unknown. It may have served to keep a chest lid from opening more than 90 degrees. The U-shaped portion and wire are round in cross section and 5/32 inch in diameter, while the chain links are rectangular and measure 5/32 by 3/32 inch. Two more U-shaped brackets were found in the east barracks (Level II) and the sally port (Level III). This was the only chain found at Fort Stanwix.
One unique fastener was found in the southwest bombproof (Level III), a J-shaped nail probably used for suspending objects (fig. 38f). It is made from a rod 1/4 inch square; the shank is 3 inches long and all four sides taper to the point. The rod is bent 90 degrees to form the bottom of the "J" which is 1 inch long. The curved portion is 3/4 inch long and hammered flat. This may have been used as a rafter spike. Its shape is not similar to one illustrated in Watson (1968, p. 12) but corresponds very well to pitons used as climbing equipment, even though there are no cliffs close by.
Two wood screws were found (table 2). One found in the east casemate (Level III) is 1-1/4 inch long, 1/4 inch in diameter and tapered. The head is oval with a slot and the threads stop 1/2 inch below the head. The other, found in the sally port (Level II), is probably identical but the tip and part of the shank are missing.
Rivets were a compromise between nails and bolts (fig. 38). They were easier to make than nails because they had blunt ends, but, as with bolts, holes had to be bored through the objects to be fastened before rivets could be used. A rove, or washer, was often used with a rivet if two pieces of wood or other soft materials were being joined; the rove could be omitted if the material was hard (Mercer, 1960, p. 246). The distribution of rivets is shown in table 2.
The rivets were made from nail rods and round rods of several sizes, and were made to order rather than to a set size standard. The tip was not pointed, but often was split so that it would spread more easily when struck (Mercer, 1960, p. 246). Roves were also made to order, their size dependent upon the diameter of the rivet and material of the objects being joined. In general, the ones found are larger than the heads, but only one half as thick and are cut from flattened sheets of iron. All rivets found exhibit rose-heads of one variation or another. They range in length from 2 to 3-3/4 inches and in diameter from 3/16 to 1/2 inch.
Type 1 (fig. 38g). The heads of 15 specimens were formed by enlarging the end of the rod. Roves are still attached to 12 specimens.
Type 2 (fig. 38h). One specimen has a separate head welded to the shank.
Nuts and Bolts
A total of 15 bolts and 11 nuts, nine of which are still attached to bolts, were recovered (fig. 38). All except one were found within the limits of fort buildings (table 2).
Type 1. Bolts made from round rods, the shank hammered square above the threads to the head.
Variety a. Flat, square heads (fig. 38k). The heads of eight specimens are slightly larger on all sides than the shank. On some, the underside of the head is at a 90-degree angle from the shank; on others the head tapers down to the size of the shank which has a diameter from 1/4 to 5/8 inch. The shanks are approximately 5 to 6 inches long, approximately one-fourth of which is threaded. Three smaller specimens came from the north casemate (Level II), east casemate (Level III) and southwest bombproof (Level V). They are fragmentary pieces, missing parts of the head or shank, or both. All are 1/4 inch in diameter. These average slightly thinner than Fort Ligonier bolts (Grimm, 1970, p. 51) but just as long.
Variety b. T-head slightly convex rather than flat. One specimen is 5 inches long, shaped from a rod 5/16 inch in diameter. The head is 1-1/2 inches long and 5/16 inch wide. The tip is tapered. The threads began 1/4 inch above the tip, and continued for a distance of 1-1/4 inches.
Type 2. U-bolt (fig. 38i). One U-bolt has a 5/16 inch diameter round shank pounded flat with squared corners to form the "U." The "U" is 3-1/2 inches long and 1-1/4 inches wide with a 1-inch threaded section on both arms. One nut is still attached. It came from the north casemate (Level I).
Type 3. L-shaped (fig. 38m). Two bolts are modified pintles; at least the end of one specimen was serviceable as such. One was fashioned from a rod 1/2 inch in diameter. The hinge pin, at 90 degrees from the shank, is 3/4 inch long, the shank is 8 inches long and has two distinct portions. The half toward the hinge pin was beaten flat and exhibits two holes drilled 1 inch and 3 inches from the pin to receive nails or other fastening devices. The other half of the shank is round and the end is threaded for 1-1/2 inches of its length. A washer 1-1/4 inches in diameter and a nut are attached. They were found in the west barracks (Level I) and the southeast casemate (Level I). These may date post-1781.
Type 4. Hemispherical heads and round shanks.
Variety a. One-piece construction (fig. 38j). The head is approximately 3/4 inch in diameter and flattened on the sides to facilitate turning. The shank is 1/4 inch in diameter. It is broken, and the original length cannot be determined. Threads run the entire length of the shank to the base of the head. The specimen came from the sally port (Level II).
Variety b. Two-piece construction (fig. 38l). The shank was welded onto a preformed hemispherical head. Both specimens are missing part of their shanks, and have no threads on the remaining portion. Their identification is, therefore, tentative. Their heads both measure 1-1/4 inches in diameter; one shank is 1/2 inch in diameter, and the other 1/4 inch in diameter. The larger specimen is from the sally port (Level II) and the smaller from Feature 60 (Level II) in the west barracks.
Category 1, nuts. Two nuts, both 1-1/4 inches square, with flat top and bottom surfaces were found, one in the east barracks (Level I) and the other on the east scarp (Level II). They are 1/4 inch thick and one fitted a 9/16 inch diameter bolt and the other a 3/8 inch diameter bolt.
Eight flat metal washers were found (fig. 39a). One from the west barracks (Feature 63, Level III) is a 1-inch square piece of sheet metal with an irregular hole cut in it; the remainder are doughnut-shaped. The smallest specimen is 3/4 inch in diameter with a 5/16 inch hole and may be the only one to have been stamped out. All the other circular washers are irregular in thickness and symmetry between the outside and bore rims. These measure from 1 to 4-1/2 inches in diameter, 3/8 to 1-1/4 inches in hole diameter, and 1/16 to 1/2 inch in thickness.
These were used to reinforce and/or fasten two pieces of wood. Some may have been used on wagons or gun carriages.
Three L-shaped braces were found, one in the south ditch (Level XIII), one in the ravelin (Level II) and one in Feature 3 (Level IV) in the west barracks. They are strips of iron bent 90 degrees, one leg of which is two to three times longer than the other, and with one or two attachment holes (fig. 39h). They range in length from 4 to 53/8 inches and from 1/8 to 1/4 inch in thickness.
A brass strap was found on the east scarp (Level I) . It is 8 inches long, 7/8 inch wide and 3/16 inch thick. A hole has been bored 1/4 inch from each end and partially countersunk.
A large wrought strap was found in the northwest bombproof (Level II). Each end is split into two flat prongs approximately 3 inches long, 1 inch wide and tapering from 1/4 inch down to 1/8 inch thick at the tips. The total length is 29 inches and the main stem is 1-1/4 inches wide and 1/2 inch thick. Two countersunk holes are bored in each prong for attachment and one screw was still in a hole.
A wrought iron strap found in Feature 57 (Level III) in the west barracks may have been a reinforcing band around a timber up to 4-1/2 by 5 inches in diameter (fig. 39i). The strap is 1-3/4 inches wide and 3/16 inch thick. The band is broken so the length cannot be determined. The ends of the straps were bent away from the beam and were either riveted or bolted together or to another beam. One U-shaped bracket found in the south ditch (Level X) was a wrought iron strap 5-1/2 inches long, 2 inches wide and 3/16 inch thick (fig. 39j). The arms of the "U" are 1-1/4 inches long and the bottom 4 inches long. Two attachment holes are bored in each arm approximately 1/2 inch from the ends, while one hole, 3/4 inch in diameter, is in the approximate center of the bottom. This may have been a pivot plate for a light, four-wheel wagon.
One support with a raised center found in the east casemate (Level II) was a metal strap with a squared U-shaped offset (fig. 39g). It is 10-1/4 inches long. Each arm is approximately 4-1/4 inches long, while the offset is 1-1/2 inches long and 3/4 inch above the arms. The strap is 1-1/2 inches wide and 1/4 inch thick. Two countersunk holes are bored through each arm and another through the center of the offset.
Hinges were nailed or screwed to a shutter or door that swung open (see also the section entitled "Pintles"). The size and shape of the hinge depended upon the size, shape and weight of the object to which it was attached and also whether the hinge was to add a decorative effect (table 3).
Type 1. Strap hinges that worked in conjunction with a pintle. One end of the strap was wound around a mandrel and welded to form an O-ring that fitted over the pin of the pintle.
Variety a. Flat, with straight sides for half the length, then tapered to a point. One specimen measures 8-1/4 inches long, 1-1/4 inches wide at the pin, 1/8 inch thick and has two attachment holes 1-1/2 and 4-1/2 inches from the pintle end.
Variety b. Similar to Variety a, with the exception of an expanded disc end. They range in length from 6-3/4 to 11 inches, width from 1 to 3/8 inches and are 1/8 inch thick (fig. 40f) . Each hinge has three holes for attachment, two along the strap and one at the tip.
Variety c. Similar to Variety b, with the exception of a spear-shaped tip rather than a disc. These range in length from 10-1/2 to 18 inches, width from 1 to 1-1/2 inches and have a thickness of 1/8 inch (fig. 40c, e). The number of attachment holes varies from three to six; there is no correlation between the length and the number of holes. One specimen from the northwest bombproof passageway (Level III) has five nails still attached. They were clinched through wood 1-1/2 inches thick.
Variety d. A bulbous expansion in the strap approximately 1 inch from the pintle end, then tapered gradually to a point. Specimens measure 11-1/2 to 16 inches in length, 1-1/4 inches in width on the strap, 2-1/4 to 3 inches on the expansion and 1/8 inch thick (fig. 40a). Each has five attachment holes, two near the outer edges of the expansion and three along the midline of the strap. The holes extend from the pin ring along two-thirds of the length of the strap and there are no holes in the last one-third of the strap.
Variety e. Parallel sides along the entire length and the end split into two prongs. One prong continues in an upward curve ending in a point, while the other prong curves downward to form a hook pointed back toward the strap (fig. 40d). This specimen is missing the pintle ring and a short section of strap. The remaining portion measures 18-1/2 inches long, 1-3/8 inches wide and 1/8 inch inch thick. Six holes are bored approximately 2-1/2 inches apart along the strap. It came from the southwest casemate (Level II).
Type 2. One-piece strap hinge with a fixed pin.
Variety a. Straps shaped like isosceles triangles with the widest part at the pin. About one-third of one section is missing. The projected length is 3 inches; width, 5/8 inch; and thickness, 1/16 inch. Two holes were bored near the pin in one section. One hole was bored approximately in the middle of the other section. This came from the north casemate (Level I). It may be post-1781.
Variety b. Rectangular straps (fig. 39d). One specimen measures 3-1/2 inches long, 1-1/8 inches wide and 1/8 inch thick; the pin being 3/16 inch in diameter. Five nails, the longest being 1-3/4 inches, were still through it. It was found in Feature 60 (Level II).
Type 3. A special hinge designed to work with the Type 1c Pintle to hinge the drawbridge (fig. 41a). One was found at the bottom of the south ditch (Level XIII) where the bridge once stood. It was driven into a piece of wood like the pintle instead of being attached by nails or bolts. The shank is wedge-shaped with a 1-inch diameter hole bored through the thickness of the shank, 5 inches from the tip; it is barbed on two opposite edges to grip the wood. It measures 15-1/2 inches long, 1-7/8 inches wide, 1-1/8 inches thick and has a ring diameter of 2 by 3 inches.
Type 4. An L-shaped hinge designed to work with a pintle (fig. 39f). This has a vertical leg 3-5/8 inches long, 5/8 inch wide and 1/8 inch thick with three attachment holes. The 3/8-inch-diameter pintle ring is at the end of the horizontal leg which is 1-1/4 inches long, 5/8 inch wide and 1/8 inch thick, with no attachment holes. It came from the southwest casemate (Level II).
Type 5. Brass box hinge (fig. 39e). The patterns were stamped out of thin sheets of brass and folded over to form one half of the hinge. There were two patterns, differing only at the fold, so that the two parts would interlock and a pin, a separate piece, could pass through. Each section has three attachment holes in a triangular pattern similar to modern box hinges. It came from Feature 57 (Level III).
Type 6. Butterfly hinge. The narrowest part of the hinge is at the pin. The two sides flare out like the wings of a butterfly, and measure 1-1/2 and 2-1/2 inches maximum width, 3/4 and 1-1/4 inches minimum width, 2+ and 3+ inches in length and 1/16 and 1/8 inch in thickness. They came from Feature 58 and the sally port (Level II) respectively.
Type 7. Hook and eye hinge (fig. 39c). One half of the hinge is flaring, and the other half is a tapered strap with a hook attached through the eye. The eye section is 4 inches long, 7/8 to 2 inches wide, and 1/8 inch thick. The strap is 3-1/2+ inches long, 3/4 inch wide, tapering to a point, and 1/8 inch thick. The flaring half has three attachment holes and the strap, two. It came from Feature 76 (Level II).
Category 1. Three specimens with their tips missing belong to Varieties 1 a, 1b, or 1 c. They were found in the ditch (Level XV), north casemate (Level II) and a cellar Feature 56 (Level III) in the west barracks.
Category 2. Seven pintle rings with 1 inch or less of a Type 1 strap were found. They came from widely separated areas. The range of the diameters of the rings of all Type 1 hinges is from 3/8 to 7/8 inch which corresponds well to the range of diameters of 1/4 to 3/4 inch for the pins.
All the pintles and hinges, except the brass box hinge, were forged from wrought iron. The only unusual specimens are the two that we presume belonged to the drawbridge which were made by special design and the Type 7 hinge. The remainder are typical of that period and were made in large numbers.
All Type 1 hinges were capable of carrying considerable weight. They probably were used for doors, heavy shutters and lids to large storage boxes. Types 2 through 6 are rather light, and were probably used on cupboards, trunks and small chests for personal effects. Type 7 is quite unusual, its use is unknown.
Pintles were attached to the immovable frames of doors or windows to support hinges (table 3). They were made of two constituent parts: the shank driven into the frame, and the pin on which the hinge pivoted. All but one of the pintles found at Fort Stanwix correspond to Stone's Series B at Fort Michilimackinac (L. Stone, 1970, p. 487).
Type 1. Two-piece construction with the shank wrapped around and welded to the pin (L. Stone's Series B, Type 1; L. Stone, 1970, p. 488). The pin was made from a round rod and the shank from a rectangular strap forged to the desired shape.
Variety a. Shank with square cross section at the pin (Stone's Series B, Type 1, Variety c; L. Stone, 1970, p. 488). The top and two sides are tapered to the tip (fig. 40i). These range in length from 4-1/2 to 7-1/2 inches and are 1/2 to 3/4 inch square at the pin. The pins are 1/2 inch in diameter and protrude 1-1/2 to 1-3/4 inches above the shank.
Variety b. Shank rectangular in cross section. The thickness is uniform from in front of the weld to the tip; the top tapers from the weld to the tip (fig. 40j). The pin is 1/2 inch in diameter and beaten to a rectangular shape where the shank is welded to it. It protrudes 1-3/4 inches above the shank.
Variety c. Very large (fig. 41b). This specimen was found in the southeast casemate (Level II) near the area of the drawbridge. Judging from its size and location, it probably hinged the drawbridge. The shank is 13-1/2 inches long and flares in width from 1-1/8 inches at the pin to 1-3/4 inches at the tip, and tapers in thickness from 1-1/4 inches at the front of the pin to a point. A hole 1 inch in diameter was drilled through the thickness of the shank 3 inches from the tip. Barbs were cut on the edges to help keep the pintle from sliding out of the wood. The pin is 1-1/2 inches in diameter and rises 2-1/2 inches above the shank.
Type 2. One-piece construction (Stone's Series B, Type 2; L. Stone, 1970, p. 488). All are fashioned from rectangular rods. One end of the rod is hammered to a circular shape, then bent 90 degrees to form the pin. Varieties a and b are beveled at the base of the shank below the pin.
Variety a. Shank rectangular in cross section. The maximum size is at the juncture with the pin, and it tapers on the top and two sides to the point (fig. 40k). One specimen from the bridge area (Level I) is still attached to a hinge. The length of the shanks range from 3-1/4 to 6 inches; maximum widths range from 1/2 to 1 inch, maximum thickness ranges from 3/8 to 1-1/2 inches; pin lengths range from 1-1/4 to 1-1/2 inches; pin diameters are the same measurements as the maximum thickness of the shank.
Variety b. Shanks that were beaten flat. The top of the shank tapers to make the point, while the sides and bottom are parallel (fig. 40h). These range in length from 4-3/4 to 5-3/4 inches; in maximum width from 1/2 to 5/8 inch; in maximum thickness and pin diameter from 3/8 to 1/2 inch, and in pin length from 1-1/8 to 1-3/4 inches.
Variety c. Very crude with a blunt point (fig. 40g). The tip is missing but there is no taper to any of the shank; consequently, it probably had a blunt end. The one specimen is over 6 inches long, 1/2 inch wide, 3/4 inch thick and has a 2-inch-high pin. It came from the sally port (Level II).
Type 3. Two-piece construction with the pin welded to the shank (fig. 40i). The pin of this specimen is 1/2 inch in diameter and 2-1/4 inches long. The shank was made from a square rod 1/2 inch thick. Part of the shank and point are missing, but it appears that the top and sides taper to a point.
Type 4. One-piece construction with the shank pounded flat to a triangular shape (Stone's Series A, Type 1; L. Stone, 1970, p. 487). This pintle was attached to the frame with three nails or screws (fig. 40b). The pin is offset from the shank by 1/2 inch and is 3/8 inch in diameter and 1-1/2 inches high. The tip of the shank is missing but its projected length is 3-1/2 inches. It is 1 inch wide at the pin, about 2 inches wide at the tip and 1/8 inch thick. It came from Level III in the northeast bombproof.
The hinges that were attached to the pintles were so constructed that all were capable of supporting doors to structures. Except for the Type 4 pintle, they would appear to have been driven into wood. Generally, a hole slightly smaller than the pintle was bored in the wood and then the pintle driven in (Orville Carroll, personal communication). This is particularly true of the large pintle for the drawbridge. In addition, a hole per pendicular to the length of the drawbridge pintle had to be bored to pass a pin through the hole in the shank to lock it in.
Two type of hasps were found: a two-part hinge type and a one-piece hasp secured by a staple (table 3). The one-piece specimens were more crude in appearance and were probably used to secure doors and shutters rather than box or chest lids.
Type 1. Hinged (fig. 42c). This elaborate specimen is made of steel, and has the configuration of a spear point. The hasp is 4-1/2 inches long. 1-1/8 inches in maximum width where the eye was cut, and 1/8-inch thick. The eye is 7/8 inch long, 3/16 inch wide and 2-1/2 inches from the pivot. It came from Level II in the southwest casemate.
Type 2. Slit strap.
Variety a. Strap with cut slit (fig. 42a) . This is an iron strap 3/16 by 3/8 inch in which a large slit was cut and spread to form the eye, and a smaller slit cut at the other end for attachment. It is approximately 9 inches long, 1-1/4 inches in maximum width, and has a 90-degree bend. The eye is 2 inches long and 1/2 inch wide. A U-shaped fastener was used to secure the hasp and to give it the mobility necessary to slip on and off the catch. It came from the southwest bombproof (Level II).
Variety b. Two rods welded together (fig. 42b). These were constructed of two square rods bent and welded together. The specimen from the guardhouse area (Level II) was made from 1/4-inch-square rods welded immediately in front of the eye. The eye is 2 inches long 9/16 inch wide. It was broken behind the eye. The second hasp was made of 1/2 by 1/4 inch rods. It also was broken just behind the eye. The elliptical eye is 2-3/4 inches long and 5/8 inch wide.
Three door pulls were found, one each from the southwest bombproof (Level III), northeast bombproof (Level I), and the ditch (Level XI). The first two have spear-shaped plates with a rectangular latch hole in one plate. One specimen was made from a round rod 1/2 inch in diameter (fig. 42d), and the other from a rectangular strip 1/2 by 3/16 inch (fig. 42e). The pull made from the round rod is approximately 8 inches long with plates 2-1/2 inches long and 2 inches wide. The handle is 3 inches long and offset 1-1/4 inches. There are three attachment holes in each plate. The specimen made from a strap is approximately 9 inches long with one plate approximately 2-1/4 inches long and 2 inches wide. The other plate has rusted away. The handle is 4 inches long and offset 1-3/4 inches. The plate has three attachment holes. The third pull, from the ditch, has semicircular plates. It is made from a strap 1/4 by 1/2 inch. It measures 7-7/8 inches long with plates 2 inches wide and 1-1/4 inches long. The handle is 4 inches long and offset 1-3/4 inches. There are three attachment holes in the plates.
Five door hooks without eyes were excavated (fig. 42f). Staples were probably used as eyes, and also to secure the hooks to either the door or frame. Square rods were used to form the hooks. One end was wound around a mandrel to form an eye, the shank has been twisted a number of times to give it a decorative effect, and the end opposite the eye is bent 90 degrees to form the hook, which is tapered to a point. Of the five specimens, two are from the north casemate (Levels I and II), two from the west barracks (Level I and Feature 60, Level II) and one from the east scarp (Level II). They range in length from 3-3/4 to 5-1/2 inches.
Stock locks were commonly used for doors during the 18th century. A large specimen from the guardhouse area (Level I) and one small lock each from the southwest casemate (Level II) and east scarp (Level II) were found. The large lock has a cover plate 7 inches long and 4 inches wide. The arms of the U-shaped spring are 3 inches and 3-3/4 inches long. The ward and bolt are missing. The small locks are fragmentary and badly corroded. They are approximately one half the size of the large specimen. No 18th-century pad locks were found.
In addition to the gates, all chests, trunks and probably some building entrances were under lock and key. Eleven whole and fragmentary keys were found (table 3). Three specimens are made of brass, seven of wrought iron and one of steel.
Type 1. A large and fairly complex web which extends in one direction from the pin.
Variety a. Solid pin with no collar (fig. 42g, h). A large specimen from Feature 76 is 7-1/4 inches long, and may have been used with a gate lock. The semicircular bow is 1-1/8 by 2 inches, the stem is 4-1/2 inches long, the web 1-1/4 by 1 inch, and the pin 1-3/8 inches long, extending 3/8 inch in front of the web. A smaller specimen from the sally port (Level II) is missing the bow; the stem is 1-1/2 inches long, the web 3/4 by 1/2 inch and the pin 1 inch long, extending 1/4 inch in front of the web.
Variety b. Hollow pin (fig. 42i). This key revolved on a pin in the lock. Two incomplete specimens were found, both about 3-1/2 inches long. One came from the southwest bombproof (Level II) and the other from the north casemate (Level I).
Type 2. Narrow web expanding in two directions from the pin (fig. 42k). One brass specimen has a heart-shaped bow with a round hole cut in the center. The key is 3-1/8 inches long, the stem 1-1/8 inches long with no collar. The pin is hollow, 5/8 inch long, and does not extend in front of the web. The web is 1/8 inch wide on both sides of the pin. It came from Feature 60 (Level III).
Type 3. No web. The stem has a specially shaped hollow pin (fig. 42j). Two similar specimens were found, one of cast brass from Feature 76 and one of iron from Feature 63 (Level II). The iron key has a longer stem but otherwise is identical to the brass specimen. The keys measure 2-1/2 and 3 inches in length; the bows are oval shaped, 1 by 1-1/2 inches. The stems are 3/4 and 1-3/8 inches long, ending in diamond shaped pins 5/8 inch long and 3/8 inch wide on a side.
Category 1. Five additional iron specimens are too fragmentary to type.
Window glass was found scattered over most of the site but clearly much was of 19th- and 20th-century origin. A total of 3,076 fragments were found in pre-1781 context. These range in thickness from .8 to 3.2 mm. The mean and standard deviation for this sample is 1.59 mm. and .4079 respectively. A "refined" sample restricted to the undisturbed cellar holes of the two barracks (644 specimens) has a mean thickness of 1.57 mm. and a standard deviation of .3736. There is a statistical significance at a level of .01 in the thickness of the glass between several of the cellars and trash dumps of the fort (table 4) but this cannot be explained in terms of time, spatial distribution, or the type of cellar in which the glass was found. John Walker (J. Walker, 1971, p. 78), has suggested that window glass in the 19th century was made thicker with the passage of time. The window glass at Fort Stanwix supports his conclusions but does not refine his method since there is obviously a minimum thickness for window glass which was still in use until ca. 1840 (J. Walker, 1971, p. 78). A conversion to means instead of ranges as used by Walker might enable us to extend the utility of the hypothesis but we lack the data from other sites necessary to develop this. By presenting our data in some detail we hope to stimulate such a development.
No clustering of glass fragments was found which might indicate the location of windows in the structures. Two panes, one in the north casemate (Level III) and one in the west barracks (Level II), were 7 inches along one edge and at least 3 inches in the other dimension. The Thomas DeWitt powderhorn (dated 1778), shows windows with cross pieces suggesting four panes of glass for all the buildings on the parade ground, but this may have been artistic license.
During the siege, St. Leger sent emissaries to the fort. These were "conducted blindfolded into the fort, and received by Colonel Gansevoort in his dining room. The windows of the room were shut, and candles lighted . . ." (W. Willett, 1831, p. 55). This room was probably darkened by shutters over the windows. Such shutters would have been necessary in the winter to conserve heat, and probably served incidentally to protect the glass from breakage during a siege.
At Fort Stanwix in 1758, the British had ". . . got 40000 Bricks ready to Burn for the Chimneys & propose another kiln of 100000" . . . (Abercromby Papers: Stanwix to Abercromby, Sept. 5, 1758). It is very probable that all the bricks used in constructing the fort were produced locally. There are no comparable American records to verify this, however. The bricks were fired an orange-red; a few were over-fired and glazed on one side. All were hand made, and contained a large amount of sand. X's were scratched on one side of four bricks, and another was impressed by the paw of a large dog.
Dimensions of the measured bricks: length, 6-7/8 to 8-1/2 inches, median 8-1/4 inches; standard deviation, .4575; width, 3-1/4 to 4-3/8 inches, median 3-7/8 inches, standard deviation, .1969; thickness, 1-5/8 to 2-1/2 inches, median, 2-1/8 inches, standard deviation, .1546. Although they average slightly longer and narrower than 19th-century bricks from the site, it should be noted that the difference is not statistically significant at a level of .05.
Bricks were concentrated in the cellars of the east and west barracks and on the parade ground side of fireplaces in the casemates. The latter were probably pushed over by the weight of the collapsing curtain walls, and the former were apparently dumped into the cellars while clearing the site since they were not found in the fill adjacent to the cellars.
All mortar found in the fort consisted of burned lime with a high percentage of sand. Color was generally a light cream white. Samples were saved for comparative purposes, and as models for the reconstruction. "Lyme-stone" was sent to the fort along with boards for the 1777 reconstruction (Gansevoort-Lansing Papers, May 14, 1777).
Only scattered parts of weapons were found (table 5). Most appear to have been broken and then discarded. Several pieces of brass furniture from guns had been burned in a fire, probably one of those that destroyed the fort in 1774 or 1781. Most of the brass furniture was marked on the back with engraved Roman numerals. The greatest concentration of musket parts was recovered in the west barracks. Sources used in identifying parts are Darling (1970), Neumann (1967) and Peterson (1968).
Table 5. Distribution of weapons, weapon parts and accouterments. (Ratios
given are the number of specimens per 10 square feet of excavated area within the
Sling Swivels. Nine sling swivels were excavated, one of brass from the ravelin (Level II), the remainder of iron (fig. 43a). Dimensions: length: 26 to 30 mm., average 29.5 mm.; width, 52 to 59 mm., average, 55.4 mm.; thickness, 4 to 5.5 mm., average, 4.5 mm. These were probably from British Land Pattern-like (Brown Bess) muskets.
Triggers. All specimens were of iron (fig. 43b), and all but one were fragmentary. The complete specimen is 53 mm. high, 25 mm. long and 14 mm. wide. Two others are 14 mm. wide also. These were probably from British Land Pattern-like muskets.
Trigger Plates. Two brass trigger plates with slots 35 mm. long, off-center to the right (fig. 43c) are 19 mm. and 18 mm. wide; one is 9.1 cm. long. One specimen, from Feature 57 (Level IV), is marked on the back with "VIII" and the other, from Level II on the east scarp, "VIIII." These would be from British Land Pattern muskets.
Trigger Guards. All were made of brass and so fragmentary that complete measurements are not possible (fig. 43e). Four specimens are marked: one from Level XI below the bridge with two stamped "W's" and an "XIIIII"; one from Feature 69 (Level II) with a stamped "Y"; one from the east casemate with "VIII" and "VIIII" and one from Feature 57 (Level III) simply "VIII." All were from British Long Land Pattern muskets.
Lock Plates. All were made of iron and contained the firing mechanism of flint lock firearms. None are marked and all are badly corroded.
Type 1. One specimen (fig. 43g) is 14.5 cm. long and 2.8 cm. high with a flash pan 2.5 cm. wide. It has five screw holes and still retains the cock with a French gunflint, sear, sear spring, tumbler and frizzen spring. Judging by its size, it was probably from a large pistol or American rifle. It was found in Level I of the north casemate.
Type 2. Three specimens, all with 10 screw holes, are 17.2 cm. long and 3.2 cm. high with a flash pan 3.0 cm. wide (fig. 43h). All had been stripped of parts. The curvature suggests early British Long Land Pattern muskets, but these may have been American copies. They were found in the north casemate (Level II), Feature 69 (Level II) and the sally port (Level II). Two found at Fort Ligonier (Grimm, 1970, p. 91) are 18.1 and 18.3 cm. long and are more likely from British issue weapons.
Cocks. All specimens are of iron; only three are complete. They held the gunflint which struck a steel frizzen to ignite a priming charge in the flash pan. All are from British Land Pattern muskets, except Types 2 and 4, which are probably pistol or rifle cocks.
Type 1. Four specimens have a height of 90 mm.; jaw widths of 26 to 27 mm. and a jaw screw length of 49 mm. (fig. 43i). Three jaws have hollows and two have teeth for gripping gunflints. They were found in the east casemate (Level I), in the guardhouse area (Level II) and in Features 48 and 76.
