Академический Документы
Профессиональный Документы
Культура Документы
Excavations in the Southeast Line, Knoxs Artillery Cantonment Pluckemin History & Archaeology Series, Report No. 5
Middlebrook Cantonment was a far cry from the miserable log cabins and hovels in which the Continental Army shivered during the previous winter at Valley Forge. What occurred within those buildings and in the surrounding area was as remarkable as their construction. Officers were taught the science of gunnery and warfare, men were drilled in gun evolutions, craftsmen made and repaired everything from muskets to canteens, supplies were brought into warehouses from around the country, and the larger Continental Army quartered to the south at Middlebrook was resupplied for the campaign of 1779 (Seidel 1987). Pluckemin is a story of American success, and it is a story very much worth telling.
Figure 1. The 1778-1779 Pluckemin Cantonment, as depicted by Captain John Lillie area of investigation highlighted by the red circle
This report is the fifth in a series that tells the Pluckemin story. Its focus is on the supply side of the American story. One of the things that set the Pluckemin Cantonment apart from other winter camps is the fact the the field arm of the Military Stores Department was located there. In addition, two companies of artificers and a company of armourers were at work in the shops at Pluckemin (Sekel 1972; Seidel 1987). Artificer units were made up of skilled craftsmen, such as carpenters, blacksmiths, coopers, wheelwrights, and masons, while the company of Continental Armourers was comprised of gunsmiths and specialists in
munitions. Historical and archaeological research revealed that the industrial and supply portions of the site were located in the southeast and south, in the two long buildings depicted on the right in Figure 1. The long building at the base of the hill, facing the viewer (top right), has been referred to by archaeologists as the Southeast Line. Careful plotting of surface artifacts in the Southeast Line, and subsequent analysis, gave investigators some strong clues as to what took place in the building (Seidel 1987, 2012b). At the far right end of the building, slag, scrap metal and a variety of broken tools suggested blacksmithing. To the left of this smithy area, roughly centered in the red circle in Figure 1, lay a concentration of gun parts, indicating the probable loctation of a gunsmiths shop, part of the armourers workshops. Just to the left of the gunsmiths shop was an area defined by large quantities of sheet metal, perhaps the remains of a tinsmiths shop. In contrast, the left side held fewer industrial materials, but an abundance of the bone, shell, ceramics, glass and personal items that typically are found in living quarters. This suggested the location of barracks rooms for the artificiers and armourers, an hypothesis that was supported by subsequent historical and archaeological investigation as described in the fourth rpeort of this series (Seidel 2012d). This report focuses on excavations in one room of the Southeast Line at Pluckemin, the gunsmiths shop. It offers an exciting complement to the investigations done in the artificers living quarters, offering us an opportunity to see what these men worked on during the winter of 1778-1779.
Excavating at Pluckemin
Although the various reports in this series build upon one another, they are designed so that they may be read alone, without reference to earlier reports. This means that a certain amount of repetition is necessary in laying out the background and the methods used in the work. Those who have read previous reports and are familiar with the techniques used by the archaeologists may want to skip ahead to description of these specific investigations. Archaeological excavation is not the same as digging in the garden, nor is it at all related to the methods used by pothunters or those with metal detectors. It is instead a careful and precise approach to dissecting the layers of soil that have accumulated over time, looking for how those layers have been laid down by people and isolating the clues with which to reconstruct past ways of life. At Pluckemin, as with most sites, excavations areas were laid out inside a surveyed grid, so that every point, be it the corner of a fireplace or the location of an artifact, could be measured and recorded via precise grid coordinates. Each of these excavation areas was numbered, and then each layer of soil or identifiable feature such as a posthole, wall, or pit was given its own unique number. In the Pluckemin Project, each layer or feature was called a locus (a neutral term, simply meaning that it was something the archaeologists wanted to number and record). These layers were laid down by people and nature over time, and on a site that was occupied for a long time, the earliest deposits can be quite deep. Typically, excavators start at the surface and carefully dig or scrape away each layer of soil, gradually working their way down and back in time. As each new layer is exposed, it is photographed and drawn to scale, and the artifacts recovered are separated according to the layers or loci in which they are found.