Type 2. Two specimens have a height of 70 mm., jaw widths of 18 to 19 mm. and a jaw screw length of 36 mm. (fig. 43j). One was broken and repaired by brazing. They were found in the guardhouse area (Level II) and Feature 69 (Level II).
Type 3. One fragment with a jaw screw length of 45 mm. was recovered. This specimen had been broken and repaired by brazing. It came from the sally port (Level II).
Type 4. One specimen with a height of 84 mm., jaw width of 23 mm. and jaw screw length of 42 mm. (fig. 43g). This is the cock on the lock plate found in the north casemate (Level I).
Frizzen. One steel frizzen fragment was found in the ravelin (Level I) (fig. 43k). It is 32 mm. wide, and was probably from a British Land Pattern musket.
Frizzen Springs. Two steel frizzen springs were recovered. One from Feature 50 measurements 7.0 cm. long. It is probably from a British Land Pattern musket; the second is 4.7 cm. long and on the lock plate found in the north casemate. These specimens held the frizzen in place over the flash pan to receive the blow of the gunflint.
Main Springs. Two steel main springs 82 mm. long were found (fig. 431). These forced the cock forward when the trigger was pulled. The specimens are probably from British Land Pattern muskets. They were found in the southwest casemate (Level I and Feature 48).
Side Plates. All were made of brass and all were fragmentary. They were mounted on the stock opposite the lock plate and were fastened to it with screws.
Type 1. A thin excised brass plate, probably from a pistol, came from Feature 3 (Level III) (fig. 43m). It is over 10 cm. long but broken.
Type 2. Ten plano-convex "serpentine" plates, probably with two screw holes (fig. 43n). Five are marked on the back. One specimen from the north casemate (Level I) has a broad arrow and the number "1" stamped on it, and the engraved Roman numeral "XV." Two others from Feature 57 (Level III) were stamped "XIII." One, from Feature 57 (Level III), has a stamped "G," and part of a Roman numeral is stamped on another from Feature 60 (Level II). These were from British Land Pattern muskets.
Type 3. One specimen is a flat plate, probably with two screw holes (fig. 43p). This has been identified as from a 1751 Spanish infantry musket by an arms collector, but we have doubts about this identification. It came from Feature 57 (Level II).
Escutcheon Plates. Brass plates mounted on top of the stock and fastened with screws to the trigger guards (fig. 43d). One measures 62 by 28 mm. and is engraved on the front with "L x C". On the back it has an "XIII" and a stamped "C". It came from Feature 57 (Level IV). Another, from the sally port (Level II). measures 65 by 24 mm. and is engraved on the front with "17 Rt","C", "61". A stamped broad arrow appears on the back. This specimen is probably part of a musket taken from the British 17th Regiment of Foot in 1779 at Stony Point, New York since it was in a post-1776 context. Two companies of the 17th were stationed at Fort Stanwix in 1763 (Gage, 1931, p. 210). The third specimen, from the east casemate (Level II), is partially melted, but has a Roman numeral on the front. These were from British Land Pattern muskets.
Screws. Four brass screws, 24 mm. long, probably came from weapons. Two are 4 mm. in diameter and two are 5 mm. in diameter. The heads range in diameter from 10 to 15 mm. These were probably from British Land Pattern muskets.
Butt Plates. These specimens are all made of brass.
Type 1. Three plates made from sheet brass were nailed to the butt of the stock (fig. 44q). Length, 12.8 cm. (one specimen); width, 3.4 cm., 4.4 cm. and 5.0 cm.; thickness, 1 mm. These were probably American made. They came from the north casemate (Level III), Feature 72 (Level II) and the west ditch (backfill).
Type 2. Five specimens cast with a tang projecting over the stock, and held on by screws (fig. 44p). Length, 12.3 cm. (one specimen); width, 4.9 to 5.0 cm.; thickness, 2 to 3 mm. These were from British Land Pattern muskets. They were found in the east casemate (Level I), on the east scarp (Level II), in Feature 76, in the bakehouse area (Level I) and in the Campbell collection.
Nose Caps. Made of brass and fitted over the end of the stock near the muzzle of the weapon (fig. 43f). They were probably from British Land Pattern muskets.
Type 1. Two specimens are 22 to 25 mm. long. These had a single hole for mounting; one retained a brass rivet in place. They were found in Feature 57 (Level III) and the north casemate (Level I).
Type 2. One specimen, 23 mm. long, with two mounting holes and a groove on the underside for the ramroad came from Feature 57 (Level III).
Gun Barrels. These specimens were made of wrought iron.
Type 1. Two specimens were from .75 cal. smoothbore muskets. Both are fragmentary, but one from the north ditch (Level II) still retained the breech plug. It tapers from 3.2 to 2.3 cm. in diameter. The other, 2.3 cm. in diameter, is a short fragment with an iron bayonet stud brazed onto it 5.0 cm. from the muzzle. This came from Feature 72 (Level II). These were from British Land Pattern-like muskets.
Type 2. One specimen from Feature 76 is a .69 cal. smoothbore musket barrel with an exterior diameter of 2.3 cm. A mounting notch for a ramrod pipe is situated 20 cm. from the muzzle. The grain of this barrel makes a complete turn every 13 cm. One end was hammered flat and the other showed pounding from use as a tent peg. Origin is uncertain.
Type 3. One specimen from the north ditch (Level VIII) is a .66 cal. smoothbore pistol barrel 14.9 cm. long, with the breech plug intact (fig. 45c). It tapers from 2.9 to 2.0 cm. at the muzzle and has a mounting notch for a ramrod pipe 6.2 cm. from the muzzle.
Type 4. Two specimens are .44 cal. smooth bore musket barrels with an exterior diameter of 1.8 cm. One specimen from Feature 57 (Level III) has a notch for a ramrod pipe 11 cm. from the muzzle. One end has been flattened and a hole punched through it. The other came from the Campbell collection. These were probably American made.
Forward Band. A single iron forward band, probably from a French weapon, was found by Campbell.
Ramrod Pipes. These specimens secured the ramrod to the bottom of the stock and were fastened to the stock by pins. All were made of brass.
Type 1. Cast forward pipe with two mounting lugs. One specimen, 10.3 cm. long, (fig. 44d), was from a British Land Pattern musket with a steel ramrod. It came from the sally port (Level II).
Type 2. Cast forward pipes with one mounting lug. Four specimens were broken or partly melted, but are about 6.0 cm. long. These were probably from British Land Pattern muskets. They were found in the east barracks (Level I) the bake house hearth and Feature 57 (one in Level III and one in Level IV).
Type 3. Cast 2d pipes with one mounting lug. Nine specimens are from 3.1 to 4.0 cm. long (fig. 44b). These were from British Land Pattern muskets.
Type 4. Cast tail pipe with one mounting lug. Length, 74 mm. (fig. 44a). This was from a British Land Pattern musket and was found in the south ditch (Level XI).
Type 5. Sheet 2d pipes bent to form a mounting lug. Three specimens are from 2.8 to 3.7 cm. long (fig. 44c). These were almost certainly of American origin (Neumann, 1967, p. 96). They came from the north ditch (Level X), the sally port (Level II) and the southeast casemate (Level I).
Ramrod Tips. No iron "button-tipped" ramrods were found, but six brass tips for wooden ramrods were recovered (fig. 44e). Two from the sally port (Level II) and Feature 76 are marked with stamped broad arrows. All are truncated brass cones with an end brazed onto them. The following calibers were found: one .70 cal., two .64 cal. and three .63 cal. These undoubtedly served larger caliber weapons, probably .75 cal. British Land Pattern muskets.
Gun Worms. These specimens were made of iron and designed to extract lead balls from a musket to unload it.
Type 1. Single prong. One specimen from the north casemate (Level I) measures .31 cal.
Type 2. Double prong (fig. 44f). Two specimens measure .56 cal. and .62 cal. They came from the sally port (Level II) and the north casemate (Level I) respectively.
Vent Picks. These were used to clean the vent hole from the flash pan through the barrel of the weapon.
Type 1. Brass wire picks on brass chains (fig. 44g). Five specimens had a wire pick with a loop at one end to attach it to the chain. Three consist of chains only. One from the sally port (Level II) measures 14 cm. long from the tip of the pick to a loop that probably fastened to a button hole on a soldier's waistcoat.
Type 2. Cast brass picks (fig. 44h). Two specimens were cast or stamped with a drilled decorative finial. They are 9.2 cm. and 6.7 cm. long. One from the guardhouse area (Level II) has a heart-shaped finial and the other from the southwest casemate (Level II) appears to be a thistle.
Steel bayonet fragments were scattered over the entire site. All had triangular blades and were probably of a British pattern (Peterson, 1968, p. 85), including one from the northwest bombproof (Level II) which is stamped "US" on the blade (fig. 45a). Another from the west casemate (Level II) is stamped with what appears to be a rampant lion, but not the British East India Company mark (Darling, 1970, p. 54). The most complete blades are 36.5 cm. long (14.4 inches) with a maximum width of 3.0 to 3.5 cm. The sockets have an interior diameter of 2.2 to 2.5 cm., and none have a clasp to secure the bayonet to the musket. The blade edges of two fragments from Feature 69 (Level II) were hammered together until they touched, but the function of these is unknown.
These brass hooks were used to attach scabbards to a belt. Apparently they pulled out of the leather quite easily, as several were intact and leather was still attached to two specimens. A scabbard hook similar to Type 1, but made of iron, was found at Fort Ligonier, 1758-1766 (Grimm, 1970, p. 127). Type 3 was also the most common type at Fort Ligonier (Grimm, 1970, p. 127).
Type 1. Three specimens, 51 to 52 mm. long (fig. 44i), were attached by inserting a triangular projection through the leather. They were found in the west casemate (Level II), Feature 60 (Level I) and a disturbed area of the west barracks.
Type 2. One specimen from the north casemate (Level II) is 35 mm. long (fig. 44j). It appears to have been soldered on to another plate which was then attached to the leather.
Type 3. Twenty-four specimens (table 5) are from 47 to 57 mm. long with an average of 53.2 mm. (fig. 44k). They were attached by two rivets and a backing plate to the leather. The leather was 2 to 4 mm. thick.
Type 1. A folded brass sheath with a brazed seam and a finial in the end was identified as a scabbard tip. It came from the east scarp (Level II) and is 56 mm. long and 23 mm. wide.
Type 2. These specimens are made of solid brass and three varieties were recovered.
Variety a. Short solid post (fig. 441). Four of these specimens have the following dimensions: length, 21 mm.; diameter, 10 to 12 mm. One still had leather adhering to it. They were found in the southwest casemate (Level II), the east casemate (Level I), Feature 76 and the sally port (Level II).
Variety b. Long solid post (fig. 44m). Three of these have the following dimensions: length, 24 to 32 mm.; diameter, 9 to 11 mm. They came from the southwest casemate (Level II), the guardhouse area (Level I) and Feature 76.
Variety c. Hollow post (fig. 44n). Three of these have the following dimensions: length, 27 to 43 mm.; diameter 9 to 10 mm. The largest from the sally port (Level II) may have been a Type 2 specimen with a brass sheath around the post. The others came from the guardhouse area (Level II) and the east scarp (Level II).
These specimens were concentrated in the northwest bombproof (Level II) and the south ditch (Levels I and XV). They are made of wrought iron and none are identical (fig. 45b). All have two tangs between which the wooden shaft was seated. Iron rivets were driven through the tangs and shaft. Blade length: 17 to 29± cm., average, 20.7 cm.; blade width: 3 to 5 cm.; total length: 21+ to 38.5+ cm., average, 30.4+ cm.; spear shaft diameter: 2 to 3 cm. On August 8, 1777, Lt. Col. Willett and Lt. Stockwell each left Fort Stanwix to get help carrying "a spear . . . eight feet in length . . ." (W. Willett, 1831, p. 60).
No record has been found of flint knappers stepping forth to ply their trade, and no gunflints recovered have been made from native chert in Revolutionary War contexts. The journal entry indicates that gunflints were in short supply in the American army.
Gunflints have been divided into five main groupings based on method and locale of manufacture (Witthoft, 1967, pp. 12-49). These are: Aboriginal (American Indian), Nordic, Dutch, French and English. The Dutch gunflints also have been referred to in the literature as gunspalls. Morphologically, there was little difference in the method of manufacture of American Indian and Nordic gunflints, since both were bifacially chipped and square to rectangular in shape. The Dutch gunflints were struck individually from cores, and generally were plano-convex and D-shaped, with a prominent bulb of percussion on the dorsal surface. They were made in a variety of colors indicative of Riss glacial outwash material (Witthoft, 1967, pp. 25-26). French gunflints were generally of a honey-colored or blond translucent French flint, and were made from segments of long blades struck from a core. English gunflints were similarly made but more prismatic in shape and the flint graded in color from blond to black. Some of the lighter colored English specimens may have been made from imported French flint (Witthoft, 1967, p. 32).
A total of 3,826 gunflints were found at Fort Stanwix, 3,412 of these from a single deposit in Feature 69 (Levels I and II) in the east barracks. Of the total sample, 62.9 percent were Dutch and 37.1 percent French, with one English specimen found in a post-1781 context (Level I). It probably dated from the early 19th century.
Woodward (1960, pp. 29-39) has cited a number of standard sizes for gunflints including Pocket Pistol, Rifle, Musket, Long Dane, etc., for English gunflints, but we believe this was a late development and that prior to the 19th century there was no standardization into size categories. However, there was a wide range of sizes (especially French gunflints) and this needs to be further explored. More recently, Lyle Stone (1971) has proposed a formal classification based on such properties as manufacturing technique, shape, material and color. We applied this method to the Fort Stanwix sample, both as a means of describing that sample and to make comparison easier with the Fort Michilimackinac sample from northern Michigan.
The following terms are used to describe gunflints: The bed is the bottom surface; the edge is the striking surface at the front; the heel is the back of the gunflint opposite the edge; the face is the uppermost surface, usually parallel to the bed, and the length is the distance from heel to edge along the axis of the position of a gunflint in the jaws of a gun cock. Length measurements are recorded only for those specimens that do not appear to have been used. Lyle Stone (1971, pp. 11-19) has established length-width ratios for his types which were tested with our samples.
Basically, Stone's formal analysis was accomplished by classifying the Fort Michilimackinac sample into three series, A, B, and C, based on manufacturing techniques. Series A, corresponding to what we call French gunflints, was subdivided into four types based on their longitudinal cross section. Series B and C, which we call Dutch, had only a single type each but Series C, Type 1, was subdivided into three varieties based on color. Specimens that could be classified by series, but not by type because they had been modified or broken, were placed into categories within the series.
We found no objection to the method as presented, but we encountered one difficulty in its application. Many fragmentary specimens could have been included in either Series B or Series C categories. This problem has been solved by converting Series B and C into types within the same series. At Fort Michilimackinac, Series B gunflints accounted for only .2 percent of the total sample, while at Fort Stanwix they were 10.0 per cent of the total sample. Our larger sample shows the same color range as the Series C gunflints. This is in contrast to what Stone found with his much smaller sample (1971, p. 16), which were all of the darker variety (b). In this report we will refer to Series A as French, and Series B and C as Dutch, Types 1 and 2. Table 6 shows the break down of these types in the Dutch series by variety.
Table 6. Unmodified Dutch gunflints separated on the basis of color (a, light gray; b, dark gray to black).
We feel that the transverse flake scar which distinguishes the Dutch Type 1 specimens from those of Dutch Type 2 is a by-product of the removal of a previous gunflint from the nodule, or core preparation. Therefore, it is likely that both types could have been produced from the same core and there would be no temporal difference. A few examples of Type 2 exhibit flake scars on the bed at the edge or on the side of the gunflint.
The same situation exists for the French gun flints. The flake scars that characterize the four types set up by Stone (1971, pp. 11-16) could have all been produced on a single core. In view of the mass production methods of the 19th century, it seems unlikely that many blades were considered so badly struck to be of no further use, although in some cases there is sufficient data to show that the finished product was later sorted by size or shape (Woodward, 1960, p. 32; C. Smith, 1960, pp. 60-61). Although we feel that these types (not to be confused with the series) and varieties have little historical significance in terms of relative popularity, this is an unproven assumption, and they are retained for comparative purposes.
The following is a description of the types found and their frequencies on the site. A break down by area is shown in table 7.
Table 7. Distribution of gunflints. (Those found in the East Barracks include 2,145 Dutch and 1,267 French specimens from a single deposit.)
French, Type 1. Three transverse flake scars form a beveled front and back, and a face parallel to the bed. The heel is generally rounded off with secondary chipping (fig. 44r). Three hundred and two specimens (one chalk heel and 277 burned) have the following dimensions: length, 21.5 to 28.0 mm., average, 24.4 mm.; width, 23.1 to 42.1 mm., average, 28.36 mm.; standard deviation, 4.74; thickness, 4.6 to 12.2 mm., average, 6.80 mm.
Stone (1971, p. 11) presents a regression formula, based on 18 specimens, for estimating length from the known width of used gunflints of this type. We applied his formula to our 15 unmodified French Type 1 gunflints and found that the results were not predictable. The Fort Stanwix specimens average 2.4 mm. longer than the formula would have made them, with individual variations from -1.0 mm. to +5.3 mm. Since length, the second dimension produced by this formula, is based solely on the width, no new information was generated from its application. While length may have been a more critical variable (Stone, 1971, p. 12), there simply are not enough unmodified specimens for meaningful comparisons. During the process of manufacture, there should have been a consistent length-thickness ratio, but this was obscured by secondary shipping on the heel in the finishing process. We attempted a scattergraph of the width-thickness ratio, and concluded that the two variables were unrelated in the final form.
Plotting the distribution of the widths alone shows two distinct size clusters, with a breaking point at 31 mm. There are 105 specimens below 31 mm. in width, and 22 specimens larger than 31 mm. where the width could be measured. Four of the latter were used with strike-a-lights; the smaller specimens were not used in this way. Since most of these specimens were burned and some partly destroyed, it is not possible to make a definitive statement that the larger specimens were intended for use with strike-a-lights. We feel, however, that they are too large to have served as gunflints and they do exhibit wear. If they were not repeatedly struck at the same point they would not exhibit the deep notching that we consider diagnostic of strike-a-light flints. Thus, we tentatively add a second trait, large size, to our definition of strike-a-light flints. We do not mean to imply that some individuals did not occasionally use small gunflints with strike-a-lights. At Fort Ligonier, 1758-1766, (Grimm, 1970, p. 89) there were 18 French gunflints over 31 mm. wide and 12 less than 31 mm. wide, which suggests that the size of French gunflints was decreasing through time although the range was the same for both sites.
French, Type 2. Two transverse flake scars form a beveled front and a face parallel to the bed. The heel is rounded off with secondary chipping (fig. 44t). Dimensions of 415 specimens (one chalk heel and 393 burned) are: length, 23.9 to 34.0 mm., average 26.4 mm.; width, 22.0 to 40.0 mm., average, 27.06 mm.; standard deviation, 3.63; thickness, 3.7 to 13.5 mm., average, 6.74 mm.
As with Type 1, Stone (1971, p. 13) presented a formula for estimating length from width. The Fort Stanwix specimens average 2.7 mm. longer than they should have, according to the formula. A breaking point also occurs in this type with 22 specimens wider than 31.0 mm. and 194 narrower than 31.0 mm. All three of the observed strike-a-lights are of the larger size.
French, Type 3. Two transverse flake scars form a beveled front and back with no face. Heel usually shows secondary chipping (fig. 44s). Dimensions of 79 specimens (one chalk heel and 75 burned) are: width, 22.0 to 39.1 mm., average, 27.50 mm.; standard deviation, 3.36; thickness, 4.2 to 11.0 mm.; average 6.97 mm. No lengths could be measured This type also could be broken into two width size classes at 31.0 mm., with three specimens in the larger class and 33 in the smaller. The only strike-a-light in this type is in the large size.
French, Type 4. A single broad flake scar forms a face parallel to the bed. The heel shows secondary chipping, although on a used specimen it was sometimes difficult to distinguish the heel from the edge. Dimensions of seven specimens are: width, 26.7 to 37.1 mm., average, 31.92 mm.; thickness, 5.7 to 8.5 mm., average, 7.10 mm. No lengths could be measured. There are too few specimens to make any statement about size classes.
French, Category 1. This category includes those specimens which were identifiable as French gunflints, but could not be typed due to breakage, modification or burning. There are 610 specimens in this group; 596 were burned and nine used with strike-a-lights.
Dutch, Type 1. A single bulb of percussion on top of the heel, with a transverse flake scar across the back of the heel on a bevel. There is some secondary chipping on the sides and occasionally on the edge and heel (fig. 44u). Dimensions of 379 specimens (two chalk heels and 338 burned) are: length, 22.0 to 40.0 mm., average 31.18 mm; width, 25.0 to 40.5 mm., average, 33.09 mm.; standard deviation, 3.13; thickness, 4.0 to 12.3 mm., average, 8.65 mm. Unlike the French specimens, these gunflints appear to be a homogeneous sample. Although there is a broad width spectrum, no breaking point could be defined. Seven gunflints were modified by use with strike-a-lights, but none could be measured. It is our impression that the strike-a-light flints have the same size range as the gunflints. The Fort Ligonier (Grimm, 1970, p. 90) sample had the same range and average width as the Fort Stanwix sample. Although most of these specimens were burned, a few were not, and these show color variations ranging from light gray to black. The significance of these variations will be discussed in the following type but the data is presented here for the record.
Variety a. Gray to brown in color. Dimensions of 26 specimens (one used as a strike-a-light) are: length, 25.0 to 36.8 mm., average, 32.52 mm.; width, 28.0 to 39.0 mm., average, 33.3 mm.; thickness, 5.9 to 11.4 mm., average, 9.27 mm.
Variety b. Dark gray to black in color. Dimensions of 15 specimens (two chalk heels and three used as strike-a-lights) are: length, 22.0 to 40.0 mm., average, 32.00 mm.; width, 25.0 to 38.9 mm., average, 33.83 mm.; thickness, 5.2 to 12.3 mm., average, 9.74 mm.
Dutch, Type 2. A single bulb of percussion on top of the heel with secondary chipping on the heel, and some secondary chipping on the sides and edge. A few exhibit a transverse flake scar on the bed (fig. 44v). Dimensions of 1,411 specimens (eight chalk heel and 1,254 burned) are: length, 21.0 to 38.3 mm., average, 30.48 mm.; width, 19.5 to 42.1 mm., average, 33.09 mm.; standard deviation, 3.39; thickness, 4.2 to 13.2 mm., average, 8.38. Some of these gunflints could be divided into two color varieties.
Variety a. Gray to brown in color. Dimensions of 65 specimens (one chalk heel and seven strike-a-lights) are: length, 24.3 to 34.6 mm., average 32.5 mm.; width, 26.9 to 42.1 mm.; average, 33.7 mm.; thickness, 5.8 to 12.0 mm., average, 8.8 mm.
Variety b. Dark gray to black in color. Dimensions of 37 specimens (six chalk heels and seven strike-a-lights) are: length, 25.4 to 36.8 mm., average, 31.9 mm.; width, 31.1 to 42.0 mm., average, 35.5 mm.; thickness, 6.2 to 12.0 mm., average, 9.8 mm. Lyle Stone (1972, pp. 47-49) has suggested that the black Dutch series gunflints (Types 1b and 2b) date prior to the American Revolution and were most prevalent in the 1745-1765 period. Although the sample from Fort Stanwix is small, it seems likely that the black gunflints continued in use later than at Fort Michilimackinac. In the dump on the east scarp (Level II) and in the sally port (Level II) which date post-1776, they constituted 39 percent of the sample (of 33 specimens).
Dutch, Category 1. This category includes those specimens which are identifiable as Dutch gunflints, but could not be typed due to modification or burning. There were 622 specimens in this category; 617 burned and two were used as strike-a-lights.
English. Three transverse flake scars form a beveled front and back and a face parallel to the bed. Small bulbs of percussion appear on each side of the face and there is some secondary chipping on the sides. Only one example was found; length, 24.1 mm.; width, 28.1 mm.; thickness, 7.9 mm. The specimen was found along a 19th-century retaining wall in an area heavily disturbed by tree roots (redout, Level I). It is probably post-1781 in time.
Discounting the large deposit of 3,412 gun flints in Feature 69, there were remarkably few gunflints found at Fort Stanwix, considering the span of time it was occupied. In the Feature 69 deposit, all the flints were burned and concentrated in the fill of one corner of the clay-lined cellar hole and the adjacent ground surface. Perhaps they represent a deliberate collection of discarded gun flints, since nearly all appear to have been used. An unresolved question is why anyone would bother to gather up worn-out gunflints in the first place.
A series of wood or clay-lined cellar holes was situated under the east and west barracks. Two of the wood-lined cellars (Features 56 and 63) were intruded by clay-lined cellars (Features 57 and 64) and we suspect that the wood-lined holes were pre-1776, while those that were clay-lined date from the Revolutionary War period. Although the presence of gunflints alone is not proof of this hypothesis, no French gunflints were found in the wood-lined cellars (23 found were Dutch) and, with the exception of the large deposit in Feature 69, there were 22 Dutch and 22 French specimens in the clay-lined cellars.
Although there is some frequency variation between units as shown on table 7, this variation was not statistically significant for the size of the samples. Even the known post-1776 deposits do not differ markedly from the rest of site.
Comparative data on gunflints are scant. It has been suggested elsewhere (Hanson, 1970, pp. 51-58; 1971b, p. 109) that the relative frequency of gunflints could be used as an indicator to the age of a site. Table 8 lists selected late 18th-century sites and the percentages of Dutch and French Series gunflints reported as a test of this hypothesis. There is no doubt that French gunflints were supplanting Dutch specimens in the last half of the 18th century, and that this change was gradually increasing in pace toward the end of the century. The Fort Leboeuf sample, while small, serves to point out that this process was subject to local variation due to factors we cannot control at the present time. Given the known shift in popularity of gunflint series it should be possible to develop a regression formula such as Binford (1962) has done for pipe stems or South (1972) for ceramics. The main problem is a lack of large documented samples. At the present time we feel that Dutch gunflints first appeared about 1650 (Witthoft, 1967) and remained in use until after 1780. French gunflints first appeared at North American sites about 1735 (Stone, 1971) and remained in use until about 1820 (Witthoft, 1967). English gunflints appeared after 1780 (Witthoft, 1967). The increase in French gunflint frequency (at the expense of Dutch gunflints) started accelerating between 1765 and 1770 so that "pure" Revolutionary War sites should have a relatively high percentage of French gunflints to Dutch (over 30 percent) and no English gunflints.
Table 8. Comparison of sites by frequency percentage of gunflints.
Four lead gunflint pads were found in the guardhouse area (Level II), main gate area (two) (Level II) and on the east scarp (Level II) (fig. 45d). These are scraps of lead with a hole cut in the center and folded over the gunflints to hold them firmly in the jaws of a musket cock. The hole was cut to fit around the jaw screw.
Thirty-two solid iron cannon balls were excavated at Fort Stanwix (table 9). These were typed by weight. The only distinction that can be made between them is the presence of broad arrows on a few specimens. Mold markings appear to have no significance, although the variety suggests at least three sources of supply. These consist of the presence or absence of a mold seam, a depression on one side away from the seam, or a depression surrounded by a depressed ring on one side away from the seam. Some balls show none of these markings. The only correlation is that all the depressions with a ring around them are present on balls with mold seams.
Type 3. 3-pounders (fig. 46a). These have diameters of 2.68 to 2.79 inches with an average of 2.75 inches.
Type 4. 4-pounders. These have diameters of 3.00 and 3.01 inches.
Type 6. 6-pounders (fig. 46b). These have diameters of 3.41 to 3.52 inches with an average of 3.47 inches.
Variety a. Plain, except for mold markings. Seven specimens were found. The largest from Feature 69 (Level II) is malformed, almost hemispherical in shape. It was not intended for bar shot and appears to be a miscast ball.
Variety b. Marked with a broad arrow. Four specimens were found. Three were in the southwest casemate (two in Level II and one in Feature 51) the fourth on the east scarp (Level II). The broad arrow is indicative of British ordnance, but there is no evidence whether these dated from the British occupation, were fired at the fort, or were captured stores.
Type 12. 12-pounder. These have diameters of 4.38 and 4.43 inches.
Type 16. 16-pounder. This has a diameter of 4.56 inches. No records could be found of a cannon of this weight ever having been at Fort Stanwix. The ball was located just outside the entrance to the northwest bombproof (Level II) and may have been used as a doorstop.
During the siege of 1777, the British used 4.5-inch "Royal" mortars to lob shells into Fort Stanwix (fig. 46c). Shell fragments were found scattered over most of the fort area, but most fragments apparently had been gathered up after the battle. They were concentrated in Feature 73 on the parade ground and Feature 58 by the guard house (probably temporary privies), in Feature 34 in the north casemate and Level II in the east casemate, and in Features 69 and 72 in the east barracks. It is significant that mortar bomb fragments were absent in wood-lined cellar holes, except one fragment in Feature 52, and the trench (Feature 76) before the ravelin. It is quite possible that these structures may have been filled prior to the siege. Two whole 4.5-inch mortar bombs were found in Features 69 and 72 in the east barracks.
One of the mortar bomb fragments from the ditch (Level I) was from a 10-inch shell. Its origin is a mystery, but it is suspected that some soldier brought it back from the Great Lakes as a souvenir and discarded it at Fort Stanwix. It weighs 1,491 grams (approximately 3 lbs., 4.6 oz.) and lay near the top of the counterscarp next to the south end of the bridge.
The weight of the bomb fragments indicates that there was only enough metal to account for 11 mortar bombs in the collection. Much of the metal was probably gathered at the time and shipped to foundries to be re-melted into weapons for the Americans.
One hand grenade is located in the Campbell collection, but its provenience is unknown. It is 3.75 inches in diameter. Another fragment reportedly found years ago on the site is in the Rome Historical Society collection.
Whether these iron balls (fig. 46d) were cannister or grape shot is moot, since the distinction was based largely on how they were fixed for loading into cannon: the former were loaded in cans and the latter were tied in a bag to a wooden sabot. References to the ammunition in the garrison referred to both.
One cluster of 45 balls from the passageway to the northeast bombproof (Level II) were either loose or in a bag since a can was not found with them. However, there were can remnants adhering to two shot in Feature 69 (Levels I and II). Most of the cannister shot (368 specimens) was in the fill of Feature 69 in the east barracks (the same cellar with the concentration of gunflints), suggesting that ammunition was stored in the building at the time it burned.
Nearly all the shot shows casting marks in the form of seam lines and sprues. The size varied from area to area. In general, the shot fell into three sizes: 20 to 22 mm., 31.1 percent; 24 to 28 mm., 45.2 percent; and 37 to 40 mm., 6.1 percent. These sizes were not uniformly distributed (table 10).
Table 10. Comparison of cannister shot sizes in selected areas. (The Campbell collection was not measured.)
A total of 1,008 lead musket balls were recovered from the excavations (tables 11 and 12). The highest concentrations of these were in the northeast bombproof (Level II), the east scarp (Level II) and the west casemate (Level II). The lowest concentrations were in the northwest bombproof, the ditch, and the headquarters area. Most of the fort had four to five balls per 100 square feet.
Only seven balls appeared as if they had been fired and were flattened upon impact. Their distribution (table 7) suggests that they were fired from several locations, and the ones in the north casemate (Levels II and III) were probably souvenirs carried into the building. More balls had been cut or mutilated by chewing (fig. 46i) or were found with the sprue still untrimmed (fig. 46h). Joseph Moore (Simms, 1882, Vol. I, p. 590) remembered that two privates in 1776 who were whipped ". . . . . did not utter one word of complaint; but each taking a leaden bullet in his mouth, bit upon it as the punishment was inflicted." Two hundred and nine untrimmed balls were excavated, and, along with the lead waste encountered in various places, attested to the manufacture of musket balls in the fort. For details see entries under sprues and lead waste in this chapter.
Over half of the musket balls were .69 caliber, or 69 hundredths of an inch in diameter (fig. 46f). These could have been fired in a smooth bore .75 caliber weapon which was standard issue for both American and British Forces (Peterson, 1968, pp. 29-30). The smaller diameter ball was necessitated by the black powder used to fire the piece, which left a carbon deposit in the bore, and the paper cartridge used to seat the ball against the powder. All of the .66 to .72 caliber balls were probably used in .75 caliber weapons.
The next most frequent size found are .63 caliber balls (fig. 46g). These, with others ranging from .58 to .64 caliber, were probably made for imported French muskets of .69 caliber (Peterson, 1968, p. 37).
A final group of musket balls range from .47 to .56 caliber, dominated by the latter. These were probably used with American made rifles. To insure a tight fit in the rifled grooves of the bore, these balls were made with a smaller tolerance, and were wrapped in a greased patch rather than a paper cartridge. Some of these may have been made for pistols. The only pistol barrel in our collection is .63 caliber in size. At Fort Ligonier, 1758-1766 (Grimm, 1970, p. 109) the most common sizes were .69, .62 and .52 caliber. The first two sizes are comparable to the Fort Stanwix sample but apparently there is a shift from .52 caliber to .56 calibre during the period these forts were occupied.
There were 102 lead buckshot found in the excavation (table 13), (fig. 46j). These range in size from .25 to .44 caliber, and about a third still have sprues attached. No attempt was made to compute a mean since it is evident from the data that the buckshot cluster at .25, .31, .34 and .40 calibers.
Table 13. Distribution of buckshot and birdshot. (Ratios given are the number of specimens per 10 square feet of excavated area within the structure.)
There were 198 lead birdshot found on the site (table 13) (fig. 46k). These range in size from .06 to .21 caliber, and most are sub-spherical in shape. The sizes cluster at .135 to .165 calibers. Some were molded, others probably were made in a shot tower, and a few were cut from a small diameter lead wire. These were uncommonly numerous on the parade ground, largely due to a single concentration in a small area near its center.
Sprues and Lead Waste
Sprue fragments, untrimmed musket balls and lead waste comprise abundant evidence of the manufacture of musket balls, buckshot and birdshot at the fort (table 14). This industry appeared to have been concentrated in the north and southwest casemates, and the southwest bombproof. The high concentration of lead waste in the barracks probably represents storage, since it was not accompanied by sprues or untrimmed balls.
Table 14. Distribution of sprues and lead waste.
Individual sprues indicate molds for 1, 2, 5, 6, 9 and 15 balls (fig. 46e). The spacing of the sprues, however, is too close for balls larger than buckshot in size, and most made balls of birdshot size only. On "gang" molds the largest ball size possible was .56 caliber, and most are in the .20 caliber range. Inasmuch as larger balls were found with cut but untrimmed sprues, the question of where the sprues went remains. Probably, because there would have been much waste, they were remelted. While possible, it is unlikely that all large balls were made with single-shot molds.
One specimen was found in Feature 51 in the southwest casemate, one in the east casemate (Level II), six in Feature 69 (Level II) in the east barracks, one in the south ditch (Level XI) and three in the sally port (Level II). They consist of a hollow iron tube 5 mm. in diameter attached to a concave iron funnel 22 to 25 mm. in diameter (fig. 461). They were identified from a comparable specimen at Fort Ligonier (Grimm, 1970, pp. 74-76) for which the identification was uncertain. Three of our specimens have nail holes in the funnel part with the nails driven from the outside as if to attach the object to the end of a pole. This makes the identification even less certain. Nine of these were associated with post-1776 deposits.
A brass grenadier's match case was found in the northwest corner of the west casemate (Level II) (fig. 47). This is a perforated cylinder with reinforcing bands around the top and bottom. There is a perforated cone in one end. The whole case is 7 inches long and 1 inch in diameter. It was held by two rivets to a brass plate 4.2 by 2.1 inches in size which was sewn to a belt. It was designed to carry a slow match for igniting incendiaries, such as grenades. The slow match and pull ring were not found. A similar specimen is illustrated by Darling (1970, p. 1). It was probably a relic of the British occupation.
Clothing and Ornamentation
Nine fragments of cloth were found in pre-1781 contexts at the site. They had been preserved by charring or contact with metal. All were a simple over one and under one weave with course thread. Where it could be discerned, the thread had a Z-twist. All could be classed as coarse cloth. A piece from Feature 51 retained a hem-stitched seam. The other specimens came from the southwest casemate (Level II and Feature 48), the north casemate (Level IV), the west casemate (Levels II and III), the northwest bombproof (Level II) and the west barracks (Feature 60, Level II).
Five fragments of braid were found. One, from the north casemate (Level I), consists of three tassels joined together. The tassels, formed from loops of heavy Z-twisted two-ply thread, were preserved by contact with a Type A-1 lead button.
Two strips of silver braid were recovered, one on the east scarp (Level II) and one in the sally port (Level II). The first specimen was backed with cloth with a Z-twist twined weave and was stitched to the garment by a single thread down the center (fig. 48a). It is 9 mm. wide and over 18 cm. long, and was cut or broken on both ends. The other specimen was woven over two-ply warp threads in an over-and-under technique to produce a twisted rope pattern (fig. 48b). It is 8.5 mm. wide and over 5 cm. long, and was broken at both ends. A brass specimen 7 mm. wide and over 2.7 cm. long, was found in Feature 57 (Level IV). It has a Z-twist twined weave method of manufacture. The fifth piece is of brass, and is too poorly preserved to be studied. It was found on the east scarp (Level II).
Buttons were a common find, particularly in the barracks and casemate areas (table 15). Some specimens could be identified by their markings as post-1775 in vintage. A total of 558 buttons were recovered, and in order to handle this large group, they were divided into types based on the method of manufacture and the materials used. It is not possible to classify an additional 55 specimens because they are too fragmentary. The method of analysis follows L. Stone (1971) but the types are our own. References will be made to Stanley South's typology (1964, pp. 113-134) since this is a standard work on the subject. We have not followed this typology because it does not meet our descriptive needs. This typology does not incorporate all of South's types, because some were not found in the pre-1781 component at Fort Stanwix. We have added new types found at Fort Stanwix.
Table 15. Distribution of button tpes. (The column
headed with a question mark in Series D represents specimens not
identifiable as to type; ratios given are the number of specimens per
10 square feet of excavated area within the structure.)
Basically, the buttons at Fort Stanwix fall into two size ranges: 11 to 20 mm. and 18 to 25 mm. in diameter. Some types are generally larger than others which accounts for the overlap. However, the number of specimens within these ranges is so high that we can state with confidence that, despite variations from one type to another, only two sizes served the needs of the soldiers. The smaller specimens were probably used on waistcoats and in lieu of knee buckles; the larger ones on regimental coats and the waist band of pants. Diameters were measured from edge to edge and the thickness includes the length of the shank.
Series AOne-piece cast buttons
Type 1. White metal with a mold seam on the back running over the shank. The seam around the edge was usually trimmed off (South's Type 11). Dimensions: small: diameter, 15 to 20 mm., average, 17.3 mm.; thickness, 8 to 12 mm., average, 10.7 mm.; large: diameter, 22 to 28 mm., average, 23.2 mm.; thickness, 9 to 13 mm., average, 11.0 mm. There are several varieties of this type, distinguished by their lettering or decoration. This type appears to be far more common at Fort Stanwix than at Fort Ligonier (Grimm, 1970, pp. 63-162).
Variety f. Face cast with "New Jersey" in script. One specimen (fig. 48h) has a diameter of 23 mm. and a thickness of 12 mm. This button was in Level II on the east scarp. It was a New Jersey regimental button of the Revolutionary period; the 3rd New Jersey Regiment was at Fort Stanwix from July to October, 1776, (Luzader, 1969, pp. 56-61). Another similar specimen (but not from the same mold) was found at Fort Ticonderoga (Calver and Bolton, 1950, p. 91).
Variety g. Face cast with "PS", and "R" and beaded edge. Three specimens (fig. 48i) are 24 mm. in diameter. No thickness measurements were possible. One was found in the north casemate (Level I), one in the guardhouse area (Level II) and one in Level II in the sally port. The button was probably worn by troops of the Pennsylvania State Regiment organized in April, 1777. It became the 13th Regiment of the Pennsylvania Line in October, 1777, and was incorporated into the 2nd Regiment of the Pennsylvania Line in July, 1778, (Lefferts, 1926, p. 52). No records could be found of this unit having been at Fort Stanwix. Another specimen was found in a grave near Philadelphia (Calver and Bolton, 1950, p. 140).
Variety h. Cast with "IX" on face. One specimen is too fragmentary to measure. Found in the east casemate (Level II), it probably belonged to the 9th Massachusetts Regiment of the Continental Line, elements of which were at Fort Stanwix in August, 1777, (Luzader, 1969, p. 106).
Variety i. Cast with an Arabic numeral on face. Nine specimens (fig. 48j) have the following dimensions: diameter, all 22 mm. except one, 17 mm., which has the numeral "8" on it; thickness, 10 to 11 mm.
One of these buttons was in the north casemate (Level II), two in the east casemate (Level II), one on the east scarp (Level II) and five in Level II in the sally port. They exhibit the numerals "7" (two), "8" (two), "9" (one), "12" (two) and "14" (two). These were probably worn by American militia units.
Variety j. Cast with Arabic numerals and other designs on face. Three specimens (fig. 48k) have diameters of 17 mm., 25 mm. and 17 mm. No thickness measurement is possible. These were British regimental buttons of the 26th, 29th and 53rd regiments of Foot, respectively. The 26th button has the number "26" in the center with a rope border. The 26th Regiment served in New Jersey in 1767, (Darling, 1970, p. 56). Part of the 26th was captured at Fort St. John's, N.Y., in November, 1775 (Dupuy and Dupuy, 1963, pp. 88-70) and the regiment later served out of New York City until December, 1779. This button was probably grafted onto a New York uniform at St. John's or Montreal. It was found in the east casemate (Level II). Others like this were found at a British camp in New York City (Calver and Bolton, 1950, p. 115). The 29th button exhibits the number "29" in the center with a broad arrow border. This regiment arrived at Boston in 1768 (Lefferts, 1926, p. 172), serving primarily in Canada. Two companies were with Burgoyne at Saratoga (Bird, 1963, p. 280) where they were captured. It was found in the ravelin (Level II). Similar specimens have been found on Carleton Island (Calver and Bolton, 1950, p. 57) and at Fort Erie (Calver and Bolton, 1950, p. 118). The 53rd Regiment button has a "53" on a depressed ribbed center. This regiment arrived at Quebec in 1776, and served with Burgoyne as a rear guard at Fort Ticonderoga (Bird, 1963, p. 94). It was found in the sally port (Level II). While none of these regiments were at Fort Stanwix, all had stores captured by the Americans between 1775 and 1777.
Variety k. Face cast from a coin mold. Dimensions of four specimens: 28 by 11 mm., 23 by 11 mm., 27 by ? mm., and 18 by 10 mm. The first of these was made from a George II halfpenny mold. The second was from a Spanish coin dated 1767 and the last two were from Spanish coins with illegible dates. They were found in the north casemate (Level II), the east casemate (Level II), the east scarp (Level II) and the sally port (Level II).
Variety l. Cast with horse and rider on its face. Three specimens (fig. 48l) have the following dimensions: diameter, 24 mm., 16 mm. and 16 mm.; thickness, 8 mm. for the smaller size. These buttons exhibit horse and rider in "curvet" position. Two were found in the sally port (Level II) and one in the guardhouse area (Level II).
Variety m. Back cast with initials "PN" flanking the shank. Diameter of one specimen: 17 mm. The shank is broken off. Buttons with these initials were found at Fort Ligonier, 1758-1766 (Grimm, 1970, p. 63), but they were probably comparable to our Type B2. This suggests a pre-Revolutionary context for this specimen which was found in the north casemate (Level I).
Variety n. Silver plated. One specimen found in the sally port (Level II) has a diameter of 17 mm. and a thickness of 11 mm.
Variety o. Face cast with urn and wreath border. One specimen (fig. 48m.) has a diameter of 25 mm. and a thickness of 12.5 mm. It came from Level II in the sally port.
Variety z. Face cast with geometric or symmetrical designs. Dimensions of 21 specimens (fig. 48n): small: diameter, 15 to 19 mm., average, 17.3 mm.; thickness, 8 to 12 mm., average, 10.1 mm.; large: diameter, 22 to 25 mm., average, 23.3 mm.; thickness, 10 to 13 mm., average, 11.0 mm. These specimens were found in the north casemate (Levels I and II), the east casemate (Level II), the southeast casemate (Level II), the east barracks (Feature 52, Level III), the west barracks (Feature 57, Level II), the ravelin (Feature 76) and the sally port (Level II).
Type 2. White metal with a hole drilled through a solid cast shank on the back, otherwise like Type 1 (fig. 48p). This plain button has a diameter of 23 mm. and a thickness of 8 mm.
Type 3. Brass with a hole drilled through a solid cast shank on the back.
Variety a. Plain face. Dimensions of two specimens (fig. 48q): diameter, 19 and 20 mm.; thickness 6 and 9 mm. They were found in the southwest casemate (Level I) and west casemate (Level II).
Variety z. Decorated face with symmetrical designs. Diameters of four specimens (fig. 48r): 16 mm., 17 mm., 19 mm., and 26 mm.; thicknesses: 6 mm., 6 mm., 8 mm. and 9 mm. No two designs are alike. One, from the north casemate (Level I) with a snowflake motif has a concave face. An identical specimen was found at Valley Forge in the Weeden Brigade area (personal observation). Another, from the east scarp (Level II), has holes punched through it as part of the design. A similar specimen was found at Carleton Island (personal observation).
Type 4. Brass specimen with two transverse holes crossing through a solid cast shank (fig. 48s). This specimen has a diameter of 17 mm. and a thickness of 6 mm. Cast on the front is the number "18" in a broken circle. This was probably a button of the British 18th Regiment of Foot which arrived at Boston in October 1774, and the officers, at least, returned to England in July 1776, (Lefferts, 1926, p. 171). Darling (1970, p. 56) listed this regiment as having arrived in Philadelphia in 1767. The button was found in Level II in the sally port.
Series BTwo-piece buttons with a brazed or flux-jointed seam and a separate shank.
Type 1. Brass specimen with a brass wire shank, a brazed seam and two vent holes.
Variety a. Plain face. Dimensions of 44 specimens (fig. 48t): small: diameter, 16 to 17 mm.; thickness, 14 mm.; large; diameter, 20 to 24 mm., average, 21.2 mm.; thickness, 14 to 17 mm, average, 15.5 mm. Eleven of these were in Feature 60 (Levels II and III) in the west barracks and seven were on the east scarp (Level II). Others were scattered about the fort.
Variety b. Silver plated. One specimen from the north casemate (Level III), is 22 mm. in diameter.
Type 2. White metal with an iron wire shank and two vent holes (South's Type 12?, Grimm's Type 1?; fig. 48u). Dimensions: small: diameter, 15 to 17 mm., average, 16.3 mm.: large: diameter, 19 to 32 mm., average, 21.3 mm. All shanks were broken off. A concentration of these buttons was found in the east end of the north casemate (Level I) and the east scarp (Level II). South's Type 12 was defined as a solid button, but apparently he had only one badly preserved specimen to describe (South, 1964, p. 119). A few of our buttons were so badly corroded as to appear solid, but the presence of vent holes and a few well-preserved specimens leads us to conclude that all were hollow. Grimm (1970, pp. 62-63) noted one-piece hollow pewter buttons with two holes and iron shanks as the most common type at Fort Ligonier, 1758-1766. These were probably two-piece buttons since it would be difficult to cast a hollow pewter button and imbed an iron shank, and the only function of the vent holes would be to allow gases to escape while the two halves were joined together. They were the third most popular type at Fort Stanwix, and the most common at Fort Ligonier which suggests that they are pre-1776 in time, or British military buttons. L. Stone (1970, p. 171) also believes that they are British, and notes that they occur in a post-1769-1781 context at Fort Michilimackinac. It should be noted that only one possible specimen came from the civilian context of Brunswick Town, North Carolina, 1726-1776 (South, 1964, p. 119).
Type 3. Brass with a brass wire shank (South's Type 6; fig. 49a). Dimensions: small: diameter, 14 to 18 mm., average, 16.2 mm.; thickness, 11 to 14 mm., average, 12.1 mm.; large: diameter, 19 to 24 mm., average, 21.4 mm.; thickness, 13 mm. One specimen from the east scarp (Level II) has a leather thong through the shank and another from Feature 73 has an iron shank. Symmetrical designs are on about 25 percent of the faces, but the faces of many buttons of this type are broken off. These were not separated from the plain specimens since we could get no accurate count.
Series CTwo-piece cast button, flux-joined with a drilled shank.
Type 1. Brass (South's, Type 1; fig. 49b). Diameters: 19 mm. and 29 mm. No thickness measurement was possible because the faces were broken off.
Series DTwo-piece button with brass face and crimped edge; separate shank on some types.
Type 1. Four-hole bone back (South's Type 3; fig. 49c). Dimensions: small: diameter, 15 to 18 mm., average, 15.9 mm.; thickness, 3 to 5 mm., average, 4.0 mm.; large: diameter, 19 to 26 mm., average, 22.2 mm.; thickness, 5 mm. Most were complete enough to subdivide on the basis of decorative motifs.
Variety a. Plain. Nineteen specimens were found.
Variety b. Gilded. Four specimens were found. One of these, from the north casemate (Level II), has heavy thread through the holes forming an "X" on the back. Another, from Feature 48, was gilded over an embossed basketweave pattern. The others came from Feature 52 (Level IV) and Feature 76.
Variety c. Silver plated. Two specimens were found. One, from the west casemate (Level II), was plated over an embossed basket-weave pattern. The other was from the northeast bombproof passage way (Level II).
Variety z. Various symmetrical embossed designs. Five specimens (fig. 49c) were found. The most common patterns were a basket-weave motif and a lattice motif.
Type 2. Specimens with a four-hole wooden back (South Type 3; fig. 49d, e). Dimensions: small diameter, 15 mm., 16 mm., 16 mm. and 17 mm.; large: diameter, 21 mm., 21 mm., 21 mm. and 22 mm.; warping of the wood prevented thickness measurements. Most were complete enough to subdivide on the basis of decoration.
Variety a. Plain. Six specimens were found.
Variety b. Gilded. One from the north casemate (Level I) was gilded over a symmetrical embossed design of six leaves and scallops (fig. 49e).
Variety z. Various symmetrical embossed designs. Eight specimens (fig. 49d) were found.
Type 3. Specimens with a bone back, a central hole and a brass wire shank (South Type 4; fig. 49f). Dimensions: small: diameter, 14 to 16 mm., average, 14.8 mm.; thickness, 10 mm.; large: diameter, 18 to 24 mm., average, 21.1 mm.; thickness, 11 to 13 mm. One specimen from the west casemate (Level II) was gilded, and a few have embossed designs.
Type 4. Stamped brass back with a central hole and a brass wire shank (fig. 49g). Dimensions: small: diameter, 15 to 16 mm., average 15.9 mm.; thickness, 9 mm.; large: diameter, 21 to 28 mm., average, 24.0 mm. No thickness was measurable. Most specimens have an embossed wild rose motif surrounded by vines or ferns. One back (the largest specimen) from Feature 69 (Level III) exhibits the initials "I F" flanking the hole.
Type 5. Brass back with a brass wire shank soldered to it. Face covers only the edge of the button (South Type 16; fig. 49h). One has a diameter of 18 mm. No thickness was measurable.
It exhibits an eight-pointed star motif and was found in the southwest casemate (Level I).
Type 6. Brass back with two crossed brass wires soldered to it (fig. 49i). Diameters: 16 mm., 22 mm. and 22 mm., and thicknesses of ?, 7 mm. and 8 mm. All were plain.
Category 1. Faces only, with the backs missing. These may have been Types 1, 2, 3, 4, or 6. Dimensions: small: diameter, 15 to 18 mm., average, 16.4 mm.; large: diameter, 23 to 34 mm., average, 25.7 mm. Most of these have symmetrical embossed designs and two from the southwest casemate (Level I) and the sally port (Level II) are gilded.
Series ECast one-piece buttons with a separate shank.
Type 1. Brass with a spun back and a brass wire shank imbedded in a casting spur (South's Type 7; fig. 49j, k). Dimensions: small: diameter, 13 to 19 mm., average 17.2 mm.; thickness, 6 to 12 mm., average, 9.0 mm.; large: diameter, 22 to 29 mm., average, 25.1 mm.; thickness, 6 to 13 mm.; average, 10.3 mm. Coarse cloth was found adhering to two of these buttons, but this appeared to be a fortuitous association. The shanks were usually broken off. Several varieties could be discerned.
Variety a. Plain. Sixty-seven specimens (fig. 49j) were found.
Variety b. Gilded. Three specimens were found. They came from the north casemate (Level I), the west ditch (backfill) and Feature 69 (Level I). These may all be post-1781, but this is unlikely.
Variety c. Silver plated face. Six specimens were found. One came from the south ditch (Level XV), one from the north casemate (Level I) and four from the sally port (Level II). One of the latter was engraved with a central flower motif, and a scalloped border.
Variety z. Engraved symmetrical designs. Six specimens (fig. 49k) were found.
Type 2. Brass with a visible mold seam and a brass wire shank imbedded in the back (South's Type 8; fig. 49l). Diameter: 24 mm. and 28 mm., thicknesses: 10 mm. and 11 mm. One specimen from the northwest bastion (Level I) is plain and the other from the east scarp (Level II) has a flower motif on its face.
Series FCut bone or shell buttons drilled for suspension.
Type 1. Bone with a single transverse hole through a node on the back (South's Type 14; fig. 49m). One specimen, 16 mm. in diameter and 5 mm. thick, was found in the west casemate (Level II).
Type 2. Bone with a single drilled hole (South Type 15; fig. 49n). Dimensions: small: diameter, 9 to 17 mm., average, 13.6 mm.; thickness, 1 to 3 mm., average, 2.1 mm.; large: diameter, 19 to 23 mm., average, 21.4 mm.; thickness, 2 to 3 mm., average, 2.4 mm. They were scattered about the site but a small concentration of partially finished buttons, the cow ribs from which they were cut, and an iron bit for cutting them, were found in Level II in the sally port (fig. 76l).
Type 3. Bone drilled with four holes (South's Type 20; fig. 49p). Diameters: 17 mm. and 25 mm., thicknesses: 3 mm. and 2 mm., respectively. This and the following type have been dated ca. 1800-1865 by South (1964, p. 121) but both specimens had an indisputable pre-1781 provenience. They came from the south ditch (Level XI) and Feature 57 (Level IV).
Type 4. Bone drilled with five holes (South's Type 19; fig. 49q). Diameter, 17 mm., thickness, 3 mm. They came from the east scarp (Level II) and the ravelin (Level II).
Series GOne-piece stamped with the shank soldered on.
Type 1. Hand stamped (South Type 9).
Variety a. Plain. Diameters of two specimens: 21 mm. and 24 mm. The latter has a thickness of 6 mm.
Variety b. Depicts a cannon pointing right with a "Grand Union" flag mounted on the trail. Diameters of six specimens (fig. 49r): 21 to 22 mm. All shanks were broken off. These were American artillery buttons of the Revolutionary period (Calver and Bolton, 1950, p. 85). Companies of Colonel Lamb's Artillery Regiment were stationed at the fort from 1777 to 1781 (M. Willett, 1777-78; Lauber, 1932). Two were found in the north casemate (Level II), one in Feature 56 (Level II), one in the sally port (Level I), one in the guardhouse area (Level I) and two in the Campbell collection.
Series HThree-piece button with a drilled back or with a separate shank.
The face was of two pieces with the outer one pierced to expose the inner one.
Type 1. Bone back drilled with four holes (South's Type 5; fig. 49s). Diameter, 17 mm., thickness, 4 mm. It came from Feature 69 (Level II).
Series ICast, faceted glass button with a shank imbedded in it. May have had a brass back.
Type 1. Brass shank (South's Type 13; fig. 49t). Diameter, 15 mm., thickness, 11 mm. It was found in Feature 56 (Level II).
Series JA wire-wrapped button.
Type 1. Wood back drilled with four holes and wrapped with brass wire in a cross-wise fashion (fig. 49u). Preservation was too poor to permit measurements. One came from Feature 48 and the other from the surface.
Thirty-three sleeve links, or fragments, were recovered from the fort. The number and distribution are shown on table 16, broken down by type and variety. Diameter and thickness were measured like the buttons.
Series AOne-piece cast with integral shank.
Type 1. Brass with a drilled brass shank coupled with a brass wire (fig. 49v). Dimensions: diameter, 13 to 17 mm., average, 14.8 mm.; thickness 5 to 9 mm., average, 5.7 mm.
Variety a. Plain. Five specimens were found. One was octagonal.
Variety b. Bust of a man flanked by the initials "PR." One specimen was found in the west casemate (Level II).
Variety c. Various symmetrical designs. Eleven specimens were found. Five are octagonal.
Type 2. White metal with a white metal or iron wire coupling. Diameters, 15 mm., 16 mm., 17 mm. and 14 mm. by 16 mm. The thickness of the second specimen is 6 mm. One has a plain face, two have symmetrical designs and one from the east casemate (Level II) appears to have been made from a coin mold (George I). One is oval in shape.
Series BOne-piece cast with a separate shank.
Type 1. Brass (fig. 49w). Dimensions: diameter, 11 to 16 mm.; average, 13.2 mm.; thickness, 4 to 6 mm.; average, 5.0 mm.
Variety a. Plain. One specimen was found.
Variety b. Stamped symmetrical design. Two octagonal specimens were found.
Variety c. Cast symmetrical designs. Two octagonal specimens were found.
Series CCast with a separate shank and a glass setting.
Type 1. Brass (fig. 49x). Dimensions: diameter, 13 mm. by 16 mm. and 12 mm. by 14 mm.; thickness, ? and 7 mm. Both are oval in shape. One has a purple faceted setting and the other is an opaque light blue. Two pressed glass settings, one blue and one amber, were found in the north casemate (Levels III and II) and a pressed clear glass setting was found in the southwest casemate (Level I). These may have come from sleeve links.
Series DCast with a drilled integral shank and a paste setting.
Type 1. Brass (fig. 49y). Dimensions: diameter, 12 mm., 12 mm. and 13 mm.; thickness, 7 mm., 8 mm. and 8 mm. One has a purple-and-white setting and the other two a blue-and-white setting.
Series ETwo-piece, cast with brazed seam, separate shank, and two vent holes in the back.
Type 1. Brass (fig. 49z). Diameter: 15 mm. by 11 mm.; thickness: 9 mm. It is oval and gilded.
Series FThree-piece, brass, shell and glass composite (South Type 34).
No measurements were possible, as the shank was broken off and the shell worn down around the edge. The shank and backing were cast as one piece with the shell as a collar below the glass setting. This was found in Level I and may be post-1781.
Type 1. Two-piece clasps (table 16) with studs and hook or eye on each piece (stock clasps).
Table 16. Distribution of sleeve links and clasps.
Variety a. Plain cast brass with three studs (fig. 50a). A piece of one specimen was found with a length of 30 mm., a width of 50 mm. and a thickness of 1 mm.
Variety b. Engraved cast brass with three studs. This has a length of 29 mm., width of 48 mm. and thickness of 1 mm. Only the hook half of one specimen was found by Campbell. A similar clasp was found at Fort Montgomery (Jack Mead, personal communication).
Type 2. Two-piece sheet brass clasps with hooks or holes for attachment to cloth (haversack or stock clasps); (fig. 50b, c). Dimensions: length, 31 to 33 mm.; width, 38 to 40 mm.; thickness .5 mm. These specimens have a row of six to eight holes along the back edge which allowed them to be sewn to cloth. One complete set from the north casemate (Level II) still retained the thread (fig. 50b). There was either a wide hook at the other end, or three rectangular holes in the center; the hook of one fits into the hole of the other. One specimen from the east scarp (Level II) which formerly had a hook, had been broken longitudinally and was converted by piercing with three crude holes (fig. 50c).
Type 3. Cut cast brass nailed to wood (function unknown). This has a length of 26 mm., width of 28 mm. and thickness of 2 mm. It has a single nail hole and was a reworked specimen, having been cut and hammered from a piece of cast brass.