Although these digging techniques are common to all archaeological sites, there are some aspects of the Pluckmin site and its topography that called for modifications. Perhaps the most striking difference between the Pluckemin Cantonment and many other sites is its short occupation. The bulk of the site was used for a period of only six months before it was abandoned. This meant that there had been relatively little time for its occupants to deposit multiple layers of soil and trash, or for buildings to be modified, repaired or replaced. In other words, the archaeological deposits were generally very thin. In addition, they were typically very close to the surface. Much of the site is located on a steep slope. Although the usual processes of natural soil build-up were at work there over the past 200 years falling leaves and each seasons dead vegetation being digested and processed by soil fauna to make new soil gravity and erosion constantly worked to wash away the new deposits, keeping the 18th century levels very close to the present day surface. As illustrated in the second report in this series, vast amounts of artifacts were to be found on the surface. Indeed, in some places all it took was a gentle raking away of leaves to expose a surface that might have been walked on by soldiers in 1778. On the typical site, with many, deep layers or strata, archaeologists spend a lot of time puzzling out these layers and trying to date them via the artifacts each contains. At Pluckemin, however, the short occupation and thin deposits made this less of a concern. More important than the vertical patterning of remains was the horizontal patterning of features and artifacts. The plotting and collection of surface artifacts (Seidel 2102b) was remarkably helpful in determining how the cantonment was laid out and what kinds of activities took place in various areas. It therefore seemed likely that a careful plotting and analysis of horizontal distributions of artifacts within rooms would be equally helpful, and the larger the area exposed, the better. This contrasts with a common excavation strategy of digging smaller squares of one meter or five feet. Ideally, archaeologists at Pluckemin wanted to see entire rooms exposed, all at once. Then it would be possible to plot the distribution of nails and other objects to figure out construction techniques and to reconstruct daily life and the utilization of space. That was the approach adopted in a portion of the artificers barracks, and it was very illuminating. Despite the shallow nature of the site, the excavations proved to be complex. Excavators had to trace many horizontal interfaces between loci. Often these interfaces shifted quickly as they were excavated, because they were due to the lensing of very shallow and thin levels. The lensing and many fine wash levels of erosion deposits make it difficult to follow soil changes on a site such as this. This is compounded by disturbances from tree roots, worms, and rodents, as well as the ever-present rocks. Nevertheless, careful attention to stratigraphy (the layering of the site), to horizontal patterns of debris and artifacts, detailed scaled drawings and photographs, and a focus on plotting individual artifact locations (as opposed to the common practice of lumping everything from a layer together), made it possible to tease out the story of the site.
Figure 2. Pluckemin site plan showing excavations, 1979-1989, with projected barracks (north at top of page) 5
Figure 3. Detail from the Pluckemin site plan showing the Southeast Line area 11G37 is to the left of the label Armorers (north at top; yellow grid square is 100 meters) 6
Figure 4. Plate from Diderots Encyclopedia (1762-1777, Vol. I), showing flintlock mechanisms 8
Figure 5. Plate from Diderots Encyclopedia (1762-1777, Vol. I), showing flintlock mechanisms 9
The men responsible for repairing some of this equipment at Pluckemin were commanded by Captain Cornelius Austin, who had been an armourer for the state of New Jersey earlier in the war (Brown 1980 - Revolutionary War Rolls, 1775-83: Monthly Muster Rolls, Continental Artillery). According to the muster rolls, Austin and his two superintendents (John Darbin and Godfrey Lowry) oversaw some twentytwo men. In addition to damaged weapons, documents show that a wide variety of tools and replacement parts were issued to the armourers. The partial listing in Table 1 gives an indication of the types of items which were delivered to the armoury between January 26 and June 15 of 1779 (from Sekel, 1972: Appendix J).