A large assortment of buckles were found on the site, most of which were made for shoes (table 17). The following classification is tenuous in part because many buckles could have served several specific functions, although they were divided on the basis of form. There is an easily discernible relationship between form and function for some, such as shoe and stock buckles, but some of the smaller specimens could have been used on haver sacks, belts, or small harnesses. All buckles are of a copper alloy unless otherwise noted. In terms of types found, frequency of occurrence and size of specimens, the Fort Stanwix sample closely resembles the sample from Fort Ligonier, 1758-1766. (Grimm, 1970). There was apparently no great change in styles between 1758-1766 and 1766-1781 except for an increase in the popularity of Type 1f shoe buckles.
Type 1. Single frame with pin post and two-part fork, and toothed loop for double straps (shoe buckles).
Variety a. Large plain rectangular buckles with sharply rounded corners and a high arch (fig. 50d). Dimensions: length, 57 to 58 mm.; width, 44 to 45 mm.; thickness, 3.5 to 4.0 mm. These are heavy buckles and probably military issue. None were found at forts Ligonier (Grimm, 1970, pp. 56-58). or Michilimackinac (Lyle Stone, 1970). One fragment from the southwest bombproof (Level I), is silver plated and somewhat larger and thinner than the others.
Variety b. Large plain rectangular buckles with angled corners (fig. 50e). Dimensions: length, 47 to 58 mm., average, 51.3 mm.; width, 43 to 52 mm., average 46.5 mm.; thickness, 1.5 to 3.0 mm., average, 2.5 mm. These were probably military issue. Peterson (1968, p. 230) illustrates one from the Philadelphia, which was sunk in Lake Champlain in 1775. Two were made of iron. This was Grimm's Type 1 (1970, p. 56) at Fort Ligonier where it was also the most common type of shoe buckle. Three of ours are stamped "TURNER" and another "S. WILLSON" on the tongues. Two tongues with the latter mark on it were found at Fort Ligonier (Grimm, 1970, p. 62). Our example came from Feature 75 on the east scarp and the three "Turner" buckles came from the north casemate (Level I).
Variety c. Large relief molded rectangular buckles with angular corners (fig. 50f). Dimensions: length, 51 mm.; width, 38 to 44 mm., average, 40.3 mm.; thickness, 1.5 to 3.0 mm., average, 2.1 mm. These were probably civilian buckles. One from the southeast bastion (Level II) was silver plated. This was Grimm's Type 5 (1970, p. 58).
Variety d. Large relief molded rectangular buckles with excised areas and angular corners (fig. 50g) (Grimm's Type 3). Dimensions: length, 58 mm.; width, 46 to 50 mm.; thickness, 1.5 to 2.5 mm., average, 1.8 mm. These were probably civilian buckles.
Variety e. Large plain rectanguloid buckles with rounded corners (fig. 50h) (Grimm's Type 2). Dimensions: length, 54 to 58 mm.; width, 41 to 52 mm., average, 45.8 mm.; thickness, 2 to 3 mm., average, 2.6 mm. Peterson (1968, p. 230) refers to this type as a British military pattern. Two are made of iron and one of pewter.
Variety f. Large relief molded rectanguloid buckles with rounded corners (fig. 50i) (Grimm's Type 6). Dimensions: length, 63 to 67 mm.; width, 44 to 53 mm., average, 47.9 mm.; thickness, 1.5 to 3.0 mm., average, 2.3 mm. This was the most common variety, which argues against its having been civilian, but probably indicates the use of civilian buckles by the military. Very few of these (five) were found at Fort Ligonier (Grimm, 1970, p. 50). One from the east scarp (Level II) is silver plated.
Variety g. Large relief molded rectanguloid buckles with excised areas and rounded corners (fig. 50j) (Grimm's Type 3). Dimensions: length, unknown; width, 55 mm.; thickness, 2 mm. These were undoubtedly civilian buckles.
Variety h. Small relief molded rectangular buckles with angled corners (fig. 50k). This specimen is 38 mm. long and 2 mm. thick. It was probably worn on a woman's shoe because of its small size.
Variety i. Small relief molded rectanguloid buckles with rounded corners (fig. 50l). Dimensions: length, 39 to 43 mm.; width, 33 to 36 mm.; thickness, 2.0 to 2.5 mm. These were probably women's shoe buckles.
Variety j. Small relief molded rectangular buckles with excised areas and rounded corners (fig. 50m). No measurements were possible except thickness: 1.5 to 2.0 mm. These were probably from women's shoes.
Type 2. Single frame with pin post and two part fork and anchor. These buckles were nearly square (knee buckles) and had steel pins.
Variety a. Plain rectangular buckle with angled corners (fig. 51a). Dimensions: length, 24 to 30 mm., average, 27.9 mm.; width, 25 to 33 mm., average, 28.9 mm.; thickness, 1.5 to 3.0 mm., average, 1.9 mm. These were probably military issue and greatly resemble Type 1b shoe buckles.
Variety b. Relief molded rectangular buckle with angular corners (fig. 51b). This is 25 mm. long, 23 mm. wide and 2 mm. thick.
Variety c. Plain rectangular buckle with rounded corners (fig. 51d). This was 20 mm. long, 20 mm. wide and 2 mm. thick. This is probably a spur buckle.
Variety d. Relief molded rectangular buckle with rounded corners (fig. 51c). Dimensions: length, 33 mm.; width, 27 to 32 mm.; thickness, 1 to 2 mm.
Variety e. Relief molded rectangular buckle with excised areas and rounded corners. Dimensions: 33 mm. long, 30 mm. wide and 1.5 mm. thick.
Type 3. Single frame with pin post and a one piece fork and anchor (function uncertain). These were wider than they were long, and all movable parts were made of steel.
Variety a. Plain rectangular with angular corners (fig. 51e). This was 28 mm. long, 32 mm. wide and 3 mm. thick.
Variety b. Plain rectangular with rounded corners. These were about 33 mm. wide and 2 mm. thick. All specimens were fragmentary and one was made of pewter.
Variety c. Relief molded rectangular with rounded corners (fig. 51f). Length, 27 to 28 mm.; width, 28 to 42 mm., average, 34.4 mm.; thickness, 1 to 2 mm. Two specimens were made of pewter.
Variety d. Relief molded oval (fig. 51g). This was 30 mm. long, 38 mm. wide and 1.5 mm. thick and had a recessed pin terminal which Grimm (1970, p. 62) considered to date ca. 1775-1795. This specimen came from Feature 57 (Level IV) in the west barracks and cannot, therefore, date later than May, 1781. As it was the only one with this style of terminal found it was probably quite late, and a beginning date for the style of ca. 1775 would be acceptable.
Type 4. Single frame plain buckle with pin post, two-part fork and hook and rounded corners. Width greater than the length (stock buckle) (fig. 51h). See also: Clasps. Dimensions: length, 26 to 27 mm.; width, 39 to 44 mm.; thickness, 1.5 to 2.5 mm. One of these, from the east casemate (Level I) was marked "A BAHLER" on the hook part.
Type 5. Single frame with tongue looped over one side of the frame (belt or harness). Two small specimens were made of brass, the remainder of iron. Grimm (1970, p. 56) identified these as harness buckles.
Variety a. Small rectangular with angled corners (fig. 51i). Dimensions: length, 21 to 42 mm., average, 31.2 mm.; width, 24 to 45 mm., average, 33.9 mm.; thickness 2 to 5 mm., average, 4.2 mm. One of these from the east scarp (Level II) has leather attached (fig. 51i).
Variety b. Large rectangular with rounded corners (fig. 51j). This is 33 mm. long, 60 mm. wide and 5 mm. thick. It was a harness buckle from Level I and may be post-1781.
Type 6. Single frame plain rectangular buckle with studs and hook and rounded corners (sword belt buckle) (fig. 51k). This is 86 mm. long, 46 mm. wide and 4 mm. thick. It was in Level II of the north casemate.
Type 7. Double frame buckle with a steel tongue wrapped around the center of the buckle (baldric and belt buckles).
Variety a. Large plain double oval (fig. 51l). Dimensions: length, 46 to 51 mm., average, 48.3 mm.; width 7.4 to 10.7 cm., average, 8.5 cm.; thickness, 2.0 to 2.5 mm. One of these, from Feature 60 (Level III), has leather associated with it; a British broad arrow is cut into the leather. It appears to have been a bayonet frog but is too fragmentary to reconstruct.
Variety b. Small plain double oval. This is 38 mm. long, approximately 40 mm. wide and 2 mm. thick.
Variety c. Small plain rectangular with angled corners. These are 28 mm. wide and 2 mm. thick. Only fragmentary specimens were found. One from the northeast bombproof still has leather around the center post.
Variety d. Small plain rectangular buckles with rounded corners. Dimensions: length, 32 to 37 mm., average, 34.0 mm.; width, 43 to 63 mm., average, 51.3 mm.; thickness, 2 to 3 mm. One of these from Feature 3 (backfill) has leather around the center post (fig. 51m).
Variety e. Small plain D-shaped buckles (sling buckles) (fig. 51n). These are 47 mm. wide and 2 mm. thick. Both specimens are fragmentary.
Type 8. Double frame with steel tongue wrapped around one side of the frame which was notched (function unknown).
Variety a. Large plain rectangular buckles with rounded corners (fig. 51p). These are 61 mm. wide and 2 mm. thick. Both specimens are fragmentary.
Variety b. Small plain double oval buckles with a folded copper strap around the center post (garter?) (fig. 51q). This is 26 mm. long, 28 mm. wide and 1.5 mm. thick. The strap is 19 mm. long.
Type 9. Single frame, large plain U-shaped buckle with raised center post and rounded corners. Steel tongue wrapped around center post (harness buckle) (fig. 51r). This is 50 mm. long, 56 mm. wide and 4.5 mm. thick.
Type 10. Single frame, small round buckles with tongue wrapped around one side of the frame which was notched (ornamental) (fig. 51s). These have a diameter of 21 mm. and are .5 mm. thick. Both examples are silver and were probably made for the Indian trade.
These are all brass wire clothing fasteners (fig. 52a). They probably served a variety of clothing. Table 18 shows their distribution. Dimensions: length, 8 to 21 mm., average, 15.0 mm.; width, 7 to 19 mm., average, 11.7 mm.
A stamped and engraved copper strip was found in the sally port (Level II) (fig. 52b). It is 82+ mm. long and 18 mm. wide and has been broken off at one end. Two engraved lines running 2.5 mm. inside each edge appear to have been part of the original decoration. This had been modified by the addition of notches on both edges, two engraved stick figures with triangular bodies, and four clusters of designs stamped with a small octagonal tool and a punch. It was flat when found but appeared to have been bent at one time, and was probably a bracelet.
One brass finger ring with an octagonal bezel 11.5 by 9 mm. was found on the east scarp (Level II). The bezel was engraved "PI" with a rocker-stamped border. This type has been referred to as a "Jesuit ring" (Noel Hume, 1970, p. 266), distributed among the Indians as marks of conversion. The initials have been found on other rings and Cleland (1972, p. 206) suggests that these initials represent stylistic drift from an "IHS" design (Isus Hominis Salvator) dating pre-1700.
Several types of pendants were found on the site. See table 18 for their distribution.
Type 1. Disc.
Variety a. Plain lead (fig. 52c). This has a diameter of 35 mm. and is 3 mm. thick. It was probably hammered from a musket ball and perforated near one edge.
Variety b. Engraved lead. This has a diameter of 43 mm. and is 2 mm. thick. It was probably made from a musket ball and was perforated near one edge. On one side it is engraved with a Masonic emblem of an arch, an altar and an open eye. It may have been a post-1781 artifact since it came from Level X of the south ditch.
Variety c. Plain copper. This has a diameter of 28 mm. and is 2 mm. thick. It was probably a worn coin perforated near one edge.
Type 2. Lead cross. This is 58 mm. long, 50 mm. wide and 12 mm. thick. It was made in a crude wooden mold with a hole molded at one end (fig. 52d).
Type 3. Silver teardrop shape. This is 15 mm. long and 5 mm. in diameter. It is beaten silver with base, side and loop soldered together. It was probably an Indian trade item (fig. 52e) and was found in Level III of the sally port.
Type 4. Amorphous stone. This is 22 mm. long, 21 mm. wide and 4 mm. thick. This small piece of siltstone was ground and notched with a hole drilled in one corner (fig. 52f). It was probably Indian made, dating from a pre-fort occupation of the site, although found in Level II in the guardhouse area.
These were all made from scraps of brass bent to form a hollow cone (fig. 52g). They were attached to the fringes of clothing. One still has a leather thong in it. See Table 18 for their distribution. All but one from the north casemate were in definite pre-1781 context. They are 12 to 40 mm. long, averaging 20.7 mm.; comparable to specimens found at Fort Michilimackinac (Stone, 1970, p. 373).
Two tin-plated brass earrings were found (fig. 52h), one in Feature 52 (Level IV) in the east barracks and one in the sally port (Level III).
Both are wire loops .5 mm. thick and 10 to 10.5 mm. in diameter. One is broken but the other is divided at one point with thickened ends.
A total of 318 beads were recovered from the site (table 18). Only three, one bone, one ceramic and one plastic, were not manufactured from glass. Fourteen of the beads were definitely from post-1781 contexts, and another 15 were types manufactured over the past 200 years and were found in questionable context. These are not described below. The bead inventory was unvaried compared to those of historic period Indian habitation sites in this area. The beads exhibited only one surface color and were rather drab and monotonous. The beads were of types that had a long time range so they were of no use in separating the British from the American occupations of the fort.
Type 1. Simple, drawn, doughnut bead. This was the most abundant type found. They range in diameter from 2.0 to 4.0 mm. and have been called embroidery beads, versus seed beads that measure under 2.0 mm. in diameter (Hsu, 1969, p. 41). These are equivalent to Stone's seed beads, Class I, Series A, Type I (L. Stone, 1970, p. 350).
Variety a. Opaque white. These have a diameter of 2.0 to 4.0 mm. with 90 percent between 3.0 and 3.5 mm. The exterior is smooth and has a luster.
Variety b. Opaque turquoise. These have a diameter of 3.0 to 4.0 mm. with 90 percent between 3.0 and 3.5 mm. A majority of the turquoise beads have erroded or pitted exterior surfaces.
Variety c. Opaque grey-blue. One bead of this variety is 3.5 mm. in diameter. It has a smooth exterior surface.
Variety d. Translucent turquoise. These have a diameter of 3.0 to 4.0 mm. with 90 percent between 3.0 and 3.5 mm. The exterior surface is pitted and erroded.
Type 2. Compound, drawn, doughnut bead. This is identical to Type 1 except for the number of constituent parts (fig. 52j). The beads are white over white but the inside layer of glass does not have the same luster as the outer layer. These are equivalent to Stone's seed beads Class I, Series B, Type I (L. Stone, 1970, p. 353).
Type 3. Compound, drawn, tube bead.
Variety a. Clear over opaque white (fig. 52k).
These have a diameter of 9.0 to 11.0 mm. and a length of 6.0 to 9.0 mm. A transparent glass layer over an opaque white glass gave this type a pearl-like appearance. This is equivalent to Stone's necklace bead Class I, Series B, Type III (L. Stone 1970, p. 307).
Variety b. Opaque red over translucent green (Cornaline d'Aleppo). This has a diameter of 3.5 mm. and a length of 16.0 mm. This was recovered during the excavations of 1965, but no provenience data was given for it. It was cut from a longer tube and the ends have been fire polished. This is equivalent to Stone's seed bead Class I, Series B, Type III (L. Stone, 1970, p. 354).
Type 4. Simple, drawn tube bead. These are equivalent to Stone's necklace bead, Class I, Series A, Type VI (L. Stone, 1970, p. 301).
Variety a. Blue, translucent. These are 2.5 to 3.0 mm. long and 4.0 to 5.0 mm. in diameter. They were cut from longer tubes and heat tumbled to smooth the broken edges.
Variety b. Black, opaque (fig. 521). This has a diameter of 5.9 mm. and is 22 mm. long. This section was snapped off a longer tube and the ends are jagged and irregular.
Type 5. Simple, mandrel wound bead. These are equivalent to Stone's necklace beads, Class II, Series A, Type VIII, Variety A (L. Stone, 1970, p. 325).
Variety a. This has a diameter of 10.5 mm. and a length of 10.0 mm. The molten glass strand from which this bead was constructed was very thick, so the contact lines are quite distinct (fig. 52m).
Variety b. Clear, painted. This has a diameter of 7.5 mm. The exterior surface was painted to imitate a pearl. The beads found at Fort Ligonier were essentially the same as at Fort Stanwix (Grimm, 1970, pp. 49-50), except for a few wound beads at Fort Ligonier that were different colors. Neither fort had complex, multi-colored beads. It seems rather unusual that more beads and bead types were not found at Fort Stanwix considering the trade activities in and around the fort. There were accounts of Indian scouts, messengers and families staying for various spans of time. In 1775, it was reported that:
During the siege, Col. Willet stripped the Indian camps of bag and baggage, and in 1779 troops passed through the fort before and after destroying the Indian villages to the southwest. Two major treaty signings and many minor meetings were held with the Indians, and Indians are known to have visited the fort many times. Despite this, trade items, in general, were not found. See also: Buckles, Type 10; Bracelet; Ring; Pendants, Type 3; and Tinkling Cones.
Many tools were used to erect fortified positions; the most common were picks, shovels, spades, axes and billhooks. Picks, shovels and spades were needed to dig ditches and form earthworks. Axes and billhooks were used to clear trails and to cut wood and brush for revetting the earthworks.
In 1776, General Philip Schuyler requested intrenching tools for his various field commands, but was told that there were very few and that he should buy and borrow suitable tools from local inhabitants (Geo. Washington to Schuyler 6/13/76, Washington Papers, 1932). In December 1776, he ordered Henry Glen to:
Capt. De La Marquese, the French engineer at Fort Stanwix, reported making helves for axes, pickaxes and spades and other implements shortly after he arrived in 1777, but he did not inventory the number of tools. When the English troops retreated from Fort Stanwix on August 23, 1777, leaving behind much of their equipment, the American troops listed 100 picks, 50 billhooks, 80 falling (sic.) axes and 106 spades as part of the captured stores (Scott, 1927, facing p. 289). An inventory of engineer's stores taken on July 1, 1778, listed 238 picks, 90 billhooks, 170 axes and 150 spades (Clinton, 1900, #1554).
Spades and Shovels
Four types of spades and one type of shovel were found at the fort. The spades were rectangular with flat edges and flanged at the top for pushing with the foot. The shovels had slightly curved blades with pointed tips and no flanges. Our spades did not fit well into Peterson's typology (Peterson, 1968, pp. 181-182).
Type 1. Rectangular with straight sides and square corners (fig. 53b). The socket for the handle attachment is primarily above the blade so that only an inch or less of the bottom of the handle rests against the back of the blade. A rivet through the socket and the wood handle kept the handle from slipping or turning. The top 1/4 inch of the blade is bent forward to form the flange. The thickness of the blade is approximately 5/32 inch at the top and gradually tapers down to 3/32 inch at the bottom. The two spades of this type were found in the fill of the ditch (Levels XI, XIII) near the bridge area. One is 6-1/2 inches wide and 8 inches long and the other measures 6 inches wide and 8-1/4 inches long. The sockets of both are broken.
Type 2. Rectangular with slightly tapering sides and square corners (fig. 53c). The socket for the handle is more a part of the blade, rather than added on at the top. Approximately 3 inches of the handle rested against the back of the blade. A rivet kept the handle in place. The thickness and the flange are the same as Type 1. Two specimens of this type were found; one in Feature 57 (Level IV) in the west barracks measures 6-1/2 inches wide at the top, 5-1/2 inches at the bottom and 9-1/4 inches long and the other, from Feature 69 (Level II) in the east barracks, was 7 inches wide at the top, 5-3/4 inches at the bottom and 9-1/4 inches long.
Type 3. Metal spade with a wood core (fig. 53a). A wooden paddle and handle were carved from one piece of wood, fitting inside the blade of the metal spade. The back of the blade is one piece of metal and another piece of metal is welded on beginning about halfway up the front. The top of the front piece was bent backward and the back piece forward over the top of the wood paddle to form the top of the spade. Each part of the blade has a 4-inch shank at the top to secure the handle with rivets plus two or more rivets through the blade and paddle. Four specimens were found, one near the bridge (Level XIII), one in the sally port (Level II) and two in the middle of the southwest casemate (Levels I, II). They are quite uniform in size, measuring 6 to 6-1/2 inches in width at the bottom, 6-1/4 to 7 inches in width at the top and 11-1/2 to 12 inches in length.
Type 4. Wooden spade with a metal casing around the edge (fig. 53d). The casing is 2 to 4 inches long and 7 to 10 inches wide with two side flanges up to 9 inches long for attachment to a wood paddle. There is a V-shaped groove on the inside edges of the casing to seat the wood paddle. According to Noel Hume (1970, p. 275) nails driven through the flange and wood secured the two parts. The flanges of the casings in our collection are fragmentary but no evidence of nails or nail holes were observed. Eight specimens, five from the southwest casemate (Level II), one each from the east and west barracks (Level II) and one outside the ravelin (Feature 76) were found. None of the specimens are complete and no wooden portions survived.
The 18th century was a transitional period in the developmental sequence of spades. Noel Hume (1970, pp. 274-275) suggests that Type 4 spades were common before the 17th century, but were gradually being replaced by Types 1, 2 and 3 during the 18th century and by the beginning of the 19th century had been replaced completely. Type 3 was probably a transitional type between Type 4 and Types 1 and 2. The fact that half our spades are Type 4 suggests that they retained their popularity in this region or enjoyed a revival due to the scarcity of iron during the American Revolution.
One shovel blade was found in the southwest casemate (Level I) 80 percent intact, and fragments of two others came from the southwest casemate (Level I) and the south ditch (Level XIII). The blade was approximately 12 inches long, 10 inches wide at the top and the sides rounded gradually to a point (fig. 53e). The socket extended approximately 3 inches above the blade and protruded about 5 inches into the blade. A rivet through the socket and handle acted as a fastener. A nearly identical specimen was found at Fort Ligonier, 1758-1766 (Grimm, 1970, p. 144).
Two types of picks were found during the excavations. Type 1 is large and heavy while Type 2 is smaller and for light duty work. No handles were found.
Type 1. Large. These measure 21 to 22 inches from tip to tip along the top of the pick. The lengths of the blades and points were 9 to 9-1/4 inches. The differences in the total length are due to the size and shape of the oval eyes. The top of the cheeks varied from flat to slightly rounded, while the bottoms formed an ear approximately 1 inch wide.
Variety a. Two piece construction (fig. 54b). The blade, point and cheeks of the eye are two halves welded together. Two of these were found, one at the bottom of the scarp in the north ditch (Level XI) and one in the ravelin (Level II). The specimen in the ravelin has a heart stamped on the underside of the blade. Two axes, one from near Ft. Ticonderoga (Peterson, 1968, p. 184 and 1972 personal communication) have similar marks but no information about the mark has been found.
Variety b. Three piece construction. The blade of the three-piece pick was separate; it was welded to the two pieces that formed the point and the cheeks. This was found in the north casemate (Level II).
Type 2. Small (fig. 54a). These are 4 to 5 inches shorter and only two-thirds the weight of the larger type. The points and blades are 1 to 2 inches shorter than those of Type 1. The blades are only 1-1/2 inches wide, or less, all of two piece construction. The top of the cheek is slightly rounded while the bottom protrudes to a rounded ear. One was found in the north casemate (Level I) and the other in the east casemate (Level II). The right cheek of an undeterminable type was found in the north ditch (Level X).
The axe was an important tool in the construction, maintenance and everyday living in the fort. They were made in many different sizes and shapes; some for general purpose cutting and others for specific tasks. A great number of names, based upon size, shape, purpose, and/or area of origin have been used to classify axes.
Three terms were used in the inventories of axes at the fort dated August 23, 1777: "80 falling [sic.] axes were captured from the English," (Scott, 1927, facing p. 289) and July 1, 1778: 170 "narrow" axes and 16 "broadaxes" were part of the engineer's stores, (Clinton, 1900, #1554). The felling axes and narrow axes were undoubtedly one and the same. A large number of axes would have been used by the Americans during the reconstruction of the fort and just prior to the siege. While finishing construction, parties of axmen were sent along Wood Creek to obstruct that waterway with trees and bushes. The British were using axes to clear the obstructions in Wood Creek and the military road between Fort Oswego and Fort Stanwix.
A wide variety of terms have been applied to the parts of an axe; some are fanciful, some are manufacturer's terms, and some are traditional. A few terms seem to be widely accepted, but even knowledgeable writers have differences of opinion as to their meaning. For this reason, we include the following definitions:
Measurements of the axes were made as follows: the maximum length was taken from the butt to the edge; blade length from the front of the eye to the edge; edge width was the length of the cutting surface, butt width was the length of the butt; butt thickness was measured from the back of the eye to the back of the butt. The distribution of the axe forms is shown in table 19.
Table 19. Distribution of axes and wedges.
Type 1. European-style axe with the butt at an angle to the edge. The cheeks are flat on top and the bottom tapers upward from the butt toward the blade. The butt is rounded and the edge is convex. Usually, the blade measures well over one-half the total length of the tool. All of the Type 1 axes are broken in the same pace with one cheek and part of the butt missing. They also show signs of having been hammered on the butt, perhaps being used as wedges.
Variety a. Top of the axe flat from butt to edge, but the bottom of the blade flared from the front of the eye to the edge (fig. 55a). Two specimens were found, one 80 percent complete and the other consisting only of the right cheek. The measurable specimen is 7 inches long, the blade 4-3/4 inches long, the edge 4-1/4 inches wide and the butt 3-1/4 inches wide. It is of two-piece construction, welded along the blade and in the midline of the butt. The right cheek is missing, probably broken off when someone hammered on the butt attempting to use the axe as a wedge. From the size of the fragmentary specimen, it appears to have been very similar in size and shape.
Variety b. Blade flared both on the top and bottom (fig. 55b). It is of one-piece construction with the weld in front of the eye. Three measurable axes and one cheek fragment are representative of this variety. They are 7-1/2 to 8 inches long; the blade 4-1/2 to 5-1/2 inches long; the edge 3-3/4 to 4-3/4 inches wide and the butt 3 to 3-1/8 inches wide. One cheek and part or all of the butt are missing on all of these specimens.
Type 2. An intermediate form between the European form and American form of axes (fig. 55c). The. butt and edge are parallel and the top is flat but the blade is still more than one half the total length. The butt is squared and there is an ear on the bottom of the cheek. Two specimens belong in this category. They measure 9-1/4 inches and 7-1/4 inches long; they have 6-1/4-inch and 4-1/2-inch blades, 4-inch and 3-3/4-inch wide edges, 2-1/4-inch and 2-7/8-inch wide butts, and 1/2-inch thick polls. The larger one has sharp lower ears and the smaller one has rounded lower ears. Although the butts are polled, the center of gravity is still in front of the eye.
Type 3. Very similar to Type 2 but the butt has a much more distinct poll. There are no complete specimens of this type, only the butt and one cheek remaining from any single axe. The butt and bit were probably parallel.
Variety a. Flat top and lower ear. Four specimens make up this variety; the butt width ranges from 2 to 2-5/8 inches and the poll thickness from 5/16 to 1/2 inch.
Variety b. Ears on the top and bottom of the cheek near the back of the eye. The blade flares outward from the cheek on both the top and bottom. The two specimens of this variety fall into the same measurement range as Type 3a axes.
Type 4. American style axe. The blade is one half the total length of the tool and the center of gravity is near the eye. The top is flat and there are ears on the bottoms of the cheeks.
Variety a. A two-piece axe, the blade, cheeks and butt fashioned from one piece of metal. Another heavy piece of metal was welded onto the butt to square off the poll and counterbalance the weight of the blade. Five complete specimens make up this variety. Dimensions: length, 6-3/4 to 7-1/2 inches; blade length, 3-1/2 to 4 inches; edge width, 3-3/4 to 4 inches; butt width, 2-3/4 to 3-1/8 inches; and poll thickness, 7/8 to 1 inch.
Variety b. A three-piece axe with a steel bit (fig. 55d). This is identical to Type 4a, except for a steel wedge that was welded to the front of the blade to form a harder and longer lasting cutting edge. The one axe in this variety is 7 inches long and has a 3-5/8-inch-long blade, 4-1/8-inch-wide edge, 3-1/2-inch-wide butt and 1-inch thick poll.
Variety c. Three-piece construction with separate polls (fig. 55e). The two sides of the blade each have a cheek with the poll welded between the ends of the cheeks. Two specimens are in this variety. Dimensions: length, 6-3/4 and 7-1/4 inches; blade length, 3-1/2 inches; edge width, 3-7/8 and 3-3/4 inches; butt width 3-1/8 and 3 inches; poll thickness, 1 and 1-1/2 inches. The longer axe has the initials "EB" under an axe stamped on the left side of the blade.
Variety d. Unknown number of pieces (at least two). This variety has ears on the top and bottom of the cheeks and the top and bottom of the blade flares outward. Only the butt and cheek remain. The butt measures 2-1/4 inches wide and the poll 7/8 inch thick.
Type 5. Small, one-piece belt axe similar in shape to Type 1 (fig. 55f).
The top is flat, the bottom of the blade flares to the edge, the eye is circular and the butt is rounded. Three complete specimens and two cheek fragments are in this type. Dimensions: length, 5 to 5-1/2 inches; blade length, 3 to 4-1/8 inches; edge width, 2-1/4 to 2-5/8 inches; and butt width, 1-1/2 to 1-5/8 inches. The incomplete ones are of comparable size.
Type 6. Small, one-piece belt axe similar in shape to Type 3. The eye is wedge-shaped and the butt is squared.
Variety a. These exhibit a flat top while the bottom of the blade flares to the edge (fig. 55g). Two specimens have the following dimensions: length, 4 and 4-1/8 inches; blade length 2 and 2-1/8 inches; edge width, 1-1/2 and 2 inches; butt width, 1-1/2 and 1-3/8 inches; poll thickness, 3/8 and 1/2 inch.
Variety b. Blade flares on top and bottom toward the edge. One specimen has cheeks with small upper and lower ears. It is 5 inches long with a 3/4-inch-long blade, 3-1/2-inch-wide edge, 1-1/4-inch-wide butt and 3/8-inch-thick poll.