Table 1. Partial listing of materials supplied to the Pluckemin armoury Item Anvils Bayonets, damaged Binding wire, lbs. Borax, ounces Brass pipes Carbines, damaged Cocks Cock pins Files, assorted " flat " handsaw " 1/4 round " 1/2 round " round rattail Fuzee cocks Fuzee cock pins Fuzee hammer springs Hammers, large Hammer springs Locks Muskets, damaged Pinchers, pair
Pistol cocks
Number 1 296 9.5 40 480 44 144 144 70 2 46 1 6 29 24 24 24 144 144 1,488 2 24 24 3 3 31 1 144 2 2
Pistol hammers Rasps, common " 1/4 round Rifles, damaged Sear springs Small grindstones Tenent saws
10
This partial listing of materials supplied to the armoury gives some idea of the extent of their activities. These materials would have been supplemented by items already in possession of the workmen when they arrived, as well as shipments for which we have no documentation. The armourers continued working at Pluckemin after the artillery left for the spring campaign. Samuel Hodgdon, Field Commissary of Military Stores, left instructions for Austin on June 15, ordering him to supervise the men in the armoury and ensure that they completed the repairs on damaged muskets (Seidel 1987, 2012a). The shop must have been closed down sometime in early to mid-July, as a letter of July 6 (Hodgdon Letterbooks: Hodgdon to Sheineman) ordered teams to Pluckemin to pick up all the repaired muskets, "seven or eight hundred New Constructed Cart[ridge] Boxes, the machine for drawing fuzes, all the spare Bayonets, and scabbards, all the Drums & Fifes," and anything else the wagons would carry. Austin himself was ordered to report to the main camp, as damaged muskets were coming in quickly. Although the documents give us a good idea of the volume of repairs and the kinds of tools in use, there are still a great many unanswered questions. What, for example, was likely to go wrong with a soldier's weapon while he was in battle? Could some simple repairs be made in the field, or were guns sent back to the armourers for even minor repairs? What kinds of guns were most commonly used, and how well equipped were the troops with accoutrements and other weapons such as bayonets? The documents are often silent on these matters, and archaeology is sometimes the sole means of answering such questions.
Excavation of 11G37
Because of an increase in the frequency of gun parts in the surface collection of ten meter square SII6 (Seidel 1987, 2012b), this area was targeted for excavation. Surface mapping had already indicated the approximate location of a structure, or at least a room within a structure, in SII6. The evidence consisted primarily of a level terrace on the hill. This terrace could be traced to the north, where excavations in 11G11, 15, & 19 had already yielded structural remains of the Southeast Line (Seidel 1987, 2012d). Closer examination of the surface in SII6 showed that there were areas strongly resembling the chimney fall which overlay the fireplace in 11G19. This area was decided upon as holding the best prospect for subsurface testing. Excavation unit 11G37 was situated so as to contain within it the area which resembled the chimney fall. The area to be examined was fairly large, measuring 6.50 meters (21.3 ft) east-west by 4.00 meters (13.1 ft) northsouth. The unit and the surface features are rendered in Figure 6. These consisted of a level area extending across the center and lower (west) portion of the square, a drop in elevation along the west section, and a hump of rock and soil which protruded into the center of the level area. It was this protruding feature which had occasioned the placement of the unit in this area. We hoped that it might signify an underlying hearth or, if our hypothesis as to this area being a gunsmith or armourers shop was correct, a forge.