Type 7. A "pipe tomahawk" of two, possibly three-piece construction (fig. 55h). The steel blade is one piece but corrosion at the juncture of the eye and bowl prevented us from determining how these were attached. The draw hole in the eye of the axe was drilled at least partially from the eye side before the eye was formed. The top and bot tom of the blade flare to the edge, and the rim of the bowl is approximately twice the diameter of the bottom of the bowl. The specimen is 5-1/8 inches long with a blade 2-7/8 inches long, bit 2-1/2 inches wide, and pipe 1-1/2 inches long with a 3/4 inch inside rim diameter. It was in Level II of the east scarp.
From the inventory lists, apparently there was no distinction between "types" that performed the same general function. Broad axes were listed, but none were found during excavations. This may have been due to the relative few on hand, i.e., 16 in July 1778 (Clinton, 1900, #1554). Also, a broadaxe was a specialized tool that required some training before any degree of proficiency was reached. The master carpenter probably had direct control over these tools and kept close track of them because they were scarce. All of the belt axes were probably personal property; none were mentioned on any of the inventories.
The American style axe (Type 4) supposedly was developed by 1750 (Mercer, 1960, p. 7) while Hodgkinson (1965) gives a date of 1775. It was a well established type in the American colonies by the time of the Revolution (Mercer, 1960, p. 4). No examples of this type were found at Fort Ligonier, 1758-1766, (Grimm, 1970, p. 47) while at Fort Stanwix 10 out of 24 felling axes were American style. It would seem then, that Hodgkinson's date is best supported by the data. Hodgkinson (1965) dates Types 2 and 3 from about 1750, but none of these were found at Fort Ligonier. The presence of Types 1 and 2 axes on the east scarp (Level II) indicates that the European style axe had not been completely abandoned by 1776. There were not enough documented sources to attribute particular types of axes exclusively to one army or the other during the American Revolution. Probably both had certain difficulties in supplying the troops in the field, so all captured equipment would have been incorporated into the captor's stores. The British were using Tories and Canadians as militia (Luzader, 1969, p. 96) and these men were probably very familiar with the American style axe.
Splitting the logs into usable proportions was a formidable task that required considerable effort and skill with an axe. A wedge and hammer could have been handled with a minimum of instruction and would have been less fatiguing to the worker. Either the fatigue parties were small, Stockwell's wedges were made of wood, or the above-mentioned order was not fully carried out because only six iron wedges were listed on the two 1778 inventories. A good deal of wood was needed just for daily cooking, baking and heating. This did not include wood needed for construction, shingles and rails for fences which had to be split, and rough splicing of joints which could best be done with wedges.
Wedges could be made from metal, wood, or a combination of a metal bit and wood top. Only iron wedges were recovered during the excavation, and none appear ever to have had a wood top. All their heads had been pounded with a metal tool. Some are badly splayed and one shows signs of having split and been repaired by welding another wedge in the crack. No wooden mallets or beatles were found, nor were any metal ferrules from them identified.
Six wedges, four relatively complete and two broken in half, came from separate locations scattered around the fort (table 19). They were approximately 7 inches long and 2-3/8 to 1-1/2 inches wide (fig. 56e). All but one are beveled on both edges and the exception is beveled on one edge at the bit. The heads are rectangular and relatively flat while all the faces and edges are straight and smooth.
Another tool for splitting wood was the frow. Only one was found, in front of the ravelin (Feature 76) (fig. 56c). A frow was used when greater cutting control or uniformity of product was desired. Its use was also faster than, and did not expend as much energy for the same results as, a wedge or axe. The blade is only partially preserved; it measures 5-3/4 inches long, 2 inches in maximum width and is approximately 1/8 inch thick. It is of one-piece construction. The eye is 1-1/8 inches in diameter and 1-1/4 inches deep.
The billhook, also called a facine knife (Peterson, 1968, p. 185), was used to clear brush and small trees from fortifications and trials, and to trim sapplings for use in making facines (bundles of sticks used in revetting fortifications). They were employed much like modern machetes are used today. Fifty were captured from the English forces in 1777 (Scott, 1927, facing p. 289) and 90 were on the inventory in 1778 (Clinton, 1900, #1554). This specimen is shaped much like a cleaver with a hooked end on the cutting edge (fig. 56k). It was found between the southwest casemate and the west barracks (Level II) and has a cutting edge 8 inches long, and a hook 1 inch long at its widest part. The back of the blade is flat, the same as a knife blade. A one-piece conical socket handle, 4-1/2 inches long, could have been used without hafting it onto a wooden shaft.
The garrison was periodically supplied with cattle which were slaughtered at the fort (see Appendix B). A single butchering tool, an iron cleaver, was found in the west barracks. The blade is 8-1/2 inches long, 2-1/2 inches wide and 1/4 inch thick. The tip is rounded and the top is flat (fig. 56j). A pointed, tapered tang 4 inches long is 3/4 inch wide at its juncture with the top of the blade.
Claw. The claw hammer has been in existence as long as the nail. The tool was used by both carpenters and farriers. Until the cheap wire nail changed the pattern of fastening in carpentry, farriers used the claw hammer even more than carpenters (Mercer, 1960, p. 264). One wrought iron specimen was found in Feature 51 in the southwest casemate. The head is 4 inches long with probably 1/2 inch of the claw missing (fig. 56a). The poll is 1-1/2 inches long and almost 1 inch in diameter. The claw was formed by splitting the metal. A U-shaped metal strap was welded inside the eye so that the top of the eye was capped. The arms of the "U" protruded 1/2 inch below the bottom of the eye to reinforce the handle.
Peen. One half of a wrought iron light hammer, similar to an upholsterer's hammer, was found in the north casemate (Level I). The poll is 1-7/8 inches long, 3/4 inch in diameter near the eye, tapering to a 1/4 inch diameter at the peen. The top, bottom and outside surfaces of the eye are flat. The other end of this tool is missing.
Sledge. One-third of what appears to have been a sledge hammer was found in the bridge area (Level X). The polls are octagonally shaped, each side 1-1/2 inches wide, and 2-1/2 inches long. The cheek is 1-1/2 inches wide, 1-1/2 inches long, and 1 inch thick. One peen has been broken off and the remaining one has a concave surface.
Category 1. Two unknown metal tools, both broken and missing the same parts, were found in the sally port (Level II) and in the southeast casemate (Level II). They were classified under hammers because the remaining portion most likely was used as a hammer (fig. 56b). The one from the sally port is smaller, better preserved and shows better workmanship. Its functional parts are distinct and well formed. The poll measures 1/4 inches long, 5/8 by 3/8 inches at the eye, and 3/8 inch square at the peen. The cheeks are 2-1/2 inches wide with two metal rivets holding the head and handle together. The eye is only 1/2 inch wide and at least 3 inches long. The handle was probably elliptical in shape. Since the eye was so wide, the missing portion may have been a large hammering surface or a cutting edge. The specimen from the southeast casemate has a poll 1-3/8 inches long, 3/8 by 1-1/8 inches at the eye, and 3/8 by 5/8 inches at the peen. The cheeks are 2-3/8 inches wide and also have two rivets for attachment to the handle. The eye is 5/8 inch wide and approximately 3 inches long. There are similarities between these artifacts and shingling hatchets or lath hatchets. However, both of these types of hatchets would have had a flat top and been of heavier construction.
Files have traditionally been divided into two classes: metalworking and woodworking. Within these two classes, there were a large number of specialized forms for various craftsmen. Eighteenth century files were made by cutting grooves into the face of a blank bar with a cold chisel. The spacing of the grooves was judged by eye (Mercer, 1960, p. 293). Files could be singly cut with parallel grooves or doubly cut with cross hatched grooves. The inventories from the fort did not specifically mention files. However, they listed one carpenter's chest of tools and one set of blacksmith's tools which probably contained files. Several types of saws were listed and these had to be continually sharpened if they were put to much use. An inventory of sundry stores sent from Boston to Springfield, Bennington and Albany in January of 1777, had a large number of files listed (Gates Papers, Reel XI). Some of these were probably sent on to Fort Stanwix.
Metalworking Files. Used for sharpening, honing and smoothing. The grooves are relatively shallow but well-defined. All are made of steel.
Type 1. Elongated diamond cross section with two knife edges. The one specimen from Feature 69 (Level II) in the east barracks is incomplete; most of the tang and one-third to one-half of the blade survive. It has a flat tapering tang, round heel and straight edges. It is a single cut file with approximately 20 grooves to the inch. The ridge along the face of the blade is slightly off center and the blade is 1-1/4 inches wide.
Type 2. Rectangular cross section. The heel is square, the tang tapering to a point and the blade tapering slightly from the heel to a square or rounded tip. It is a single cut file with grooves on the faces and edges. Two specimens are in the collection. A complete file was found in the east barracks (Level I) and one blade was in the southeast casemate (Level III). The complete file is 9-3/4 inches long, 3/4 inch wide, tapering to 5/8 inch, 3/16 inch thick and has a 7-3/4 inch long working surface with approximately 40 grooves to the inch. The other is 1 inch wide, 1/4 inch thick and has approximately 32 grooves to the inch.
Type 3. Triangular cross section with tapered tang and blade. These were probably used primarily for sharpening saws. The three faces are equal in width ranging from 1/4 to 7/8 inch. None of the specimens are complete but they probably ranged in length from 7 to 10 inches.
Variety a. Single cut files with 50 to 60 grooves to the inch. One came from the north casemate (Level I) two from the west ditch (Level I) and one from the east casemate (Level II).
Variety b. Double cut files with 40 to 60 grooves to the inch. Two came from the north casemate (Levels I, II) two from the passageway of the northeast bombproof (Level II) and one from Feature 64 (Level III) in the west barracks.
Category 1. Three badly deteriorated specimens were in a tool kit found in Feature 60 (Level II) of the west barracks.
Type 4. Plano-convex cross section. One in complete specimen from the southwest casemate (Level III) measured 1/2 inch in width, tapering to a point, and 3/16 inch thick. The flat face was double cut and the convex face was single cut. It was too badly corroded to obtain a reasonable estimate on the number of grooves. Two more were found in Feature 60 and are described under tool kit.
Type 5. Knife file (Mercer, 1960, pp. 295-296). This has the same shape as a chef's knife, except for a uniform thickness (fig. 56d). The tang is a continuation of the upper half of the blade and is approximately one-half as long as the blade. Two specimens, one from the west ditch (Level I) and one on the parade ground near the headquarters (Level I) were complete. They had 2-3/8 and 2-1/2-inch-long blades, 1-1/4-inch-long tangs, 1/2 and 9/16-inch widths and were 1/8 inch thick. All of the grooves had been obliterated. This type of file was definitely used for sharpening saws.
Woodworking files. These are generally coarse with deep cut grooves. They are not rasps because the cutting surface was made by cutting grooves into the face of the metal rather than punching points out from the back of the cutting surface.
Type 1. Rectangular in cross section. The sides taper slightly to a square tip and the tangs on both specimens are missing. They have double cut faces and single cut edges, with approximately 16 grooves to the inch. They are 1-1/16 and 1-1/8 inches wide, 3/16 inches thick and approximately 9 inches long. One was found in the north casemate (Level I) and the other in the east casemate (Level I).
Two natural sandstone cobbles were found in the west end of the north casemate (Level II) which had been used as whetstones. Dimensions: length, 13.9 and 14.0 cm.; width, 31 and 47 mm.; and thickness, 31 and 38 mm. A fragment of a shaped slab of fine-grained sandstone, 49 mm. wide and 10 mm. thick was found in the west casemate (Level II).
Saws appeared on all the inventory lists that have been found for the American occupation of the fort. They were categorized into three types: crosscut, whip and hand. Presumably, hand saw was a miscellaneous category for all small saws. One small fragment of a steel saw blade was recovered from the sally port (Level II). The blade is 7/8 inch wide and less than 1/16 inch thick. The teeth are short and number 10 to the inch.
One small saw dog was found in Feature 76. The wedge-shaped tangs are 1-1/2 inches long and spaced 4-1/2 inches apart. This was used to hold a log in position while it was being trimmed and shaped.
A metal punch from the southwest casemate (Level I) was recovered. It is a heavy, stout tool that could have been used on either soft or hard material, but the head does not show heavy hammer marks. This tool was not on any of the inventory lists but neither were many other tools that artificers used. The punch was 5-1/2 inches long; 3-3/8 inches of the upper part of the stem is round and 5/8 inch in diameter. The lower part is square, tapering uniformally on all four sides to a sharp point.
Augers and Gimlets
Although there were no specific mention of the use of augers, the carpenters probably used them quite often since all the structures and furniture were constructed of wood. Pegging pieces of wood together was a common 18th-century method of joining. One fairly complete auger from Feature 57 (Level II), two possible shanks of augers from the northeast bombproof (Level I) and the east casemate (Level II), one auger bit from the southwest casemate (Level II) and one gimlet from the sally port (Level II) were recovered during excavations. The complete auger measures 7-1/8 inches long with a 5-inch-long shank, and a bit 2-1/8 inches long and 1/2 inch in diameter. The bit has two parts, a short screw tip and a spoon-shaped cylinder that cut out the wood clockwise (fig. 56g). The shank is 1/4 inch square and flattened out on two sides to a point to fit into a rectangular chuck.
The two shank fragments are 5 and 6 inches long respectively. The shorter one has a 1/4-inch-square shank with a round end, and the other has a 1/4-inch-diameter shank with a square tapered end like a modern auger (fig. 56h). The bit fragment is missing the shank and the upper part of the bit. It is a nose auger (Mercer, 1960, p. 185), that is, all the cutting occurred at the tip by the knife-shaped cutting edge that is almost perpendicular to the length of the bit. There is no screw tip, so a hole was started for this auger and pressure was applied via the handle in order to cut the hole. The bit is a half cylinder and had to be continually withdrawn from the hole to discharge the shavings. This bit cut a hole 1 inch in diameter with a flat bottom. The bit cut in a clockwise turn.
A gimlet is a small diameter bit for drilling wood. Ten "Gimbletts" were listed on the April 1778, engineers' inventory (Clinton, 1900, 1554), while all other references to "Gimblets" were under artillery stores. Gimlets could have been used like vent picks for cannon as well as for boring holes in wood. There were other uses as well.
Only one gimlet fragment was found. The bit is missing but the shank is 4-1/8 inches long, round at the end and square towards the bit.
A total of 20 bone buttons (see Buttons, Type F-2), 33 bone blanks and three button bits were found. All of the blanks except three came from the sally port area (Level II). One complete bit, 1 inch in diameter, was found in the sally port (Level II) (fig. 76l). One with the cutting edge missing, about 3/4 inch in diameter, was found in the north casemate (Level I) and the shank of a third was found in the bridge area (Level II). Several of the cuts in the used blanks could be matched with the larger bit found with them. This type of bit dates back to Roman sites of the First Century A.D. (Mercer, 1960, pp. 195-198).
The bit consists of three pointed cutting surfaces on a flat blade; one in the center and two at the edges, the area between the center and the edges arched and never cutting deeply into the bone. Most of the buttons have circular striations on both faces which indicates cutting from both sides of the blank. The bit is 3-1/8 inches long, the end of the tang 5/16 inch wide and the sides flaring out to 1 inch at the tip. It is uniformally 1/8 inch thick and has flat faces and squared edges. The three points are all of different lengths; the center point being the longest, 1/16 and 1/4 inch longer than the other points. The center point is beveled on three surfaces to come to a sharp tip, while the outer points are beveled on the inside and outside surfaces to form a knife edge. The bit was turned clockwise when cutting out the button; it could be turned by hand without a handle but it was probably fitted with a wooden handle with a rectangular socket that fit over the end of the bit. There was a hole in the shank for securing the handle.
The second bit was incomplete and the cutting surface had rusted away. It has a tang 2 inches long and 1/4 inch square, a shaft 3/16 inch in diameter and 1-1/4 inches long and a flat blade 1/8 inch thick and about 3/4 inch wide. This bit would have required a handle or a brace. The shank of the third was shaped the same as the complete bit but no measurements could be taken.
Two chisels were found in the excavation, one in the bridge area of the ditch (Level X) and the other in the southwest casemate (Level II) (fig. 56f). They may have been part of the blacksmith's kit or perhaps a brickmason's kit. They were different from wedges in that there was a distinct handle for the worker to grasp and the blade flared out from the handle to the bit. The chisel from the casemate is heavy and massive. It is 8-1/2 inches long, with a 2-inch-wide bit. The handle was 1 by 1-1/4 inches wide and 4 inches long and the head and bit are splayed. There was a broad arrow stamped into one face. The other chisel was 7-1/8 inches long with a 2-1/4-inch-wide bit and has a round handle 3/4 inch in diameter and 3-3/4 inches long. The head is slightly splayed from hammering.
Although screws were not common fastening devices during the late 18th century, one screw driver was found in the southwest casemate (Level II) (fig. 56i). The blade is 4-1/8 inches long, 1/4 inch square at the guard and flattened out to 3/8 by 1/8 inch at the tip. The tang is 1/4 inch square at the guard, 1-3/8 inches long and tapered to a point. The guard is 1/2 inch square and 1/8 inch thick. This tool does not have a well-defined head distinct from the shank; the tip is merely blunted.
Pieces of five assorted files, one punch, an iron bar and a nail, all wrapped together in a piece of coarse cloth and bound with twine, were found in Feature 60 (Level II). Three of the files are triangular (Type 3) and the other two are plano-convex (Type 4). One triangular file is 1/4 inch on a side; the others are smaller but too fragmentary to measure. Grooves were barely distinguishable but no count could be made. One plano-convex file was broken, but both halves were put into the kit. It is 10-1/4 inches long, 1 inch wide and 1/4 inch thick. The pointed tang is flat on the top and bottom and the blade is widest about 2 inches in front of the tang, tapering to a square tip 1/4 inch wide. All the grooves on the flat surface were obliterated. The top was double cut, but not enough remained to obtain a count. The fifth file is the tip half of another plano-convex file. It is smaller than the other and has a pointed tip. It is 3/4 inch wide and was double cut on both faces; the grooves were too badly deteriorated to count. The punch is only 3 inches long and 1/4 inch in diameter at the head, and 1/8 inch in diameter at the point. The metal bar may have been the middle section of a handle or a tang. It is rectangular in cross section, 1/4 inch wide at one end and 3/16 inch wide at the other. The wrought iron nail is corroded and incomplete.
There was such a mixture of items that this kit could not be easily identifiable with any particular craft. All the files appear to have been used for metalworking. This kit probably belonged to some sort of a "smith."
Parts of at least two iron sheaves were found in Feature 69 (Level II) in the east barracks. Rope, three poles and some sheaves are all that were needed to make a gin for lifting cannon tubes onto their carriages, or hoisting barrels and bales of supplies. The cellars of the barracks were probably used to store supplies, much of which were packed, shipped and stored in barrels that weighed up to 200 lbs. apiece (Leonard, 1780). The sheaves were approximately 3-3/4 inches in diameter and 1 inch wide at the rim. The center of the sheave rim was recessed 1/2 inch to hold a rope. It was impossible to tell how the sheaves were mounted or whether they were part of a block.
Three lead net sinkers were found ranging from 30.5 mm. to 35.5 mm. in length with a hole at one end (fig. 52i). These were found in Feature 76, on the east scarp (Level II) and in the northeast bombproof (Level I). Occasional fish bones in the fill indicated that fish were a part of the soldiers' diet (Appendix B) and the absence of fish-hooks coupled with these net sinkers suggests that fishing was not an individual enterprise but was an organized group activity. Netting usually requires cooperation between several individuals.
Overland travel on the frontier was principally by foot, although wealthy gentlemen and ranking officers used horses when they could. Horses and oxen were used primarily as draft animals to pull wagons and sleighs although we have no direct references to oxen at the fort.
Fourteen whole and fragmentary horseshoes (table 20) were found scattered around the site. A farrier must have been shoeing horses for so many of them to have been found here.
Table 20. Distribution of horseshoes and oxshoes.
Type 1. Flat surfaces with no calkings. The nail holes are rectangular in shape with no consistent pattern of spacing within the type.
Variety a. U-shaped with parallel branches (fig. 57a). The ground surface was convex and the foot surface (Noel Hume, 1970, p. 237) was flat. The branches taper slightly toward the heel but otherwise the shoes are of even width throughout. The complete shoe is 4 inches long and 4-3/8 inches wide. These were made from straps 3/4 inch wide and 5/16 inch thick. Seven holes, three on each branch and one in the toe were cut into the complete shoe. The other shoe has three holes on the branch.
Variety b. U-shaped with converging branches and fullered. The complete shoe has a flat foot surface and convex ground surface while the fragmentary shoe has two flat surfaces. They measure 3-3/4 inches and 4-1/2 inches in width; the complete specimen is 5 inches wide. The metal strips from which they were made measure 1-1/8 by 1/4 inches for the larger and 5/8 by 1/4 inch for the smaller. The larger shoe has three closely spaced rectangular holes on each branch. The smaller one has two holes and only a very narrow and shallow fuller.
Variety c. U-shaped with diverging heels. A fragment of one shoe was found. The metal strip is 3/4 inch wide while the thickness varies from 1/16 inch on the outside edge to 1/8 inch on the inside and increased to 3/8 inch at the heel.
Type 2. Branches with calkins.
Variety a. U-shaped with converging branches and fullered (fig. 57b). The calkins vary in size and shape. Small ones are the width of the shoe, 1/8 to 1/4 inch and projected 1/4 inch from the ground surface. Large ones are 1/4 to 1/2 inch wide and project 3/8 inch. Each branch has three or four nail holes, flat foot surfaces and convex ground surfaces. The one complete specimen is 4-3/4 inches long and 5 inches wide; the metal strip from which it was formed is 3/4 inch wide and 1/4 inch thick.
Variety b. U-shaped with converging branches. The calkins on both specimens are approximately 3/8 inch long and project 1/8 inch. They are 5-1/4 and 5-1/2 inches long and 4-3/4 inches wide. The larger specimen was made from a strip 1 inch wide and 3/16 inch thick and has three holes in one branch and four in the other. The other is 3/4 inch wide and 3/16 inch thick and has two holes in one branch and three in the other. They have flat foot and convex ground surfaces.
Type 3. Branches with calkins and cleated toes.
Variety a. U-shaped with parallel branches. This was the heaviest shoe found. The metal is 1 inch wide at the toe, tapered to 5/8 inch at the calkin and is 14 inch thick. The toe cleat is 1-1/2 inches wide, the same width as the shoe and 14 inch high. At least two nail holes were found approximately 1-1/2 and 2-1/2 inches from the point of the toe. It is 4-1/2 inches long and probably more than 5 inches wide.
Variety b. U-shaped with parallel branches and welded cleat (fig. 57c). The toe is 1-1/4 inches wide and the width narrows to 1/2 inch at the calkins. Three nail holes were made in each branch. A 1 by 1/2 by 1/4 inch iron cleat was welded on the front ground surface. Both surfaces are flat. The shoe is 4-3/4 inches long and 4-3/4 inches wide.
Type 4. Branches with calkins and clipped toe. This specimen is badly worn and misshapen. The point of the toe is bent up to form the clip and the calkins are indistinct knobs at the ends of the branches. One branch has two nail holes and was fullered, the other has four nail holes and was not fullered. The shoe is 5 inches long and 4-1/8 inches wide. The metal is 3/4 inch wide and 1 /8 inch thick. Both surfaces are flat.
An ox's hoof is split, so two shoes for each hoof are needed. Since oxshoes serve the same functions as horseshoes, they have many features in common and the terminology is the same. Since the shoes are not symmetrical, left and right side of the hoof is easily identifiable. Four complete shoes were found (table 20). The toes are the narrowest part and the branch widens out to the maximum width at the heel.
Type 1. Cleated toe and branches with calkins.
Variety a. Without fullering. One right side shoe was found. It is 4 inches long, 1/8 inch thick and 1/2 to 1-3/4 inches wide. The cleat and calkin project 1/8 inch below the shoe and both surfaces are flat. Five nail holes were bored along the branch.
Variety b. With fullering. One left side shoe was found. It is 4 inches long, 1/8 inch thick and 1/2 to 1-3/8 inches wide. The cleat and the calkin are 1/8 inch high. Three nail holes were bored in the branch.
Type 2. Flat surface with fullered branches. Two specimens, both right side, were found (fig. 57d). The one from the east ditch (Level II) is considerably larger than the one from the ravelin (Level II). The former may date post-1781. They measure 5-1/8 inches and 4 inches long, 1-3/4 inches and 1-3/8 inches maximum width and 1/4 inch thick. The larger one has five nail holes and a shallow narrow fuller while the smaller one has four nail holes and a wide deep fuller.
Two iron stirrups were found. One, found by Campbell, has a divided foot bar and an opening at the top for the harness. The other, from Feature 56 (Level III) in the west barracks has a foot bar divided into four rods and an iron loop welded (?) to the top (fig. 57p). It is 6-1/2 inches high including the loop and 5-1/2 inches wide.
Two iron spur fragments were found, a wrought one by Campbell and a cast one (fig. 57l) on the east scarp (Level I). The latter has a rowl mounted horizontally. It is 4 inches long and 3 inches across, cast with a scale motif. The one Campbell found is fragmentary but the rowl was mounted vertically.
Two parts of an iron snaffle bit were found in the north casemate (Level I) (fig. 57q). It is approximately 9 inches long with 4-1/2 inch cheek pieces on either side. It might date post-1781. However, identical specimens were found at Fort Ligonier, 1758-1766 (Grimm, 1970, p. 51).
Two iron frames with a loop at one end and a post riveted across the other (fig. 57k) were interpreted as harness parts. Both showed wear in the loop and a notch in one end of the rivet similar to a screw. They came from the north casemate (Level I) and the east casemate (Level II). Another possible fragment with an iron ring in the loop was found in the northeast bombproof (Level II) (fig. 57m). A hook with a T-shaped tip and a worn-through eye (fig. 57n) was found in the sally port (Level II). See also: Buckles, Type 5.
The metal parts, hubs, reinforcing bands and fasteners have survived, but no wooden parts identifiable as wheel spokes or felly sections were found.
Wheel Hubs. Two wheel hubs, one large and one small, were found (fig. 57i, j). The large specimen found in Feature 51 in the southwest casemate was made from an iron strap wrapped around a round mandrel with the ends bent outward and welded together. This ridge, approximately 1/2 inch high and one nipple 3/8 inch in diameter and 1/2 inch high kept the wooden part of the hub from slipping. It is 2-1/4 inches wide and 5/8 inch thick with an axle bore of 2-3/4 inches in diameter. The small hub, found in the sally port (Level III) may have been for a wheel barrow or small cart. It is cast steel 1-1/8 inches in diameter, 1-1/2 inches wide and 1/4 to 3/8 inch thick. The bore tapers from 5/8 to 1/2 inch in diameter. There were two 1/4-inch high wedge-shaped protuberances directly opposed to each other on the exterior.
Flange. A wrought iron circular disc, concave on one side and convex on the other, was found in the sally port (Level II). It is 3 inches in diameter and 1/2 inch thick; the concavity is 1/2 inch deep. A hole bored in the middle is 1 inch in diameter and the wall is 3/4 inch thick. This may have been part of the pivot assembly of the front axle on a wagon.
Reinforcing Band. A metal band approximately 4-1/2 inches in diameter was found in Feature 56 (Level III). It is 1-1/4 inches wide and 1/2 inch thick. When found, it was broken and too deteriorated to clean. Remnants of one nail driven from the exterior were still visible in the only hole found. See: Ferrules.
Cotter Keys and Pins. These two items served the same function, keeping pins and washers locked in place. Four keys were found (see: Linch Pin, Type 1a for distribution). They are shaped in an isosceles triangle, approximately 3 inches long, 3/4 inch wide at the base and 1/8 inch thick. They were inserted into the slot of the pin and bent 180 degrees to prevent them from slipping out. Cotter pins have not changed in form in the last two hundred years (fig. 57e). Two were found in Feature 51 (Level II), one in Feature 69 (Level II), one in the north casemate (Level III), one in the sally port (Level II) and one in the ravelin (Level II). Two were made from nail rods and the other four from flat metal strips. They vary in length from 1-3/4 to 2 inches long.
Linch Pin. Linch pins were inserted through holes at the end of axles to lock wheels on. They varied in length depending upon the diameter of the axle and hub.
Type 1. A slot at one end for a cotter key and the other end split so the prongs could be divided once the pin was set in place.
Variety a. Made from round steel rods with a slightly tapering point on the slotted end (fig. 57g). Four specimens, one from the guardhouse area (Level II), one from the ravelin (Level II) and two from the Campbell collection were found. All are broken off where the rod splits; the exact lengths are unknown. They measure 1/2 inch in diameter, approximately 8-3/4 inches from tip to beginning of prongs and 8 inches from key to prong. The slots are 3/8 to 1/2 inch from the tip, 3/8 to 1/2 inch long and 1/8 inch wide. All four pins still retained their keys.
Variety b. Made from a rectangular iron bar 5/8 by 3/4 inch (fig. 57f). It was found in the ditch (Level II). The prongs are broken off and the key was not attached. It measures 6 inches from the end to the beginning of the prongs and 5-3/4 inches from slot to prongs. The slot is 3/8 inch from the end and measures 3/4 inch long and 14 inch wide.
Type 2. Shaped like an arrow so the expanded "arrow head" replaced split prongs (fig. 57h). One specimen was found in the bakehouse (Level II). It measures 4-3/4 inches long and the shaft is 1/2 inch square. The maximum width of the triangular head is 1-1/4 inches and the slot is 1/2 inch from the end. The slot measures 1/2 inch long and 3/16 inch wide. The key was missing.
Ceramics have proven a valuable archeological tool for assigning dates to sites and interpreting the economic level of the occupants. Since the dates for Fort Stanwix are known, the taxonomic system set up here is geared to a description of the ceramics and an interpretation of the vessel forms represented in the collection. The basic breakdown into series is based on the paste of the vessels, which is a product of the types of clays used in making the vessels and the temperature at which they were fired. There are three series: porcelain, stoneware and earthenware. These are further divided into types based on color and treatment of the paste (coarseness), and the kind of glaze applied to the surface. A final division was made on the basis of surface decoration. It should be kept in mind that all these types were essentially contemporaneous in that they were in use between 1758 and 1781, and that the introduction of a new type did not necessarily mean that earlier types ceased to be made or used.