11
Figure 6. Excavation unit 11G37 prior to excavation - hachures represent sloping areas (north at left)
12
Without wishing to try the reader's patience with a detailed description of how the various strata were peeled back and sorted out, this particular area is best understood by examining the excavations as they proceeded through four stages of excavation. The strata in this area were somewhat complex, and an understanding of their distinctions and their stratification is crucial to the interpretation of the remains. To keep the discussion more comprehensible, the discussion of artifacts will, for the time being, be kept to a minimum. These may be examined in more detail in a later section
Stage I
The first stage in excavation was to strip off the shallow layer of topsoil across the unit. On the slope at Pluckemin, this initial stage generally had some predictable effects. Remembering that downhill portions of structures are often eroded, while uphill segments are buried under debris, the outcome is not hard to foresee. In this case, as with others, cultural levels were exposed from the outset on the downhill side (west), while slump and debris were prevalent on the upslope side. The various loci exposed after topsoil removal are shown in Figure 7. Loci 002 and 003 were natural slump and topsoil levels. Locus 004 was subsoil, but it appeared to have been water washed and contained slag, charcoal, and scrap metal. It lensed out to the west and washed over locus 005, which also resembled subsoil, but was far more compact than natural subsoil levels elsewhere on the site. As will be demonstrated shortly, this locus is actually subsoil which was placed on the hillside during construction to form a level platform.
Stage II
Locus 005, the platform material, had been exposed at such a shallow level that it was suspected that it had been subjected to a good deal of erosion. Consequently, its upper surface might be at a lower elevation than the original floor surface. Excavation therefore proceeded first on the uphill side of the unit, in loci 002 and 003. At a very shallow depth in the east end of the unit a new level was found, extending under both loci 002 and 003. With the exception of a few bits of slag on the upper surface, this level, locus 007, was devoid of artifacts and was a natural soil horizon, perhaps consistent with the original slope surface two hundred years ago. As locus 002 was peeled away from this level to the west, locus 007 abruptly terminated in a straight north-south line, and locus 002 dropped down in depth a further ten centimeters (4 inches) or so. This drop-off essentially defining a cut into the slope - can be clearly seen on Figure 8, along to the west edge of locus 007.
13
14
Figure 8. 11G37 at the second stage of excavation slump and eroded material have been removed, revealing the rear of a room (Stage II)
15
As locus 002, the slump, was removed farther to the west over the terrace, underneath it two new loci became apparent. The first of these was locus 009, which was a clay soil mixed with rock and mounded up towards the interface with locus 007. This was a rather loose matrix, with rocks oriented at random angles, suggestive of fill. In a semicircle around this mound was a dark soil (locus 008) containing large amounts of slag and charcoal, as well as various iron artifacts. This new level was quite compact, as if it had been intentionally beaten firm. Although locus 008 obviously was the same material across its entire extent, different segments could be graded in terms of color and compaction. The darkest and most compact area is shown south of locus 009, and the compaction and color became noticeably lighter as the locus progressed around to the north side of the mound.
Stage III
By the next stage of excavation, loci 007 and 009 had been removed, and a complete occupation surface had been exposed (Figure 9). Locus 007 was excavated down to subsoil (012), which was encountered at the shallow depth of 0.15 meters (6 inches) below surface.
As locus 009 was removed, it became more compact and noticeably redder. The color and compaction was very similar to the reddened soil found underneath hearths elsewhere on the site, suggesting great heat in this area. As locus 009 was excavated, its edges retreated inward, exposing more of locus 008. Thus it was apparent that locus 009 had washed out over this underlying level. The last traces of locus 009 were removed to reveal a smaller underlying feature, locus 011. This was a reddened area measuring approximately 0.90 meters (3 feet north-south) by 1.20 meters (4 feet east-west).
Loci 009 and 011 were clearly subjected to a great deal of heat, which is consistent with the existence of a fireplace or forge. No forge or hearth base was encountered, however, so excavation of locus 008 proceeded. This quickly revealed an intensely dark and compact area south of locus 011. It was distinct enough to be treated separately, and was designated locus 010. In the southeast corner of the locus, a compact and dense cluster of rock was exposed. In addition, a concentration of artifacts was disclosed to the east, along the base of 012.
16
Figure 9. 11G37 at Stage III of excavation a void once occupied by the forge is designated 011, surrounded by a work surface designated 008 and 010
17
Figure 10. 11G37 at Stage IV of excavation; a compacted work surface south of the forge (011) is designated 016, cut into by a pit
18
Stage IV
In the last stage of excavation, both loci 008 and 010 were completely removed. Underneath locus 010, a more general but less intense version of locus 010 was found and designated locus 016. This second, underlying surface is shown in Figure 10.
A pit was discovered at the same time, cut down from the level of locus 016. As the pit was emptied of its contents, it was shown to be broad and shallow, and had been filled with broken tools, scrap metal, and soil littered with fine iron filings. The pit cut down into sterile subsoil (make-up for the original floor), and locus 016 was shown to rest on this level.
Figure 11. East view of 11G37 with excavation completed Figure 11 provides a view of the unit looking east, after this last stage of excavation. The ledge running from left to right in the photograph (not the edge of the excavation unit to the rear, but the area with the vertical scale in front of it) is the cut made into the subsoil by the builders for the rear wall. It is interrupted to the right by a probing excavation dug into the subsoil originally the ledge continued to the right. Most traces of the forge have been removed in this photo; it was located to the right of and behind the sign board. 19
Ultimately, the platform material itself was removed, revealing the original slope of the hill on the west side (locus 017). No fossil topsoil horizon was visible, probably due to extensive root and worm action.
Interpretation
The interpretation of this area is perhaps best done by examining the features in reverse order from above. It is possible to suggest a sequence of events in the construction, use, and abandonment of a structure which stood here. The first event for which we have evidence is a leveling operation on the hillside. It is clear that the builders cut into the hill on the uphill side of the room, the cut being visible both on plan as the interface between loci 012/017 and locus 016 (Figure 10). Soil from this cut, and perhaps from elsewhere, was deposited downhill as locus 005 (Figures 7-9). This displaced subsoil provided a level platform and constituted an earthen floor. Excavation showed that the builders packed larger amounts of stone into the front edge of the platform. This too is shown in section, although the degree of erosion makes it hazardous to suggest a specific location for the front wall. The distance from the uphill cut to this point suggests a room depth of approximately 4.80 meters, or 15.75 feet. A depth of between 16 and 18 feet is most likely. The kind of structure which was built must remain conjectural. In fact, there is little structural evidence to suggest that a building of any sort was erected around this area, as no evidence of sills or postholes were found. However, even in the absence of John Lillies drawing showing a structure here (Figure1), the winter occupation argues forcefully that activities which took place here were sheltered. Also, the 126 nails and 51 pieces of pane glass recovered are suggestive. Mirroring the findings in the artificers quarters to the north (Seidel 1987, 2012d), pane glass was found predominantly along the west side, or front, of the structure. Adequate light would have been necessary for smithy operations, hence the relative increase in pane glass as compared to the North Line. Subsequent to leveling the floor, a forge was installed. This is made apparent by the combination of thermal discoloration and compaction, as well as the overwhelming predominance of artifacts which can only be associated with forge work. The forge was located along the rear edge of the room and measured roughly 0.90 meters by 1.20 meters (3 feet by 3.5-4.0 feet). As it was dismantled or destroyed at some point, its original position may be traced only at its base, locus 011. Artifacts indicative of forge work include slag and charcoal, which were found in abundance (27.776 kilograms (61.23 lbs.) of slag). Discarded tongs, files, chisels, and hammer heads were also found, along with innumerable scraps of waste metal. Aside from architectural fragments, the largest single category of identifiable materials was gun parts. Gun repair must have been a dominant activity in this room. The stains and compacted areas (loci 008, 010, 016) around the forge base (loci 009, 011) are the result of smiths standing near the fire to work and scattering charcoal and debris around the forge. The bulk of the activity appears to have taken place on the south side of the fire, as this was consistently more compact and was associated with greater numbers of artifacts.