An effort was made to identify and count the minimum number of vessels in each type. This was accomplished by sorting all the sherds by their attributes (of which rim form and decoration were the most important) into separate clusters until no further matchups were possible. At this point we discovered the cross-mends between the north casemate and the east scarp (Level II) and sally port (Level II). A vessel might, therefore, be represented by from one to any number of potsherds. This provided us with an entirely different perspective to the ceramic inventory, and allowed us to develop some hypotheses regarding the economic status of various buildings' occupants. At the same time, it was noted that certain ceramic types were quite restricted as to their vessel forms, at least for the period and locality under study. Table 21 lists the sherd counts for the various units of the fort and tables 22 and 23 list the minimum vessel counts based on the sorting of the sherds.
Stanley South (1972) has developed a formula for dating pottery samples by averaging the sum of the product of the median dates for ceramic types (of their known time range) and the number of potsherds found for each type. We applied his formula to our total sherd sample from the site. We omitted porcelain and Type 1 earthenware because their known time range is too long to give meaningful median dates. The site has a median date of 1770.0 ± 11.5 years. For the potsherds (2705 sherds) we obtained a mean date of 1772.7 ± 14.3 years, with one standard deviation. Using the same method with the minimum vessel count (365 vessels) we arrived at a mean date of 1767.9 ± 12.9 years. Both of these mean dates are well within tolerable limits for the site (± 2.7 years and -2.1 years off respectively). We conclude that the formula is applicable to the Upper Mohawk Valley and yields accurate results. It would appear that minimal vessel counts yield more accurate (not necessarily better) dates, but this is an illusion created by using absolute numbers. A more accurate way of looking at these figures would be to say that Fort Stanwix was occupied from 1758 to 1781, while the potsherds indicate an occupation from 1758 to 1787, and the minimum vessel count indicates an occupation from 1755 to 1781. On the east scarp (Level II) and in the sally port (Level II) were large deposits which can be dated from 1776 to 1781. Dates derived from the potsherds in the deposits were 1755 to 1782 and by minimum vessel count, 1755 to 1783. This substantiates what we had already suspected, that these deposits included a lot of pre-1776 material in their fill. For comparison we applied the South technique to Fort Ligonier, 1758-1766 (Grimm, 1970, pp. 159-166) and obtained dates of 1756-1767 for the pottery sample (1761.8 ± 5.47 years).
No marked specimens of pre-1781 vintage were found at Fort Stanwix. Consequently, the identification of specimens rests largely upon a body of comparative data built up over the years and by documented specimens in various collections. Some names, such as creamware, are applied so generally that we use them in the discussion in preference to "refined buffware coated with a clear lead glaze." Other names, like Queensware, were trade names that have no utility in an archeological report. Josiah Wedgwood used Queensware to refer to his high quality creamware, but since we lack any marked pieces it is impossible to be certain that any of our creamware was made by Wedgewood.
Dimensions of vessels were taken whenever possible. Because most vessels were fragmentary it was necessary to measure the component parts and give composite dimensions expressed as the size ranges of the component parts. The measured components were rim diameters, foot-rings or basal diameters, and heights. Vessels referred to as "complete" were represented by sherds large enough to allow us to make the three basic measurements. All measurements were taken in the English system since this was the way they generally had been measured in the literature (Noel Hume, 1970), and to the nearest half inch to minimize the inaccuracies of projecting rim curves from small fragments, and because shrinkage during firing required potters to compensate by making vessels larger than desired. Variations in the paste of a vessel or the heat of the kiln would affect the size of the final product. On some porcelains this shrinkage might be as high as 11 per cent of the original size. (Pascale, n.d., p. 63).
Type 1. Hard paste Oriental. Chinese porcelain was made from a combination of kaolin and petuntse fired at a temperature of about 1400 degrees centigrade. Most Chinese porcelain was manufactured in Ching-te-Chen, China, and shipped from the port of Nanking (Pascale, n.d., p. 255). No clearly identifiable Japanese ceramics were found. Out of deference to the possibility, we refer to these sherds as Oriental rather than Chinese. The bulk of the sherds were decorated in under glaze blue, a few were painted in red around the lip and some were painted in red and gold over an underglaze blue, or in several overglaze colors.
Variety a. Undecorated. Most of the sherds in this variety probably came from undecorated portions of decorated vessels. A minimum of only six vessels were counted, two saucers, one cup and three bowls. One bowl has a foot-ring 2 inches in diameter.
Variety b. Decorated in underglaze blue paint (fig. 58a-c, e, h). There were a minimum of 98 vessels in this variety, eight plates, 36 saucers, 33 cups, three teapots and 18 bowls. Decoration was confined to the upper surfaces of plates and saucers, the exterior of cups and teapots and the interiors and exteriors of bowls. In most cases, the designs on the body are landscape scenes, and the rims are decorated with a ribbon of geometric motif. While the ribbon designs on the rims are often found on many vessels, the landscape designs are not repeated. Most of the decoration was done in light blue with well-defined lines, but a few specimens (fig. 58e) exhibit a darker smudged blue decoration typically found on early 19th-century "Canton" Chinese export porcelain (Noel Hume, 1970, pp. 262-263). These specimens are, however, clearly associated with pre-1781 deposits and the designs are not the "Canton" motifs.
One saucer, from the southwest casemate (Level II), measures 5-1/2 inches in diameter and 1 inch high with a 3-inch foot-ring and has a dark brown glaze on the exterior down to the foot ring (fig. 58b). Noel Hume, (1970, pp. 259-260) suggests that this type of glaze was popular ca. 1740-1780. This specimen dates between 1758 and 1781. One plate has a rim diameter of 6-1/2 inches while four others have foot-ring diameters of 4 to 5 inches. Rim diameters of saucers range from 4 to 5-1/2 inches and four measurable specimens are 1 inch high. The relationship of rim diameter to foot-ring diameter for these four specimens was 5-1/2 inches to 3 inches, 5 inches to 3 inches, 4-1/2 inches to 3 inches and 5 inches to 2-1/2 inches. Cups have rim diameters of 3 to 3-1/2 inches, a foot-ring diameter of 1-1/2 inch and one was 2 inches high. Bowls are 4 to 6 inches in diameter at the rim, 2 to 3 inches diameter at the foot-ring and one specimen was 2 inches high. This had a rim to foot-ring relationship of 5 inches to 2-1/2 inches. Miller and Stone (1970, p. 88) have suggested that there is a proportional relationship between the rim diameter and foot-ring diameter of oriental porcelain punch bowls on the order of 2:1. Since archeological specimens are rarely intact, it occurs to us that it might be worthwhile to check the dimensions of other vessel forms by type to see if some useful information would result. This will be explored further in the summary to this section.
Variety c. Decorated with overglaze red paint. Only a minimum of four vessels were found, three saucers and one cup. These vessels are generally thinner than the other varieties and decoration is confined to a narrow area below the lip on the interior of the vessel. This variety continued to be used into the second quarter of the 19th century, having been found in Feature 8, a privy, dating no earlier than 1840 (Hanson, 1974, p. 35). No complete specimens were found. One saucer has a 5-inch diameter while another has a 3-inch foot-ring diameter. The cup has a 3-inch diameter.
Variety d. Decorated with overglaze red, black, white, green and/or gold (fig. 58d, f, g). There were at least 19 of these vessels, three plates, eight saucers, three cups and five bowls. Two of the saucers also have underglaze blue decoration. Two complete saucers have rim diameters of 5 and 6 inches and foot-rings of 3 and 3-1/2 inches in diameter, respectively. Both are 1 inch high. A complete cup is 3 inches wide at the rim, 1-1/2 inches wide at the foot-ring and 1-1/2 inches high. A bowl is 4-1/2 inches wide at the rim with a 2 inch foot-ring and is 2-1/2 inches high. Another specimen has a 2-1/2-inch foot ring.
English ceramic production of the mid-18th century was dominated by the manufacture of a thin, white, salt-glazed stoneware. By the late 1730's slip-casting and press-molding in alabaster and plaster-of-paris molds enabled the manufacture of plates and other vessels in a variety of intricate relief patterns (Mountford, 1971, pp. 29-34). The salt glazing left a surface that can best be described in appearance as like that of an orange peel. This and the relief decoration, made these vessels hard to clean despite their durability and they were later replaced in popularity by refined earthenware (Godden, 1966, XV). No examples of white slipped over gray paste specimens (Noel Hume, 1970, pp. 114-115) were found on the site.
Type 1. White salt glazed.
Variety a. Undecorated, except for relief molding (figs. 59, 60). There were at least 162 vessels of this variety, 58 plates, 38 saucers, 31 cups, four sugar bowls, seven tankards, one punch bowl, 22 small bowls and one drug jar. Slip-casting and press-molding came into vogue in the 1730's and from 1740 to the 1770's were in common use to produce plates with elaborately molded rims (Noel Hume, 1970, pp. 115-118). A number of rim patterns were introduced. Those found at Fort Stanwix are:
Diaper Pattern. Twenty-four plates have alternating panels of basketry, dot and lattice or basketry, star and lattice designs (fig. 60a). These have 9- to 9-1/2-inch rims and one of the latter, from the southwest casemate (Level II) has a 4-1/2-inch base and is 1 inch high.
Gadrooned. Twelve plates have a simple gadrooned or "bead and reel" design along the rim edge (fig. 60b). These are 9-1/2 to 10 inches in diameter with 6-inch basal diameters and are 1 inch high.
Barley. Seven plates have panels of small ovals separated by ribbing (fig. 60d). One plate, from the east scarp (Level II) is 9-1/2 inches in diameter with a 5-inch base and is 1-1/2 inches high.
Queens. Two plates from the sally port (one from Level II and one from Level III) have a ribbed edge and plain panels separated by ribbing. One is 9 inches in diameter with a 6-inch base and is 1 inch high. The other has a rim diameter of 8 inches.
Barley and Wavy Lines. One 10-inch plate from the east scarp (Level II) has alternating panels of small ovals or nested wavy lines separated by ribbing (fig. 60e).
Feather-edge. One 10-inch plate from the sally port (Level II) has a feather edge design.
King of Prussia. One plate celebrating England's alliance with Prussia during the Seven Years War (fig. 60c) (Noel Hume, 1970, pp. 115-117) was found in the guard house area (Level II) and cannot be dated more closely than 1758-1781. It has a 9-inch diameter with a 6-inch base and is 1 inch high.
An additional five plates have undecorated rims (fig. 60f). These range in diameter from 9 to 10 inches with 6-inch bases and are 1/2 to 1 inch high. There are five complete saucers (fig. 59a); four with 4-1/2-inch rim diameters, 2-1/2-inch-foot-rings and 1 inch heights. The other is 5 inches in diameter with a 3-inch foot-ring and is 1 inch high. Other fragments have 4-inch rim diameters or 2-inch foot-rings. One cup has a 3-inch rim diameter with a 1-1/2-inch foot-ring and is 1-1/2 inches high (fig. 59b). Other fragments have a 4-inch rim diameter or foot-rings 1 to 2 inches in diameter. The identification of four sugar bowls as such is tenuous, but there seems little likelihood they could be anything else. One has a relief molded pattern similar to the diaper pattern on plates. It might have been a creamer. One plain specimen has a 1-1/2-inch rim diameter (fig. 59d). Tankard fragments have 3- to 4-inch rim diameters or 3- to 3-1/2-inch bases. One punch bowl has a 12-inch rim diameter and a 5-inch foot-ring. It stood approximately 6-1/2 inches high (fig. 59e). Other bowls range in diameter from 4-1/2 inches to 9 inches with foot-rings 2 to 3-1/2 inches in diameter (fig. 59c). A drug jar fragment could not be measured.
Variety b. Decorated with incised lines filled with cobalt blue (fig. 61b, e). There were a minimum of 72 vessels of this variety, commonly known as "scratch blue," 35 saucers, 28 cups, one sugar bowl, one tankard and seven small bowls. Decoration is confined to the exteriors of the cups, tankard and bowls, except for parallel scalloped lines which sometimes occur around the interior rims of cups. Saucers are decorated on the interior and sometimes have a rouletted band of chevrons around the bottom. Saucers range in diameter from 4 to 5 inches with 2- to 3-inch foot-rings and are 1 inch high. Cups are 2-1/2 to 3 inches in diameter with 1- to 1-1/2-inch foot-rings and 1-1/2 inches high. One sugar bowl lid from Feature 52 (Level IV) has a diameter of 1-1/2 inches and a tankard fragment has a rim diameter of 2 inches. Bowls range from 4 to 5 inches in diameter and one foot-ring 2-1/2 inches in diameter was found.
Variety c. Decorated with red paint. At least two saucers have red lines painted on their interior rims. One vessel is 5 inches in diameter with a 2-1/2-inch foot-ring and is 1 inch high. The other is 4 inches in diameter.
Variety d. Decorated with polychrome paint (fig. 61a). A minimum of ten saucers, seven cups and three small bowls have floral motifs painted in overglaze enamels. Colors include black, red, yellow, blue and green. One saucer is 5 inches in diameter with a 2-1/2-inch foot-ring and is 1 inch high. Another is 2-1/2 inches in diameter with a 1-inch foot-ring and is 1 inch high. This specimen, from Feature 56 (Level IV), was originally thought to be English porcelain but taking into account the style of decoration and the fact that it appears to have been refired (by the burning of the barracks?) led us to change its designation even though the paste compares well with English soft paste porcelain sherds from Fort Michilimackinac loaned to us by Lyle Stone. Other fragments have 4-1/2-inch rim diameters or were 3 inches across at the foot-ring. Cups are 3 inches in diameter and 1-1/2 inches high with 1-1/2-inch foot-rings. One bowl is 5 inches in diameter. None are decorated in a "Fazackerly" style (see Type 1d earthenware below).
Variety e. Covered with copper-tinted lead glaze. A single teapot lid fragment was found.
Type 2. Gray salt-glazed. These vessels are slightly thicker than the white salt-glazed and wheel thrown.
Variety a. With stamped designs filled with cobalt blue. There are only five sherds of this variety, all too small to identify by vessel shape, although thinness suggests tankards or chamber pots. These were probably made in the German district of Westerwald. Noel Hume (1970, p. 280) discusses the problems of identifying the source of this material. As this pottery was out of fashion by 1775 (South, 1972, p. 85), it was not surprising that so few sherds were found.
Variety b. Undecorated or with cobalt blue paint. A tankard, a jug and a jar were identified. These vessels are thicker than Variety a and may have been made in North America. They are in distinguishable from vessels common in the 19th century. Identification rests solely on their being found in pre-1781 contexts. It is possible that a few sherds of this variety were included with the post-1781 sherds where context was in doubt. The tankard is 3 inches across at the rim and the jug has a 1-inch opening. The jar is 4 inches in diameter at the base.
Type 3. Red stoneware with no glaze. A teapot lid fragment 3 inches in diameter and a fragment of a sugar bowl, both with sprigged decoration, were found. This has commonly been called "Elers" ware which is a misnomer since it was made by other potters besides the Elers brothers (Noel Hume (1970, pp. 120-121).
Earthenwares are made from low grade clays and fired at a relatively low temperature compared to stoneware and porcelain. Distinctions could be made on the basis of whether the paste was refined or coarse (with granules of grit and sand), whether the body was fired to a red or buff color and the kind of glaze applied to the surface. Given these variables, there was almost an infinite variety of wares produced and the typology must, of necessity, be rather arbitrary.
Type 1. Tin-glazed refined buffware. A few of these sherds have pinkish colored bodies but are included here for convenience, since they are otherwise indistinguishable in glaze and decoration. Tin-glazed earthenware was the first attempt to imitate Oriental porcelain and is actually lead-glazed with a tin oxide (Noel Hume, 1970, p. 106). John Cotter (personal communication) suggests that the origin lies in the Iranian tile-making industry of the first millenium B.C. It was being produced under the name "Majolica" in Spain and Italy after the 14th century and was derived from 8th century Moorish pottery (Pascale, n.d., p. 153). By the 16th century, potters from Faenza, Italy, began production in France under the name "Faience" (Pascale, n.d., p. 157) and in Holland and England where it was called "Delftware," a name that did not become popular until the 18th century (Noel Hume, 1970, p. 106). Most tin-glazed earthenwares were decorated with blue designs in imitation of the popular Chinese porcelain. It reached its zenith of popularity early in the 18th century and by mid-century was being replaced by white salt-glazed wares. Ivor Noel Hume's observation (1970, p. 111) that delftware cups should be rare after 1750, is born out by the fact that we found only two examples in 88 vessels.
Variety a. Undecorated. Like the undecorated porcelain sherds, many of these are probably parts of decorated vessels. There were at least 15 vessels of this variety; one plate, five bowls and nine small drug or ointment jars. Bowls have rim diameters of 5 to 8 inches while foot-rings vary from 2 to 3-1/2 inches. One drug jar is 1-1/2 inches in diameter with a 2-inch base and is 1 inch high (fig. 62e). Others are 2 to 3 inches in diameter.
Variety b. Decorated with blue paint. Decoration consists of scenes in which the recurring motifs are small houses on islands, water and boats and floral sprays (fig. 62c, f-h, 63a). Bands of geometric designs around the rim are rare. The minimum number of vessels is 56; 10 plates, three saucers, two cups, 34 bowls and seven apothecary jars. One bowl rim sherd from the southwest casemate (Level I) is clearly decorated in a French style (fig. 62d); the remainder appear to be English or Dutch. No Rouen Faience was found on the site. One plate is 9 inches in diameter and 1 inch high with a 4-1/2 inch foot-ring while another is 7 inches in diameter and 1 inch high with a 4-1/2 inch foot-ring. Other rims are 7 to 9 inches in diameter while a foot-ring is 5 inches across. These plates are slightly smaller than comparable specimens at Fort Michilimackinac (L. Stone 1970, p. 439). Two saucer fragments have a 4-1/2-inch rim and a 3-1/2-inch foot-ring, respectively. A cup is 3 inches in diameter. One punch bowl, from the southeast casemate (Level II) is 11 inches in diameter. Other bowls are 5 to 9 inches in diameter with 2-1/2- to 4-inch foot-rings. Two complete specimens are 5 inches in diameter and 2 inches high with a 2-1/2-inch foot-rings, while another is 6 inches in diameter and 3 inches high with a 3-inch foot-ring. Apothecary jars are 4 to 6 inches in diameter with 3- to 6-inch bases decorated with horizontal blue bands.
Variety c. Decorated with black paint. A single bowl from the east scarp (Level II) has flowers outlined on the exterior in fine black over glaze painted lines while in the bottom is the slogan, "Success to Co [ngress]," or "Co [mmerce]." It is too fragmentary to measure the height, but the foot-ring is 3 inches across and the rim diameter is approximately 9 inches.
Variety d. Decorated with red, black, blue, green and/or yellow paint. The common motif is brightly colored floral sprays. There are at least 10 vessels, one plate, one saucer, two punch bowls, six smaller bowls and one drug jar. The plate was represented by a rim sherd decorated in a French style (fig. 62d). All but one of the remaining vessels are decorated in a "Fazackerly" style, named after Thomas Fazackerly of Liverpool, England (Miller and Stone, 1970, p. 36) (fig. 62a, b). These were ascribed to Liverpool, ca. 1750-1760 (Miller and Stone, 1970, pp. 34-36). Some of the Fort Stanwix specimens were found in 1776-1781 contexts although they may have been redeposited. A drug jar, from the east casemate (Level II), with a 4-inch rim and a pink body painted in bright enamel colors appears to be Spanish majolica or French faience. This could not be verified on the basis of comparison. The saucer has a 3-inch foot-ring. One punch bowl has a rim diameter of 9-1/2 inches. A smaller bowl is 4-1/2 inches in diameter and 2 inches high with a 2 inch foot-ring. Others have an 8-inch rim diameter or a 4 inch diameter foot-ring.
Variety e. Decorated with powdered purple (aubergene) pigment. At least four bowls are decorated on the exterior with a powdered purple pigment. One of these also has a blue painted design on the interior. There may have been unpowdered reserves on the exterior but the sherds are too small to be sure. Two bowls are 7 inches in diameter.
Type 2. Lead-glazed refined buffware. This type is characterized by a cream-colored or buff-colored paste which has all impurities removed. It generally breaks with horizontal cleavage planes. About 1740, this paste was being covered with a light yellow lead-glaze colored with various metallic oxides (South, 1972, p. 85). By 1762, Josiah Wedgwood had perfected a clear lead-glaze and was producing cream-colored earthenware on a commercial basis within a short time (Noel Hume, 1970, pp. 125-126).
Variety a. Clear lead-glaze with no decoration except relief molding. This variety is usually called "creamware" or Queensware." As noted, it was in production by the mid-1760s. At least 81 vessels are represented in the collection: 16 plates, 12 saucers, 16 cups, eight teapots, 17 tankards, one pitcher and 11 bowls (fig. 63b-h). Ten of the plates have a molded feather-edge pattern on the rim dating from ca. 1773 (Noel Hume, 1972, pp. 350-355), and one is of the Royal pattern dating from ca. 1765 (Noel Hume, 1970 p. 125). Ivor Noel Hume (1970 pp. 126-128) (South, 1972, fig. 1) has observed that creamware is generally darker in color during the period 1762-ca. 1780 and that a lighter shade was introduced by ca. 1775. This observation, although subjective, seems to be borne out by the presence of both light and dark yellow varieties in pre-1781 contexts, but only light yellow specimens in ca. 1802-1825 contexts (Features 71 and 77) (Hanson, 1974, p. 40). However, it should be noted that many pre-1781 specimens are actually lighter than an undecorated pitcher from Feature 77 (see Ramsey 1962, Plate 308d, for vessel of similar shape).
Most of the cups, tankards and bowls are decorated with a beaded band on the exterior rim and some of the saucers have a gadrooned or bead and reel band around the interior bottom. One teapot, tentatively identified because it is missing the spout (fig. 63f), has a gadrooned band around the rim on the lid. One of the plain rimmed plates is 9 inches in diameter, while the feather-edged plates range from 8 to 10 inches in diameter, 1 to 1-1/2 inches high and have 5-inch bases. The saucers are 5 to 6 inches in diameter with 2-1/2-inch foot-rings. Cups are 3 to 4 inches in diameter and 2 inches high with 1-1/2- to 2-inch foot-rings. Creamware cups and saucers tend to be slightly larger than earlier stoneware Type 1a and 1b cups. This may reflect an increase in the practice of tea drinking. Teapots have 2-1/2- to 3-inch rim diameters and 3-1/2-inch bases. One is 4 inches high. A nearly complete tankard is 2-1/2 inches in diameter and 4 inches high with a 2-1/2-inch base. Others are 2-1/2 to 4 inches in diameter or with 2- to 3-inch bases; one pitcher base is 3-1/2-inches across. One bowl is 6 inches in diameter and 2-1/2 inches high with a 2-1/2-inch foot-ring. Others are 5 to 11 inches in diameter or have foot-rings 3 to 4 inches across.
Variety b. With over-glaze red paint. There are at least two saucers, one cup and three bowls decorated with red lines around the interior of the rim. Two saucer fragments include a 5-inch rim and a 2-1/2-inch foot-ring. One bowl has a 5-inch rim diameter.
Variety c. With over-glaze brown paint. There are a minimum of one saucer and one bowl of this variety. The saucer has a brown line on the interior of the rim, while the bowl is decorated with brown panels in a band of raised squares below the lip on the exterior. It appears to have been engine-turned. Both the saucer and bowl are 5 inches in diameter.
Variety d. With over-glaze polychrome paint. These sherds are decorated with floral motifs. Only one saucer is identified by vessel shape and it has painted red and black flowers.
Variety e. With a polychrome glaze (fig. 64a-c). This variety, which has been called "clouded" ware (Noel Hume 1970 p. 123) and under more specific names as "whieldon" ware and "tortoise shell" ware, was first manufactured about 1740 (South, 1972, p. 85). The clouding was accomplished by sprinkling or sponging oxides of copper, manganese and iron onto the wet glaze and allowing them to "run." This produced a mottled effect or an imitation of tortoise-shell. There are at least 22 vessels of this variety: one plate, one saucer, one cup, 12 teapots, four tankards and three bowls. The plate has a 10-inch diameter castellated rim and a 6-inch base, and is 1/2 inch high. The teapots have 2- to 3-inch rim diameters and 2-1/2- to 3-1/2-inch bases. The bowls are 5 to 5-1/2 inches in diameter and the latter is 2-1/2 inches high with a 2-1/2-inch foot-ring.
Variety f. With green tinted glaze. There are at least one saucer and one teapot of this variety. This pottery was first produced in 1759 by Josiah Wedgwood (Noel Hume 1970, p. 124). The rich green glaze was produced by adding copper oxides to the basic lead glaze. This saucer has a 4-inch rim diameter and the teapot has a 2-inch foot-ring.
Variety g, With orange glaze. Only one sherd of this ware was found. We are not sure how the glaze was colored, only that it is unique. It was found in the sally port (Level II).
Variety h. With brown glaze. Four sherds from the same vessel, probably a teapot, were found. The glaze is a glossy dark brown with a tendency to flake off.
Variety i. With an over-glaze black transfer print. One sherd with a black transfer print of a sailing ship was found in the ditch (Level XI). It probably came from a pitcher.
Type 3. Clear lead-glazed refined redware. This type is represented, at most, by three bowls. One of these has a 5-inch rim diameter, The vessel shapes are the same as for Type 2a bowls.
Type 4. Black-glazed refined redware. This type has been called "Jackfield" and was produced after 1740 (South, 1972, p. 85) (fig. 64f). The body is thin and usually has a gray-red or purple color with a glossy black glaze applied over it. A few of our specimens are oil gilded, but no designs could be identified. At least three saucers, 10 teapots, one sugar bowl and one bowl were found. The saucers have rim diameters of 4 inches. One teapot has a 2-1/2-inch rim diameter and a 2-inch base and is approximately 4 inches high. Others have bases 2-1/2 to 4 inches across. The sugar bowl has a 2-inch rim.
Type 5. Lead-glazed slipped coarse buffware. These are vessels with a buff-colored paste containing small grains of sand and grit. Occasionally, a sherd exhibits fine veins of red clay. The surfaces are coated with refined white, red and/or brown slips and are covered with a clear lead glaze. Usually there is no slip or glaze on the pedestal feet of bowls or cups, or the undersides of platters.
Variety a. With red slip and trailed white slips (fig. 65c). There are at least seven large platters of this variety, all with notched lips. A red slip was first applied over the upper surface of the platter; parallel lines of white slip were then laid over the red slip. The upper surface was then lead glazed, changing the colors of the slips to yellow and brown. Because these platters were rarely round, it is not possible to say more than that one was 2 inches high.
Variety b. With combed red slip and trailed white slip. One platter was made the same as Variety a, except that a comb like object was dragged across the surface of the platter at right angles to the slip lines before the glaze was applied.
Variety c. White slipped with trailed and dotted red slip (fig. 64d, e). There are at least 13 vessels of this variety, eight cups and five bowls. The bodies have slips like the platters, except that the white slip nearly obliterates the red, and a row of large red dots was added to the shoulder or rim. The pedestal feet are not slipped or glazed except where the glaze ran. Cups have one thick strap handle and the bowls have two, all on the shoulder of the vessels. The cups have rim diameters of 3 inches with 1-1/2- to 2-inch bases and one is 2 inches high. Bowls are 4-1/2 to 5 inches in diameter with bases 3 to 3-1/2 inches across.
Variety d. With marbled red, white and brown slips. One platter was found which has swirled red, white and brown slips intermingled to produce a marbled effect. It has a notched lip.
Type 6. Lead glazed coarse redwares. This type is largely restricted to utilitarian vessels. It is sometimes difficult to distinguish 18th and 19th century sherds, particularly Variety a. Distinctions were made on the basis of provenience and only those from pre-1781 contexts are described here. All are lead glazed and have inclusions of fine sand and grit in the paste.
Variety a. With no slip (fig. 65b). These vessels are coated with a clear lead glaze. There are at least two bowls, 23 pans and two platters. The pans have thick outrolled rims with straight, flaring walls and are pedestral footed. They are glazed only on the interior. The platters have notches on the lip, but finer than those on the coarse buffware (Type 5). They are glazed only on the upper surface. One bowl has a 5-inch rim diameter. The bowls are glazed on the interior and exterior. Rim diameters for pans range from 8-1/2 to 15 inches and one fragment of a 7-inch diameter base was found.
Variety b. With white slip. There is a minimum of one bowl and two pans of this variety. The slip, while generally applied, did not coat the entire vessel. The vessel shapes and glazing over the slip are the same as Type 6a. The bowl has a 3-inch-diameter base. The pan fragments are from a 7-inch rim and a 4-inch base.
Variety c. With trailed white slip (fig. 65a). These vessels are decorated with thin lines of white slip. There are at least one bowl, six pans and one platter. The pans are the same shape as Type 6a, and the platter the same as Type 5a. The decoration on the platter is parallel lines, while the lines on the interiors of the pans meander. In some cases the lines might have spelled words, but the sherds are too small for us to be certain. The trailed white slip meanders on the interior of the bowl, while on the exterior it is confined to one small area, suggesting a mark. We could make no sense of it. The bowl is 6 inches in diameter and 3 inches high with a 2-1/2-inch base. Two pan fragments include an 8-inch rim and a 6-inch base.
Variety d. With marbled white slip and metallic oxides in the glaze (fig. 65d). A single tankard of 2-pint capacity, had a swirled white slip covered by a lead-glaze with copper (green) and iron (brown) oxides sprinkled on it. It was 3-1/2 inches in diameter and 6-1/2 inches high with a base 4 inches across.
Variety e. With a copper-lead glaze and no slip. There are at least 16 vessels of this variety: one bowl, 13 pans, one jug and one jar. The pans are the same shape as Type 6a. The jug is a 3-1/2 inch base fragment only, and the jar is a 6-inch open mouthed crock. With the exception of the bowl, which is glazed on both surfaces, all are glazed on the interior only. Copper-lead glazes are common on French buffware vessels in Canada (author's examination). This variety and the following two are probably not of French origin, as the vessel shapes are different and the glaze is not as deep a green. The bowl has a 3-inch basal diameter. One pan has a 10-inch base and two others have 12-inch and 15-inch rim diameters.
Variety f. Copper-lead glazed over a white slip (fig. 65e). Only a single 9-inch bowl, 3-1/2 inches high, with a 3-1/2-inch base could be identified.
Variety g. Copper-lead glazed over a trailed white slip. Two pan fragments 7 inches and 12-1/2 inches in diameter were found.
Type 7. Black glazed coarse. These sherds superficially resemble Type 4, commonly known as "Jackfield," except that the paste is coarser and redder, the vessel walls are thicker, and the glaze is thick and not glossy. There are a minimum of one handled cup, three bowls, one pan, one drug jar and one jug in this type. Identical specimens, in the drug jar form, were found in the Weeden Brigade encampment at Valley Forge (author's examination). The cup has a 3-1/2-inch rim diameter and the bowls are 4 to 10 inches in diameter. The drug jar is 3 inches in diameter and the jug is 3-1/2 inches wide at the base with a 2-inch neck (1-inch opening).
Comparison of the rim and foot-ring dimensions shows that the proportions of plates, saucers, cups and bowls were fairly uniform regardless of the type of pottery or whether made in China or England. The foot-ring diameter is generally one-half to two-thirds as large as the rim diameter. This means that if we have only the foot-ring of one vessel form we should be able to predict the overall size of the vessel with a fair degree of accuracy using a conversion factor. Our sample of measurable vessels was too small to use for calculating conversion factors but the data suggests that plate rims were 1.50 to 2.11 times as large as the diameter of their foot-rings; bowl rims were 1.38 to 2.86 times the diameter of their foot-rings; saucers, 1.30 to 2.25 times as large; and cups, 2.00 to 2.67 times as large. We hope to test this hypothesis with the post-1781 sample of vessels from the site which is large enough to provide adequate samples of each vessel form. The utility of the conversion factor, if its validity can be shown, lies not so much in predicting rim diameters as in demonstrating a relationship, since foot-rings are more durable than vessel rims. It might also be possible to define some temporal changes in the rim to foot-ring proportions. For example, at Fort Stanwix, Type 1 earthenware bowls have rims 2.00 times the size of the foot-ring while Type 2 earthenware is 2.20 to 2.40 times as large, Type 5 earthenware is only 1.33 to 1.43 times as large and Type 6 earthenware is 2.40 to 2.86 times as large. Although the number of vessels involved is small, these observations suggest that there were different standards in the British ceramic industry, and if we can identify those standards we might be able to compare the output of different districts or manufactories.
There are five types of bottles, classified according to form. The most common are round free-blown bottles with tooled string rims. Next in popularity are square mold-blown "case" bottles with outrolled lips. Finally, there are round pharmaceutical bottles, octagonal mold-blown bottles and free-blown oval bottles with tooled string rims.
Round Bottles. By far the most common type of bottle is a round, heavy based, dark green bottle with a slight constriction just above the base, a pushed up base, or kick, and a tooled string rim (fig. 66a, b). The count on these specimens was based on the number of necks (table 24). Only one specimen found in Feature 60 (Level II) is complete. Its measurements are: base diameter, 10.7 cm.; kick height, 45 mm.; shoulder diameter, 11.1 cm.; body height, 13.7 cm.; neck height, 10.4 cm.; interior lip diameter, 20 mm. and exterior rim diameter, 36 mm. It has a sand pontil mark 44 mm. in diameter. Composite dimensions for fragmentary specimens are: basal diameter, 9.1 to 11.9 cm., average, 10.6 cm.; kick height, 18 to 52 mm.; average, 34.3 mm.; neck height, 7.7 to 11.0 cm.; average, 9.3 cm.; interior lip diameter, 14 to 20 mm.; average, 16.5 mm.; exterior rim diameter, 32 to 38 mm.; average, 34.1 mm.
Evidence in kicks was found for 37 sand pontil marks ranging in diameter from 36 to 64 mm. with an average of 53.2 mm. (fig. 67a). There are two glass pontil marks 29 mm. and 32 mm. in diameter (fig. 67b). Necks are about evenly divided between those which have slightly flaring fire polished lips with V-shaped string rims below them and those which had the lips pushed down against the V-shaped string rims (fig. 66d). They seemed to be randomly distributed and contemporaneous. Most of the kicks appear to have been formed with a mollette but one was made with a rod 20 mm. square, and another with a quatrefoil rod 30 mm. square (O. Jones, 1971, p. 66). One specimen deserves special mention and this is a bottle from the ravelin (Level II) which has the name "John Sh . . o . . ." scratched into its side (fig. 66b). We have been unable to trace the name to a known occupant of the fort.
Oval Bottles. These are free-blown bottles made in an oval shape rather than round (fig. 66c). The count is based on the number of bases (table 24). The one complete specimen from the north casemate (Level I) has a tooled string rim and is slightly shorter than the round bottles. The greatest width is halfway between the base and neck and there is no shoulder. They are quite similar to modern rose bottles. They have round sand pontil marks 46 to 53 mm. in diameter. The complete specimen and part of another measure 10.1 by 15.3 cm.; and 10.3 by 15.4 cm. in maximum width. Three kicks are 24 to 37 mm. deep. The base to neck height is 13.1 cm. and the neck of the complete specimen is 9.0 cm. long with an interior lip diameter of 15 mm. and an exterior rim diameter of 33 mm. These bottles are unique so far as we know, but are probably of English manufacture since they differ from the rest only in body shape.
Square Bottles. These have often been referred to as Dutch and as containers for gin (Noel Hume, 1970, p. 62) but the evidence for this is not conclusive. They could just as easily have been English in origin and contained any sort of liquid. The bottles were blown into a square one-piece mold with the neck and shoulder finished off later. The molds were tapered slightly toward the bottom to facilitate the removal of the bottles. Consequently, the measurements taken on the bases fall short of the actual width of the specimens. No complete or reconstructable specimens were found (table 24). Bases came in two sizes, ranging in width from 73 to 78 mm. with an average of 75.3 mm. and from 86 to 97 mm. with an average of 91.0 mm. The former (four specimens) have kicks of 4 to 8 mm. with sand pontil marks 38 to 47 mm. in diameter (fig. 67d). The remainder have kicks of 13 to 18 mm. with an average of 15.4 mm. and with glass pontil marks 21 to 30 mm. in diameter, averaging 24.4 mm. (fig. 67e). Several of the larger bases have mold marks preserved beneath the glass pontil marks. These include a variety of depressions and crosses. The necks of these bottles are short with everted lips thickened with additional glass (fig. 67c). Interior lip diameters range from 13 to 23 mm., averaging 15.1 mm. and the exterior rim diameters range from 33 to 41 mm. with an average of 37.3 mm.
Octagonal Bottles. These bottles were probably used for pharmaceutical purposes but a few were rather large and may have held liquor (table 24). Grimm (1970, p. 169) suggests that they held shoe blacking or ink. About all they have in common is that four sides are wider than the alternate four sides.
Variety a. Dark green glass. This variety consists of two base fragments, possibly from the same bottle and some side wall sherds. The bases have a sand pontil mark on a slight kick. The bottle resembles the 1770 specimen illustrated by Noel Hume (1970, p. 67).
Variety b. Light green glass. These are four base fragments, none the same shade of green. Only one rectangular specimen can be measured, 66 by 51 mm. This, and one other, have glass pontil marks 20 mm. in diameter. The other two have sand pontil marks 17 mm. and 27 mm. in diameter.
Variety c. Clear glass. This rectangular base fragment is too small for a decanter and is included here. It is 52 mm. long and has a glass pontil mark 15 mm. in diameter.
Pharmaceutical Bottles. The count on these was made on the bases (table 24). There are several shades of glass ranging from clear to dark green, with one blue-green rim fragment which may have come from a French bottle. Five of the bases, three light green and two clear lead-glass, have kicks 21 to 29 mm. high made with a pointed rod. These are 36 to 53 mm. in diameter with sand pontil marks 20 to 29 mm. in diameter (fig. 67g). Ten bases range in diameter from 24 to 46 mm. and average 30.8 mm. These have sand pontil marks 16 to 20 mm. in diameter with an average of 17.1 mm. (fig. 67f). Four bases, 18 to 44 mm. in diameter and averaging 29.0 mm., have glass pontil marks 10 to 16 mm. in diameter with an average of 13.3 mm. The lips are of two types, sharply everted (nine specimens) or vertical (four specimens) (fig. 67h). The interior lip diameters range from 9 to 28 mm., averaging 14.1 mm., not counting the probable French specimen noted earlier which has a vertical rim with a lip diameter of 41 mm.
Parts of seven decanters were found (table 24) five of which are clear lead-glass. One consists of a square base 86 mm. wide on one side with a 6 mm. kick and a glass pontil mark 22 mm. in diameter. Another is a neck fragment and a third is a neck with an interior lip diameter of 31 mm. (fig. 68b). Another two, while not complete, could be reconstructed from fragments. One is cylindrical with cut fluting around the base and neck and a sand pontil mark (fig. 68c). This came from the south ditch (Level II) and is included because it resembles an 1760-1780 style decanter illustrated by Noel Hume (1970, p. 197). The other (fig. 68a) is globular with a glass pontil mark. A stopper fragment (fig. 68d) was found on the parade ground (Level I). The other two are basal fragments of green mold-blown "chestnut"-shaped bottles (fig. 67i) with glass pontil marks. Their identification as decanters is tenuous.
These were counted on the basis of stem fragments (table 24). All were made from lead glass and all of the rim fragments appear to have come from trumpet-shaped bowls except one bell-shaped bowl (fig. 68f). One foot was folded at the edge (fig. 68g), and the rest were plain conical in shape (Haynes, 1959, p. 199) (fig. 68h). Five plain stem fragments are from clear drawn stems and four from heavy conical stems in which the trumpet-shaped bowl was a continuation of the stem (fig. 68e). Two air-twist stem fragments with a Z-twist (fig. 68i) had knops (bosses) formed by compressing the glass. There were two enamel-twist stems with a Z-twist, one of which had a central gauze twist as well (fig. 68j).
Fragments of the bases of five lead-glass tumblers were found (table 24), two of which have glass pontil marks 17 mm. across. One has 10 facets around the sides but the others are round. All are less than 40 mm. across at the base.
A mold-blown lead-glass beaker 14.2 cm. high and 10.1 cm. in diameter with a glass pontil mark and eight vertical ribs on the exterior (fig. 68k) was found in Feature 56 (Level IV) in the west barracks.
One mold-blown soda glass cup 32 mm. high and 65 mm. in diameter with a 6 mm. kick and a glass pontil mark was found in the guardhouse area (Level II). It is now opaque white but was once clear glass. It was blown in a mold and there are eight vertical ribs around the exterior (fig. 681). It may have been a lamp.
Four glass ink wells were found at Fort Stanwix (table 24). They range in height from 27 to 50 mm. and in diameter from 22 to 42 mm. All are blown glass with inverted lips (fig. 68m). How these ink wells were mounted had never been determined (Noel Hume, 1970, p. 75) but we were fortunate enough to find one intact in Feature 60 (Level III) (fig. 68n). This specimen is encased in a brass jacket with a threaded opening at the top. A brass cylinder was screwed onto the jacket which had a thick paper or cardboard insert in an expansion ring just above the neck to seal the opening when it was screwed on. The function of the remainder of the cylinder remains a mystery. A brass pin and partially hollow wood dowl which may have been a pin case or a stopper was found in it. A similar specimen was found at Fort Montgomery, New York (John Mead, personal communication). Four brass cylinders ranging from 30 to 70 mm. long were found on the site, but three were not associated with ink wells (table 24). Three of the reservoirs were found in the west barracks (Feature 3, Level IV and Feature 60, Level III). The largest was in the northwest corner of the north casemate (Level III) where it had probably been put for safekeeping. This specimen also had brass associated with it which was too poorly preserved to identify as a jacket.
Iron kettles were found concentrated in the casemates, barracks and dump areas (table 25). Only two brass kettles were found, one in Feature 51 in the southwest casemate and one in the ditch (Level I). In addition, a brass lid to a pot was found in Feature 57 (Level IV) in the west barracks.
Table 25. Distribution of kettle and collander fragments.
Type 1. Iron with legs. These have the following dimensions: diameter 5 to 12 inches, average 10-11/16 inches; height 5-3/4 to 8-1/4 inches. The largest pot fragment (fig. 69a) has a raised projection on the lip for a bale attachment and another has an angular handle below the rim. Five of the 14 rim fragments have ridges inside the rim to rest a lid. The largest fragment has a raised "C" on the side but no other markings were found. The smallest kettle has four legs while the rest apparently had three. These were generally welded to the nearly flat base and are D-shaped in cross section. The bases have a casting line or reinforcing ridge under them and reinforcing ridges are common on the sides of kettles.
Type 2. Brass.
Variety a. Without legs. The example from the southwest casemate (Feature 51) is a shallow basin approximately 9 inches in diameter with an outrolled lip. It has been badly crushed and partially melted. A bale lug or handle is attached to the rim at one point with two copper rivets.
Variety b. With legs. The specimen from the ditch (Level I) is a shallow basin approximately 8-1/2 inches in diameter with the lip rolled out over an iron wire. There was a handle (broken off) attached to the lip with three copper rivets. Three iron legs were also attached to the lip with two pairs of copper rivets.
A fragment of a pot about 8 inches in diameter was found in the sally port (Level II) which had its bottom beaten flat and holes punched through it to make a colander. It had been later cut apart and this piece apparently was scrap when discarded.
Category 1. One brass fragment from the ravelin (Feature 76) was found which had been crushed to form a crude cup. There were also two detached brass bale lugs; one from the southwest casemate (Level II) and the other from the guard house area (Level II).
On the floor of Feature 57 in the west barracks lay a brass pot lid 5-1/2 inches in diameter (fig. 69b). The lid consists of a top piece crimped over a circular band to form an insert for the lip of the pot. The handle is cast brass riveted to a square plate on the under side of the top. There is an upside down "IV" stamped on the band under the rim.
Fragments of tin cans were found throughout the fort but only one complete specimen, from the passageway to the northeast bombproof (Level II) was found (fig. 69c). This is 2-7/8 inches in diameter at the base and 2-3/4 inches high with folded seams which were probably sealed with solder. The top is missing. Some of the other fragments came from rectangular cans and might have been cartridge cannisters but are too corroded and bent to be positively identified.
Wood. The charred bases of two buckets were found in the west barracks (Feature 56, Level III). One is made from white oak (Quercus) and is 8-3/4 inches in diameter, while the other is made from eastern white pine (Pinus strobus) and is 9-1/4 inches in diameter and approximately 9 inches high. Comparable buckets were found at Fort Ligonier (Grimm, 1970, p. 55). These were apparently fastened together with wooden strips as no iron bands were found. Staves are 2 to 3 inches wide and 1/4 to 1/2 inch thick on one and 2-1/2 to 3 inches wide and 1/4 inch thick on the other. The base of the first fitted into a slot 7/8 inch from the bottom of the staves. The second base has been drilled with 23 holes unevenly spaced around the edge which indicates that it was being used for a purpose other than as a bucket. A few stave fragments from a bucket or barrel made from white ash (Fraxinus) were found on the floor of the northwest bombproof (Level III).
Tin. Two tinned bale lugs were found in the west barracks (Feature 60, Level III) and in the ravelin (Level II). The former was held to a bucket with a rolled rim by one rivet. Better preserved specimens were found at Fort Ligonier (Grimm, 1970, pp. 53-55).
Two types were found, iron and copper. The latter were for gunpowder barrels.
Type 1. Iron. Only two iron hoops were identified, one from the southwest casemate (Level II) and the other just north of the southwest casemate (Level II). These are 1-1/2 feet in diameter and 1-3/16 inches wide. The ends are held together with two iron rivets. Other fragments of strap iron were found in most other areas of the fort. These probably came from hoops but we could not be certain. See also: Broilers.
Type 2. Copper. These range in width from 1/2 to 1 inch, but no hoop diameters could be projected from the fragments found. The narrower straps had holes drilled through them and may not have been barrel hoops. Two of these were found in the north casemate (Level I) with one thin strap in Feature 57 (Level IV) in the west barracks, but most fragments were in Feature 69 (Levels II and III) in the east barracks. Many of these were partly melted and one example was found with two copper rivets. One of the north casemate examples was marked with a broad arrow. Grimm (1970, p. 98) reported hoops with diameters of 13-1/2 to 17 inches which were used on powder barrels.
These were wrought iron handles with an eye at one end for suspension and a small hole at the other end for attachment to a broiler (table 26) (fig. 70b). They have the following dimensions: length, 24.4 to 31.1 cm., average, 27.4 cm,; width at center, 12 to 33 mm,, average, 19.6 mm.; thickness, 3 to 7 mm., average, 4.3 mm.; diameter at suspension end, 24 to 45 mm., average, 34.9 mm.; eye diameter, 10 to 17 mm.; average, 13.1 mm.; diameter at attachment end, 21 to 36 mm., average, 27.2 mm. Two of these still have rivets through the end and a third has part of a hook through the eye. We believe that these were attached to griddles or broilers and used in cooking.
Assorted rods and straps of iron bent into S- or C-shaped hooks and presumably used in fire places to hang pots (table 26) (fig. 70c) have the following dimensions: length, 11.5 to 28.9 cm., average, 16.5 cm. Two appear to have been made from old door hinges.
Table 26. Distribution of griddle handles, pot hooks and broilers.
Type 1. Made from iron straps. Makeshift broilers to cook meat were made from barrel hoops. Two, from the southwest casemate (Level II) and Feature 68 are bent in a zigzag fashion while another from Feature 57 (Level III) is bent into a spiral.
Type 2. Wrought iron. A specimen from the ditch (Level I) consists of two wrought iron end pieces with legs, one with a handle, which is held together by eight rods approximately 30 mm. apart, welded through holes in the end pieces (fig. 70a). It stands 69 mm. high and is 20 cm. long and 25 cm. wide. The handle is at least 20 cm. long. "Grid irons" and food stuffs were sent in May of 1777 (Gansevoort, May 14, 1777) to Fort Stanwix.
Two fragments of cast brass ornamental plates from drawers were found (fig. 71a). One came from the west barracks (Level I) and the other from Feature 69 (Level I) in the east barracks. Both were in disturbed areas but their simplicity suggests an 18th-century origin.
Three drawer pulls, two from the sally port (Level II) and one from the ravelin (Level II) were found. They could also have been used for handles on chests or boxes. None are complete, only the handle having survived. Two are made from round rods 1/4 inch and 3/8 inch in diameter, and the third from a rectangular strip 1/8 by 1/4 inch. All three are approximately 3-3/4 inches wide and 1/4 to 1-3/4 inches deep.
A number of brass tacks were found (table 27).
Table 27. Distribution of brass tacks.
Type 1. Cast convex head with welded square shank (fig. 71b). These have the following dimensions: length, 11.5 to 19.0 mm.; average, 14.2 mm.; head diameter, 8.0 to 13.0 mm.; average, 11.9 mm. Twenty-four specimens were found in a one-foot square area in Feature 51 in the southwest casemate. These have convex heads and square shanks welded on and look like upholstery tacks. These were also found at Fort Ligonier, 1758-1766 (Grimm, 1970, p. 149).
Type 2. Hammered sheet brass, one-piece construction (fig. 71c). These have the following dimensions: length, 11.5 to 26.0 mm.; head diameter, 6.0 to 13.0 mm. The shanks are rolled sheet brass cones hammered flat at the top to form the heads.
Weights and Measures
Two fragments of the balance arm of a scale were found in Feature 63 (Level III) in the west barracks (fig. 71d). It is quite similar to the ones illustrated by Peterson (1968, pp. 167-169).
Part of a conical lead weight was found in Feature 60 (Level III) in the west barracks. It is 23+ mm. long with a diameter of 31 mm.
Several ruler fragments were found.
Type 1. Brass rule scaled in .1 inch increments (fig. 71f). Two fragments from the guard house area (Level II) and Feature 51 of the southwest casemate probably are from the same ruler. The scale is engraved on one edge of one side. On the opposite edge is another scale numbered in increments of six and tied to the first by diagonal lines. The number 12 is equal to 4.8 inches, 18 to 5 inches and 42 to 7 inches. This ruler is at least 8 inches long and was probably a gunner's instrument, although no comparable example was found in the literature.
Type 2. Copper rule scaled in 1/8-inch increments. This was found in the east casemate (Level II). It is a copper strip with iron pins 1/4 inch and 1-3/4 inches from one end which probably fastened it to a strip of wood. Originally it was at least 22 inches long.
Type 3. Wooden folding rule scaled in 1/8 inch increments. (fig. 71e). These are fragments of folding wooden rules with brass ends and pivots. They were found in the north casemate (Level III) and the ravelin (Level II).
A segment of a brass protractor scaled in 1-degree increments from 60 degrees to 90 degrees was found in Feature 51 in the southwest casemate (fig. 71g).
A spherical brass plumb bob 18 mm. in diameter with a hole at the top for suspension was found in Feature 51 in the southwest casemate (fig. 71h).
As the garrison was always short of clothing, it would have been necessary to make constant repairs. The only direct reference to sewing is to the making of a flag out of odds and ends of cloth by the officers in 1777. Presumably the officers mended their own clothes, although this may have been one of the jobs of their "waiters."
These have the following dimensions: length, 12.3 to 13.6 cm., average, 12.9 cm.; blade length to pivot, 6.2 to 7.2 cm., average, 6.6 cm. All are iron and have solid loop handles (table 28) (fig. 71k). The blades are isosceles triangles with one being broad and blunt while the other tapers to a point. Two have knops and balusters on the handles.
Table 28. Distribution of scissors.
Only three brass thimbles from the fort were found, one in the north casemate (Level III) one in the east casemate (Level II) and one in the east barracks (Level I). All three have a waffle-like or check-stamped crown with rows of round indentations on the side (fig. 71i). Two have practically no thickening of the opening, while the other is fairly thick. They are 18 to 20 mm. high and approximately 16 mm. in diameter on the interior of the opening.
Three steel needles were found, one each in the north casemate (Level II), headquarters area (Level II) and the sally port dump (Level II) (fig. 71j). These are 32 to 47 mm. long, averaging 9.0 mm.
Table 29 presents the distribution of straight pins on the site. All are brass with wire wound spherical heads and the better preserved specimens are tin plated (fig. 711). We included all specimens from the site except those from documented post-1781 contexts. Round headed pins continued in use on the site until ca. 1825 (Features 71 and 77) well after the abandonment of the fort, but this later occupation was confined to the southwest corner of the fort so that there is little chance that the data is skewed. Any skewing would be restricted to the ditch, ravelin, southeast casemate and Campbell collection. The pins range in length from 17 to 35 mm. with an average of 26.97 mm., and a standard deviation of 2.7911. This is somewhat smaller than pins from Fort Michilimackinac (Stone, 1970, p. 611).
Table 29. Distribution of straight pins. (Ratios given are the number of specimens per 10 square feet of excavated area within the structure.)
A total of 7,249 white clay pipe stem fragments were recovered plus a large number of bowl fragments, one ceramic colonial pipe, one ceramic Indian pipe and seven stone Indian pipes.
Series A. English white clay pipes.
The distribution and bore diameters of the pipe stem fragments are shown in Table 30. By far the greatest concentration was a cluster of 3,178 specimens in a 10-square-foot area in the vicinity of the guardhouse north of the southwest casemate (Level II) and spilling into the sleeper trench (Feature 48) along the front wall, which partially accounts for the high number in the southwest casemate as well. All of these appear to have been made by the same manufacturer using the same molds. See Type 9a below for more information on this cluster.
At Fort Ligonier, 1758-1766, (Grimm, 1970, p. 112) the mean pipe stem bore diameter is 4.61/64 inches, while the Fort Stanwix mean is 4.12/64 inches for the period 1758-1781. The sample at Fort Frederica, Georgia, 1737-1750, has a mean bore of 4.91/64 inches (author's examination). Specimens at Crown Point, 1760-1776, have a mean bore diameter of 4.23/64 inches (Gilbert Hagerty, personal communication). These samples merely confirmed what Harrington (1954) observed, that bore diameters decrease through time. Binford (1962) calculated a regression formula for converting the mean bore diameter of a sample to a median date for the sample. We have considerable reservation about the accuracy of this formula because it assumes a constant rate of change which does not seem to occur when plotting documented samples (the four cited above for example.) Nevertheless, Harrington's original premise is valid.
Internally, the Fort Stanwix samples had very consistent means and rather low standard deviations which suggests that the samples are homogeneous. That is: (a) all areas of the site were occupied at the same time, (b) the samples were thoroughly mixed prior to 1781 or (c) there was little discernible change in bore diameters during the period 1758-1781. We can dismiss (a), above, since we have documentary evidence that the bombproofs were not built before 1776 (W. Willett, 1831, p. 49). Explanation (b) is a strong possibility since the samples from the bombproofs, on the east scarp (Level II) and in the sally port (Level II) can be dated post-1776 on the basis of documents and Revolutionary War buttons. These samples do not differ from other areas of the fort. The evidence, however, favors (c). There was little discernible change in bore diameters between the four sites noted during this period between .3/64 inches per year and .6/64 inches per year. These figures were arrived at by dividing the difference between the means of the samples by the difference between the median dates of the sites, and checked by dividing the standard deviation of each sample by the number of years the sample was accumulating (in these cases, the total time span of the sites). At Fort Stanwix, then, we would expect the mean bore diameter to shift between 4.45/64 inches to 4.81/64 inches for 1758 and 3.43/64 inches to 3.78/64 inches for 1781. In short, any mean within these limits could be considered normal.
The bowls could be typed on the basis of maker's marks and bowl configuration (table 31). These types are not intended to encompass all forms and are based solely on Fort Stanwix specimens. See Hanson (1971a, pp. 92-99) for post-1781 types from the site. The types in this series are defined on the basis of the presence or absence of markings on the bowls. Because most markings are initials (presumed to be the initials of the makers) the procedure in the past has been to judge the age of a pipe by its shape, and then ascribe the pipe to a pipe maker with those initials in that period. This method has its pitfalls and we used it with reservation to identify the source of the pipe, not to date it. No recognizable Dutch or French specimens were found in a pre-1781 Fort Stanwix context, although they occurred later at the site.
Heels are present on only 24 percent of the bowls studied and took two forms, a short truncated heel and a long conical heel which might also be referred to as a spur. The following measurements are used to describe the pipe bowls: bore diameter was taken with the butt ends of drill bits in 64th-inch sizes because standards have been set up on this system (Harrington, 1954). Bowl angle refers to the angle between the back of the bowl and the stem. Where the stem was too broken to get a line, a drill bit, securely fitted into the bore, was substituted. Although the total range was 54 to 77 degrees the bulk fell between 66 degrees and 73 degrees. As a general rule, pipe bowls were at a more acute angle to the stem in 19th-century privies on the site than in the pre-1781 components, although there were a few exceptions. Pipe bowls from the 19th-century privies on the site of Fort Stanwix rarely are less than 70 degrees to the stem. Bowl depth was measured from the lip of the back of the bowl to the bottom of the interior of the bowl. These range from 24 to 41.5 mm. in depth with an average of 32.5 mm. A ca. 1840-1860 sample (Feature 5) ranges from 26.5 to 45 mm., with an average of 32.0 mm., not significantly different from the Fort Stanwix sample. Length and width of the bowl were measured on the interior of the bowl at the lip. The bowls from the fort tended to be slightly longer than wide, while the Feature 5 bowls tended to be round or slightly wider than long and generally larger. The comparative dimensions are: Fort Stanwix: length, 15.5 to 20 mm., average, 17.1 mm.; width, 11.5 to 15.5 mm., average, 14.2 mm.; Feature 5: length, 15 to 22 mm., average, 17.8 mm., width, 15 to 21.5 mm., average, 17.4 mm. It can readily be seen that the real change in size took place in the width of the bowl rather than the length or depth. This was consistent enough that we suggest it can be used as a means of preliminary identification between 18th and 19th century specimens.
Type 1. Unmarked pipes.
Variety a. Heeless (fig. 72a). These have the following dimensions: bore diameter, 4/64 inches, bowl angle, 61 to 77 degrees, average, 67.7 degrees; bowl depth, 29.5 to 33.5 mm., average, 31.7 mm.; bowl length, 15.5 to 19 mm., average 17.2 mm.; bowl width, 11.5 to 15.5 mm., average, 14.1 mm.
Variety b. Conical heel. These have the following dimensions: bore diameter, 4 to 5/64 inch; bowl angle, 59 to 63 degrees; bowl depth, 33 to 35.5 mm.; bowl length, 20 mm. Evidence from Fort Michilimackinac (L. Stone, 1970, p. 411) suggests this type of heel was in use ca. 1750-1780.
Variety c. Truncated heel. This has a bore diameter of 4/64 inch. The bowl is warped. This specimen came from the south ditch (Level II) and might be 19th century. Stone (1970, p. 411) suggests that this type of heel dates ca. 1730-1760.
Type 2. "RT" pipes. The initials "RT" have generally been attributed to Robert Tippett of Bristol, England and his family, ca. 1660-1720 (I. Walker, 1971a, pp. 19-22), but the initials were used at Fort Stanwix after 1958 and, as in Variety e, even the name "TIPPET" appears on some bowls. I. Walker (1971b, p. 73) notes that Tippet pipes have been found at Fort Gaspereau, New Brunswick (1750-1756), and in a shipwreck in Baie-des-Chaleurs (ca. 1760). It seems likely, therefore, that the Tippet molds were being used into the 1760's, although not necessarily by Tippets. Stone (1970, p. 404) suggests dates of 1740-1780 based on specimens from Fort Michilimackinac.
Variety a. "RT" on the back of the bowl (fig. 72b). These came from the west casemate (Level I), the north casemate (Levels I and IV), the northeast bombproof (Level II), the east scarp (Level II) and the ravelin (Level I). They have the following dimensions: bore diameter, 4/64 inch; heel, none; bowl angle, 64 to 70 degrees; bowl depth, 33 to 37 mm. One also had "TIP,PET" in a circle on the right side of the bowl.
Variety b. "R°ree;T" in heart flanked by hearts on both sides of bowl (fig. 72c). These came from the guardhouse area (Level II) and the sally port (Level II) and have the following dimensions: bore diameter, 4/64 inch; heel, none; bowl angle, 68 degrees; bowl depth, 33.5 mm.; bowl length, 16.5 mm.; bowl width, 14.5 mm.
Variety c. "R, TIP, PET" in sunburst circle on right side of bowl (fig. 72d). This came from the southwest casemate (Feature 50) and has the following dimensions: bore diameter, 4/64 inch; heel, none; bowl angle, 72 degrees; bowl depth, 32.5 mm.
Variety d. "RT" in heart on right side (fig. 72e). These came from Feature 56 (Level II) and Feature 57 (Level IV) and have the following dimensions: bore diameter, 4/64 inch; heel, none; bowl angle, 68 degrees; bowl depth, 32 mm.
Variety e. "RT" with three dots above and below in beaded circle on right side (fig. 72f). This came from Feature 76 and has the following dimensions: bore diameter, 5/64 inch; heel, none; bowl angle, 69 degrees; bowl depth, 34 mm.; bowl length, 17 mm.; bowl width, 15 mm.
Variety f. "RT" in a heart on the right side with flowers around the bowl (fig. 72g). This came from the southwest bombproof (Level II) and has the following dimensions: bore diameter, 4/64 inch; heel, none; bowl angle, 68 degrees; bowl depth, 32.5 mm. Although the size agrees well with other Tippet pipes, the decoration is totally unlike other Tippet products and the letters "RT" are much larger. This is probably a late 18th-century use of the initials by another maker.
Type 3. "TD" pipes. The use of the initials "TD" on pipe bowls dates back to at least 1730 (I. Walker, 1966), and they are still being made. However, changes in the shape of the bowls and the motifs surrounding the initials are clues to the age of the pipes. None of the known English pipe makers with these initials were operating at the time Fort Stanwix was in use, and it would seem that most "TD" pipes were not made by people with those initials (I. Walker, 1966). Indeed, it has not yet been proven that anyone with these initial made "TD" pipes. All we can say, at this point, is that the shape of these bowls suggest English origin and, therefore, the marks are also probably English. Varieties a and b have been found at Fort Ligonier, 1758-1766 (Grimm, 1970, p. 114).
Variety a. "TD" with curliques above and below in a rope circle on the back of the bowl (fig. 72h). These came from the south ditch (Level XI) and the southwest casemate (Feature 50) and have the following dimensions: bore diameter, 5/64 inch; heel, none; bowl angle, 68 degrees; bowl depth, 41.5 mm.
Variety b. "TD" with curliques above and below in a circle on the back of the bowl. No measurements were possible on this specimen from the north casemate (Level II). Stone (1970, p. 403) suggests a post-1750 date for this type from Fort Michilimackinac.
Variety c. "TD" over a fleur-de-lis in a rouletted heart on the back of the bowl, raised hearts on both sides of the heel (fig. 72i). These were found in Feature 48 and Feature 56 (Level III) and have the following dimensions: bore diameter, 4/64 inch; heel, truncated; bowl angle, 66 to 70 degrees; bowl depth, 29.5 mm.; bowl length, 15.5 mm.; bowl width, 15 mm.
Type 4. "WM" pipe: "WM" with curliques above and below in a rouletted circle on the back of the bowl (fig. 72j). This came from Feature 73 and has a diameter of 5/64 inch and a bowl angle of 66 degrees. This mark has been found on pipes from Williamsburg, ca. 1750-1765, (A. Noel Hume, 1963, p. 22) and Louisbourg, 1747-ca. 1760 (Hanson, 1968, p. 20).
Type 5. "IC" pipes.
Variety a. "IC., 1" in a rouletted circle on the back of the bowl (fig. 72k). This was found in Feature 63 (Level II) and has the following dimensions: bore diameter, 5/64 inch, heel, none; bowl angle, 65 degrees; bowl depth, 30 mm.
Variety b. "I.C., 14" in a rouletted circle on the back of the bowl (fig. 72l). No measurements were possible on this specimen from Feature 3 (Level III).
Type 6. "WN" pipe: "WN" in a circle on the right side of the bowl. No measurements were possible on these specimens from the southwest casemate (Level II), the northeast bombproof (Level II) and the east scarp (Level II).
Type 7. "WO" pipe: "W" and "O" on sides of heel. This was found in the south ditch (Level II), the north casemate (Level I) and the northeast bombproof (Level II). They have the following dimensions: bore diameter, 5/64 inch; heel, truncated; bowl angle, 66 degrees.
Type 8. "T" pipe: "T" in a sunburst diamond. No measurements were possible on this specimen from the parade ground (Level I).
Type 9. Pipes with symbols instead of lettering.
Variety a. "+" on right side of bowl (fig. 72m, n). These have the following dimensions: bore diameter, 4/64 inch; bowl angle, 66 to 70 degrees, average, 68.5 degrees; bowl depth, 34 mm. Three of these are associated with the large cluster of pipe stems between the southwest casemate and the west barracks (Level II), three came from Feature 56 (Level II) and the remaining two from the southwest casemate (Level II and Feature 48). Both the bowls and the majority of the stems are crude. Inclusion of gravel in the clay, which sometimes fell out after firing left holes through the stem or bowl and there are air bubbles along the upper surface of the stem and back of the bowl. The mean of the bore diameters for this collection is 4.12/64 inch with a standard deviation of .335. All of the bits had been trimmed with a knife before the wire was withdrawn. A count of the bits indicates approximately 441 pipes in the cluster. If these were all made by the same person, he was using wires with diameters of 3/64 to 5/64 inches. Two 6/64-inch specimens in the cluster were probably intrusive, although pre-1781, and all the 3/64-inch specimens could have come from one pipe. Eliminating the 6/64-inch specimens would not change the mean and would lower the standard deviation to only .330. This cluster probably represents pipes broken in shipment and discarded at the site.
Variety b. A cross of five nodes on the right side of the bowl. No measurements were possible on this specimen from Feature 48.
Category 1. Marked fragments which could not be identified (12 specimens). Eight of these have a circle on the right side of the bowl, some of which has illegible lettering; one has a fragment of a circle and a curlique on the back like Type 3b; one has a sunburst circle fragment like Type 2c; and one is marked "?" and "W" on either side of a truncated heel.
Category 2. Fragments which were too broken to tell if they were marked (26 specimens). Seven specimens have conical heels and are probably Type 1b; the rest had no heels.
Series B. Colonial Clay Pipe.
Type 1. Unmarked red clay pipe with a heel (fig. 72q). This has the following dimensions: bore diameter, 5/64 inch; bowl angle, 65 degrees; bowl depth, 34.5 mm.; bowl length, 20.5 mm.; bowl width, 20 mm. This is made from a refined clay fired to a red-orange color. It was found on the east scarp (Level II) and may have come from Virginia.
Series C. Indian Clay Pipe.
Type 1. This was patterned after the stone Micmac pipes described below (fig. 72r). The base and stem hole are broken off but the bowl is 23 mm. in diameter and 25 mm. deep. It came from the north casemate (Level I). It is poorly fired and pinkish-gray in color. It is decorated with four incised circles (probably made with a compass) separated by groups of vertical incised lines, three groups of four lines and one group of five lines, with a row of punctations around the outside of the lip.
Series D. Indian stone pipes.
Type 1. "Micmac" pipes (fig. 72s). Five of these are made of fine grained sandstone and one of dolomite. These came from the north casemate (Level II), Feature 3 (Level III), the east casemate (Level II), the sally port (Level II) and the south ditch (Level XII). The most complete specimen is 60.5 mm. high with a bowl diameter of 19 mm. and a bowl depth of 21.5 mm. The bowl depth of two others is 22.5 mm. The nearly complete specimen has four drilled depressions on the bowl separated by groups of vertically engraved lines, three groups of four lines and one group of five lines opposite the side with the stem hole. There are engraved lines around the lip and bowl base. On both sides of the lower portion containing the stem hole are shield-shaped engraved lines bordered on the inside by drilled holes with a hole drilled transversely below the stem hole, probably for suspension. Two of the other sandstone bowl fragments have engraved lines around the lip and base and one has triangles engraved near the base with drilled depressions at the apex of the triangles. The dolomite specimen is undecorated.
Type 2. "Imitation European" pipe (fig. 72t). This specimen is made from coarse sandstone and is an elbow pipe with a vestigial "heel" at the front of the bowl. It is too broken to obtain measurements, but was obviously made with a European pipe as a model. It came from the north casemate (Level I).
Knives and Razors
Several types of knives were found (table 32). For the most part these are too corroded and fragile to be cleaned and measurements are estimated allowing for the corrosion. All blades are made of steel and are wedge-shaped in cross section.
Table 32. Distribution of knives, forks and spoons.
Type 1. Blade folded into the handle (clasp knife) (fig. 73a, b). These have the following dimensions: small: handle length, 10.9 to 15 cm., average, 12.1 cm.; blade length, 8.8 to 12.2 cm., average, 10.5 cm.; blade width, 15 to 24 mm., average, 20.3 mm. Large: handle length, 17.7 to 18.5 cm.; blade length, 16 to 16.8 cm.; blade width, 21 mm. There is no difference in the large and small clasp knives except size. All have six parts: a blade, two side plates, a divider bar along the upper side between the plates which formed a slot for the blade and served as a spring to lock the blade open or shut and two bone splints for grips which are riveted to the side plates. These bone grips are usually carved with parallel notches and two have initials carved on them. One specimen of an excised brass grip was found in the guardhouse area (Level II) and similar specimens have been found at Fort Michilimackinac and may be French (L. Stone, 1970, p. 502). The blades are notched and squared at the butt to engage the divider bar and the blade tip is convex on both the upper and cutting edges. One blade from the sally port (Level II) has "LUCA" stamped on it near the pivot.
Type 2. Blade has a square tang on it for halfting (sheath knife).
Variety a. Tang an extension of the upper edge of the blade (fig. 73c). These have the following dimensions: blade width, 17 to 30 mm., average, 22.4 mm.; tang length, 39 to 55 mm.; average, 51.0 mm. Other measurements are not possible. At least one is hafted with a wooden handle. The blade tips are straight or slightly concave on the upper edge and convex on the cutting edge.
Variety b. Tang centered on the blade (fig. 73d). These have the following dimensions: blade width, 19 to 25 mm., average, 22.0 mm.; tang length, 36 to 61 mm., average, 50.7 mm. No other measurements can be taken. This variety might be table knives (Type 3) rather than sheath knives but it is impossible to tell without a complete specimen.
Type 3. Blade with upturned blunt tip riveted to handle (table knife) (fig. 73e, f). These have the following dimensions: length, 24 cm. and 25.7 cm.; blade width, 19 to 25 mm., average, 22.0 mm. One example (fig. 73f) has a square tang but is otherwise identical. The handles are of the "pistol grip" type held to the flat tang by two to three rivets. There is usually a short ferrule between the handle and blade. These are somewhat larger than similar specimens from Fort Michilimackinac (L. Stone, 1970, p. 506).
Type 4. Small blade with square tang (surgical knife) (fig. 73g-i). These have the following dimensions: length, 9.5 cm.,? and 14.5 cm.; tang length, 26 mm.,? and 46 mm.; blade width, 8 mm., 13 mm. and 18 mm. These are tentatively identified on the basis of their unusual shape and the high quality of the steel.
Type 5. Narrow blade folding into a handle (razor) (fig. 73j). These have the following dimensions: blade length, 7.9 to 13.2 cm., average, 11.4 cm.; handle length, 8.6 cm. for the shortest specimen; blade width, 15 to 18 mm., average, 16.2 mm. These are constructed similarly to clasp knives; handles have bone grips and the blade tips were squared.
Type 6. Socketed blade (function unknown) (fig. 73k). This has the following dimensions: length: 24.7 cm.; socket length, 11.7 cm., blade width, 28 mm. This was found in Feature 51 in the southwest casemate and is probably a heavy cutting knife.
Type 1. Steel with bone handles (table 32).
Variety a. Two round tines with a pistol grip handle (fig. 74a). These have the following dimensions: length, 17.5 cm. and 18 cm.; tine length, 32 to 64 mm., average, 46.5 mm.; width at tines, 13 to 19 mm., average, 15.7. All are two-tined iron forks with "pistol grip" bone handles riveted to the tang except one, which has a two-piece antler handle with the tang flattened over the butt. This has two iron rivets while the others have three. Three specimens which have lost their hafts have square tangs for a drilled handle, but the rest have flat tangs with plano-convex bone plates riveted to them. Handles are generally plain or have small clusters of dots on them.
Variety b. Two flat tines with a ridge across the back of the tines and a straight handle (fig. 74b). These have the following dimensions: length, 15.8 cm., tine length, 26 mm., width at tines, 14 mm. The handle is two flat bone slats held to the flat tang by two rivets. The handles are engraved with chevrons pointing toward the butt except one which has the chevrons stamped into the bone. This variety occurred only in the upper level of the site and might, therefore, be post-1781. A similar example was found in a mid-19th century privy (Feature 5) which contained several early 19th-century artifacts.
Spoons (table 32).
Type 1. Small bowls with thin handles (teaspoons).
Variety a. Pewter. Only a fragment of a spoon bowl is present with a bowl depth of 4 mm.
Variety b. Brass or latten (fig. 74d). These have the following dimensions: bowl length, 34 to 36 mm.; bowl width, 17 to 19 mm.; bowl depth, 3 to 5 mm. These are all spoon bowls with the handles braized to the bottom of the bowl. A silver example of this type at Fort Michilimackinac (L. Stone, 1970, p. 463) had a total length of 10.7 cm.
Variety c. Silver plated brass (fig. 74e). Only a handle fragment was found with an upturned end.
Type 2. Large bowl with thick handle (serving spoons) (fig. 74c).
Variety a. Pewter. One bowl has a depth of 9 mm. No other measurements are possible. Three handle fragments have a central ridge on the top near the end. The bowls are oval. One bowl has engraving in it which appears to be either a small gabled roofed structure, a monogram surmounted by a "4" or a merchant's mark.
Variety b. Brass or latten.
One bowl has a depth of 9 mm. No other measurements are possible. Two bowl fragments and two handle fragments were found. The bowls are more egg-shaped than the pewter variety.
Variety c. Silver plated brass. Only a handle fragment was found.
These are all made from bone with notching to form teeth. None are complete specimens (fig. 75a). As far as we can tell, they are all toothed on two opposing edges. The number of teeth per inch ranges from 14 to 35. Two specimens have two sizes of teeth, 24 and 35 per inch on one, and 14 and 28 per inch on the other. One was found in the north casemate, (Level I), two in the southwest casemate (Feature 48 and Level II) and two in the west barracks (Feature 57, Level IV and Feature 60, Level II).
Three glass mirror fragments with silver backing were found, two in the southwest casemate (Feature 48 and Level II) and one in the sally port (Level II). All are too fragmentary to obtain meaningful dimensions. Several other mirror fragments were found on the site but they are probably of post-1781 origin.
Five hat pins (fig. 71m) were found in the north casemate (Level II), west casemate (Level II) and Feature 56 (Level III), Feature 48 and the sally port (Level II). The only complete specimen is 47 mm. long.
There were women at the fort who washed the laundry of the garrison. They were probably wives of soldiers who had to work in order to stay with their husbands, and to augment the meager wages of the soldiers. With 300 to 400 men stationed at the fort, there was plenty of laundry to keep the women busy, especially in light of the following order:
There are no specific references to ironing uniforms, but trousers and shirts made of broad cloth or linen needed to be ironed in order to retain their shape. The sad irons were probably the personal property of the women.
Three sad irons, of a shape generally associated with the early 19th century (Glissman, 1970), were recovered (fig. 75b). The largest is a surface find, the middle sized specimen came from the east scarp (Level II) and the smallest was on the floor of Feature 51 in the southwest casemate. Thus, two are from definite 1758-1781 contexts. The one found on the surface has a raised "7" on the top immediately behind the front handle attachment. Handles of all three were welded to the top of the sad iron near the front and back and had been broken off prior to their discovery. All the irons were probably sand cast. In profile, the smallest and largest irons have four tiers and the middle sized iron has three tiers. They measure 5-1/16 inches long and 3-1/8 inches wide, 5-1/4 inches long and 3-1/4 inches wide, 6-1/8 inches long and 3-7/8 inches wide and all are 1-1/8 inches thick.
A scrub brush of white oak (Quercus) was found in the sally port (Level III). It is over 11 cm. long and approximately 6.6 cm. wide and 8 mm. thick. The bristles appear to be of pig, held in place by copper wires in holes 6 mm. in diameter and spaced 11 mm. apart.
Coins and Tokens
A total of 51 coins were found. Of these, 43 are British copper halfpennys and eight are Spanish silver coins (fig. 75e-k). Campbell (1965) listed a 1743 British coin found in the bakehouse which is no longer in the collection. Table 33 presents the distribution of the coins and table 34 shows the range of dates. It will be noted that, with one exception, all the Spanish coins are clustered at the latter end of the time sequence, and the only coin minted during the Revolution is Spanish. This is not surprising since no British copper coins were issued from 1755 to 1770 (Noel Hume, 1970, p. 162). We included the dimensions of dated coins on table 34 and it can be seen that there is no consistent change in British halfpennys (which made up the bulk of the collection) over the time span represented by the collection. Therefore, one cannot date coins by their diameter when the face is worn off. The earliest Spanish coin (1699) is a silver disc stamped on both faces, denomination unknown. The other dated Spanish coins are one and two reales and the two undated coins were pie-shaped pieces cut from large Spanish dollars and weighed slightly less than one half real and one real. The larger piece (they are not from the same coin) is stamped with an "M" on one side and a broad arrow on the other. All the dated Spanish coins were minted in Mexico City.
Table 33. Distribution of coins. (Ratios given are the number of specimens per 10 square feet of excavated area within the structure.)
Table 34. Sizes of dated coins in millimeters.
One French or Belgian specimen was probably a jetton (Noel Hume, 1970, pp. 171-173) rather than a coin and is made of brass (fig. 75l). On the obverse it has a large block "L" surmounted by a crown with a fleur-de-lis on each side and the bottom. The legend reads: "LUD•XV•D•G•FR•ET•NAV•REX". The reverse has a vine scroll surmounted by a crown and the legend: "(B)ENEDICTUM•1700•.SIT•NOM•DOM. It has a diameter of 21.8 mm. and a thickness of .5 mm.
These are lead pieces, probably used as checkers or counting devices and were locally made (fig. 75m). Six of these came from the north casemate, (Feature 30 and Levels I, II and III), two from the ditch (Levels X and XV), and one each from the southwest casemate (Level I) and the sally port (Level II). Six are square, three round and one oval. Eight are cut on one or both faces with X's or symbols, or impressed with but tons or a Spanish silver coin before the lead hardened. The impressed specimens are too rough for button molds and are not graduated as weights should be, although they might have served that purpose.
A single brass jews harp with a diamond shaped cross section was found in the north ditch (Level X) (fig. 75c). The iron tongue is broken off. It is 53 mm. long, 24 mm. wide and 6 mm. thick.
A carved bone whistle was found in Feature 60 (Level III) in the west barracks (fig. 75d). It is 10 cm. long with a diameter of 15 mm. It was whittled with a knife and smoothed with a file. It has been burned.
Two brass bells were found in fort contexts. There are three others, larger than these, which probably were post-1781 and were found in the ditch (Levels I and X). The latter have engraved loops and "U"-shaped shanks like the one illustrated by Noel Hume (1970, p. 58). Two post 1781 examples are marked with the initials "WB". Bells found at Fort Ligonier (Grimm, 1970, p. 50) were an entirely different type.
Type 1. Cast with drilled shank (fig. 75p). This has the following dimensions: diameter, 30 mm.; shank height, 10 mm., shank width, 11 mm. The specimen is cast in two halves and braized together at a sharply thickened seam. The shank is a solid piece soldered to the two and drilled with a small hole. There are two holes connected by a slot in the lower half and two holes in the upper half. It has iron clackers. It came from Feature 72 (Level II) in the east barracks.
Type 2. Stamped with a wire shank (fig. 75n). This has a diameter of 21 mm. and a shank height of 7 mm. This specimen is thin brass that was apparently stamped in two halves, each of which was ground by turning and then braized at the seam. A wire loop shank is soldered to the upper half. There are four lines engraved around the sides and a broad arrow is stamped on it. It came from Feature 52 (Level IV) in the east barracks.
Lead pencils were fairly common on the site. There are two types, one made from scrap lead and the other hammered out of musket balls (table 35).
Table 35. Distribution of lead pencils.
Type 1. Scrap lead pencils.
Variety a. Round cross section (fig. 75q). Dimensions: length, 69.0 to 89.5 mm., average, 76.8 mm.; diameter, 3.0 to 7.5 mm., average, 4.5 mm. Most specimens taper to a flat tip at one end and are blunt at the other end. One is drilled for suspension at one end.
Variety b. Square to rectangular cross section (fig. 75r). Dimensions: length, 33.0 to 115.5 mm., average, 65.6 mm.; width, 2.5 to 12.5 mm., average, 6.2 mm.; thickness, 2.0 to 11.0 mm., average, 5.2 mm. These generally have blunt or pointed tips. Two are twisted, probably for greater strength and one is drilled for suspension. They frequently show hammer or trimming marks.
Variety c. Thin rectangular cross section (fig. 75s). These have the following dimensions: length, 43.5 to 83.5 mm., average, 62.2 mm.; width, 5.0 to 12.0 mm., average 7.3 mm.; thickness, 1.0 to 3.5 mm., average, 2.0 mm.
Type 2. Musket ball pencils (fig. 75t). These have the following dimensions: length, 44.5 to 60.0 mm., average, 53.2 mm.; width, 11.0 to 14.5 mm., average 12.5 mm.; thickness, 5.5 to 10.5 mm., average 8.9 mm. These were musket balls hammered to a blunt point on one end and retaining the musket ball shape at the other.
Four glass signets were found in the north and east casemates (Level II), the west barracks (Level I) and the sally port (Level II). The specimen from the north casemate is a clear glass oval, 14 by 12 mm. with the impressed bust of a man in Roman garb and a rope border. (fig. 75v). It is mounted with red wax in a brass stamp 29 mm. long. It is not a ring and was probably worn attached by a string through a hole in the mount. The one from the east casemate is 11 mm. in diameter with an impressed large house flanked by trees. The specimen from the west barracks is also 11 mm. in diameter with an impressed rosette pattern of seven flowers. These two might have been insets from sleeve links. The sally port specimen is an oval blue glass setting measuring 16 by 14 mm. It has the design of an impressed four-masted ship (fig. 75u). The back and edges are faceted. These signets were probably used with hot wax to seal letters and documents.
A single used lead bale seal was found in the ditch (Level XI) south of the southwest bastion. It consists of two discs connected by a strip. One disc has a hole in it and the other a knob on the back which has been pressed through the hole. The obverse is shown in Figure 78w. The reverse has a bisected circle with a "25" molded in the lower half. In the upper half is scratched the number "792" or "292". It is 30 mm. in diameter and may have been British, although what it sealed is unknown. Philip Schuyler (letter dated July 6, 1777) responded to a letter of John Hansen, the fort's commissary, about two bales of cloth for the Indians that were sent from Boston by the firm of Livingston and Trumbull and were numbered 172 and 173. See also: Ink Wells
This section is a catchall for those objects we could not identify as to function but which were considered important enough to describe. Many iron objects and lead and brass scraps were encountered which had no apparent function. Most of the iron specimens are probably fragments of tools. Most of the lead specimens appear to have been worked on by persons with nothing better to do and most of the brass specimens appear to be scrap.
Wrought iron ferrules were used to reinforce pieces of wood, especially ends, to prevent them from splitting. Also, they were used to keep wood or bone handles clinched tightly over the tangs of metal implements. They were all made from strips of metal cut to the desired length, bent to the desired shape and the ends welded together. The welds were skillfully done and on some specimens are very difficult to detect. Specimens identical to these types were found at Fort Ligonier (Grimm, 1970, p. 53).
Type 1. Rectangular.
Variety a. Large with rounded corners. Two specimens were found (fig. 76a). They are seven to eight times larger than the other ferrules. One fitted on a piece of wood approximately 2-1/4 by 3-3/8 inches and the other on one 2-1/4 by 4 inches. The smaller ferrule is 5/8 inch wide and 3/16 inch thick; the larger specimen is 1 inch wide and 1/8 inch thick.
Variety b. Small with no attachment holes. Two specimens were found. One is square and the other rectangular. The rectangular ferrule measures 1-1/4 inches long and 3/4 inch wide. The strap is 1/2 inch wide and 1/16 inch thick. The other is 1-3/8 inches square made from a strap 1/2 inch wide and 1/16 inch thick.
Variety c. Small with one attachment hole. Five specimens were found (fig. 76b). Four are square and one is rectangular. Three of the square ferrules measure 1-1/8 inches on a side, the other is 1-1/4 inches on a side. They are made from 3/8-inch-wide strips, 1/16 inch thick. Three still have nails up to 1/2 inch long through a hole bored in the center of one side. The rectangular specimen measures 3/4 by 1-1/4 inches and is made from a strip 1/2 inch wide and 1/16 inch thick. A nail hole is in the middle of one short side.
Type 2. Bell-shaped. Two were found (fig. 76c). They have a flat bottom, two sides that taper inward and a rounded top. They are identical in size and shape; the bottom measures 1-1/4 inches long, the sides 1-1/2 inches long, and the top approximately 1 inch in diameter. The metal strips are 1/2 inch wide and 1/16 inch thick.
Type 3. Round.
Variety a. No attachment holes (fig. 76d). There is a wide range of diameters and widths for these 29 ferrules. The diameters range from 7/8 inch to 2-7/8 inches, width of the metal 3/8 to 1 inch and thickness of the metal 1/16 to 3/16 inch.
Variety b. One attachment hole. The three specimens range in diameter from 1-1/4 to 1-5/8 inches, width of strip 5/8 to 7/8 inch and thickness of strip 1/16 to 3/16 inch.
Variety c. Two attachment holes. Two specimens were found. One is 2-7/8 inches in diameter and the other is 1-1/2 inches. They are made from strips 3/4 inch wide and 1/16 inch thick. The holes are bored approximately 180 degrees from each other.
Type 4. Oval. One specimen is 2 by 1-1/8 inches in diameter. The metal strip is 3/4 inch wide and 1/16 inch thick.
Three brass or copper wire loops were found in pre-1781 contexts. We have no idea what they were used for. Dimensions: diameter, 8.5 to 16.0 mm.; thickness, 1 to 1.5 mm. Two were in Feature 60 (Levels II and III) in the west barracks and one in Feature 52 (Level IV) in the east barracks.
Gunners Sight (?)
A brass bar fragment 25 mm. wide was found in Level I on the east scarp (fig. 76e). It is pointed at the unbroken end and has a screw hole drilled through it 21 mm. below the point and the number "219" engraved on the back. Because of its provenience it cannot be definitely related to the fort and it may be a piece of furniture hardware.
Two similar brass sheets were found on Feature 51 in the southwest casemate and in the sally port (Level II) (fig. 76f). They have the following dimensions: length, 10.5 cm. and 10.4 cm; width, 8.6 cm. and 8.1 cm.; thickness, .5 mm. Both are pierced with holes around the edge and the one in the southwest casemate still has sheet brass rivets in it.
Unidentifiable Lead Objects
There are a large number of cut scraps of lead which fell into this category, and which are probably the by-products of manufacture. These cannot be described because of their infinite variety. Two are illustrated in figure 76g. The few objects which appear to have been made for a purpose are described below.
Two lead truncated cones with rolled rims were found in the north casemate (Level II) (fig. 76h). One has nail holes in it and they probably were attached to wooden rods. They are 31.5 to 35.5 mm. long (too crushed for other dimensions).
Three lead bars with holes at each end came from the east casemate (Level II), Feature 64 (Level IV) in the west barracks, and the sally port (Level II) (fig. 76i). One is also notched at one end. We suspect these were somehow connected with weaving. Dimensions: length, 64.5 to 66.0 mm., width, 4 to 7 mm., thickness, 4.5 to 5.0 mm.
A lead ring with 10 serrations in one edge came from Feature 72 (Level III) in the east barracks (fig. 76j). It looks like a miniature crown. It is 30 mm. in diameter and 19.5 mm. high.
A D-shaped lead piece with three nail holes was found in the east casemate (Level II) (fig. 76k). Except that it is made of lead and not worn, it looks like a heel plate from a shoe. It is 28 mm. long, 45 mm. wide and 2.5 mm. thick.
A spoked, wheel-shaped object came from Feature 48 in the southwest casemate. It is fragmentary and may have been a pendant or toy.
Prehistoric Indian Artifacts
Scattered over the site, generally in Level I, were prehistoric stone tools and chert flakes. These were identified for us by Dr. Robert E. Funk, New York State Archeologist (personal communication), and range in time from ca. 4000 B.C. to A.D. 1300 with a couple of pieces that might be earlier. The types of projectile points defined by Funk are: one Otter Creek (Laurentian, 4000-3000 B.C.) (fig. 77a), two Brewerton side-notched (Laurentian, 3000-2000 B.C.) (fig. 77b, c), six Brewerton-like (Laurentian, 3000-2000 B.C.), one Lamoka-like (Lamoka ca. 2500 B.C.) (fig. 77d), four broad-stemmed (late Archaic), three crude small side-notched (late Archaic ?, ca. 2000 B.C.), two Fox Creek (A.D. 400-500) (fig. 77e) and one Levanna (Owasco, A.D. 1100-1300) (fig. 77f). The distribution of these and other tools is shown on table 36.
There are a number of utilized flakes and some drills (fig. 77g) and scrapers which cannot be identified by age. Only the Lamoka point, a bifacial tool fragment, three utilized chert flakes and eight unworked flakes can be associated with an aboriginal feature. This is a hearth (Feature 21) with fire-cracked rocks in the southeast casemate area (Level II). More Indian features were probably destroyed by the construction of the fort and subsequent occupation by the town of Rome. Three scrapers of green Normanskill chert from the Hudson Valley were found (fig. 77h). One came from the west end of the north casemate (Level II), and two from the sally port (Level II). The remainder of the artifacts appear to be local chert.
The Indian artifacts were found in all parts of the fort and many were in the top level suggesting probable disturbance or secondary deposits. Figure 78 shows the two main concentrations in relation to the fort and the original hill contour on which the fort stood. See also: Pendants, Type 4.
Last Updated: 02-Dec-2008