20
From what we know of how smiths operate, an anvil would have been located near the forge fire. The most convenient spot would be close enough so that the smith could heat his iron, remove it from the fire, and turn around to work it on the anvil. The compaction of the soil and heavy usage to the south of the fire suggests than an anvil was placed there. Usual practice seems to have been to bury a post of large diameter (one to two feet) in a hole, leaving two to three feet of the post exposed above ground as an anvil base. The pit excavated to the south of the fire was too shallow to have served as a hole for such a support, and no other post holes existed within the confines of the room. If the anvil base was installed above ground, perhaps the densely packed rock seen in both 010 and 016 (at the edge of the pit) served in some way as a base for the anvil supports. Sometime after the forge was built and put into operation, a pit was dug to the south, abutting the possible position of the anvil. This pit was broad, measuring 1.30 meters by 0.90 meters (4.3 by 3.0 ft), and shallow, extending 0.30 meters (one foot) in depth. Its purpose must remain conjectural. Although a number of explanations might be possible, it seems likely that it served as a rudimentary catchment for hot bits of scrap and scale from the anvil. If so, it may have been cleaned out on a periodic basis as it filled up with debris. The deposition of locus 010, a working floor surface, over the top of the backfilled pit makes it clear that the pit fell into disuse while the forge was still in operation. The pit fill contained gun parts, primarily pieces from musket locks, as well as fragments of tools. Particularly noticeable were pieces of files. The files, combined with the fact that the fill was impregnated with iron filings, suggest that the fill was brought in from another area. The heating and forging of metal is an operation which is different from the finish filing of repaired or manufactured pieces, which must have been done away from the forge itself. Either a separate part of the room or a different room altogether must have been used for filing and finishing. In any case, it is clear that the dominant activity around the forge was the repair of flintlocks. The smith apparently continued to work primarily on the south side of the fire after the pit was filled. The accumulation of material against the rear "ledge" (where the subsoil was cut to form the back of the building platform) shows that the broken pieces were thrown with some regularity towards the rear of the room, perhaps bouncing off the wall. The compacted soil encircling the forge to the west and north makes it clear that these areas were also utilized, although to a lesser extent. During or after abandonment of the site, the forge was dismantled. This may have been necessary to retrieve the bellows tuyere, or it could have been done by local salvors after the site's abandonment. The hump of clay and stone which overlay the "foot-print" of the forge is suggestive of the construction technique used here. It is likely that the forge took the form of a large box, enclosed with brick or stone sides, the interior was probably filled with rock rubble and clay up to the top of the sides. The top, at a convenient working height (close to waist high), would have been topped by a bed of sand for the charcoal fire. When the forge was dismantled and removed, the removal of the retaining walls resulted in the slump of the interior fill forward onto the floor of the shop, and erosion began its slow work of returning the slope to its original contours.
21
Table 2. 11G37 artifact frequencies Artifact Group Kitchen Architecture Furniture Arms Clothing Personal Tobacco Activities Total Carolina Pattern % Range 51.8 69.2 19.7 31.4 0.1 0.6 0.1 1.2 0.6 5.4 0.1 0.5 1.8 13.9 0.9 2.7 11G 37 Count 17 178 - 5.5 45 5 1 565 811 % 2.1 21.9 5.5 0.6 0.1 69.7 99.9
The actual frequencies for 11G37 are presented in Table 2. The total artifact count of 811 was reached by eliminating from the sample all modern artifacts, as well as slag and charcoal. Only four fragments of bone were found and these were also excluded. Several points are worth noting about this pattern. First, the Kitchen group contains no ceramics. It consists of 16 bottle fragments and one knife blade. The bottle fragments could represent drinking on the job, but it is equally likely that bottles were used to store linseed oil, "sweet oil" and other liquids known to have been delivered to the armourers (itemized in Sekel 1972: Appendix J). The complete absence of ceramics strengthens the conclusions reached from surface collection data, which was that this area was not a living quarter, but instead served a specialized function, most likely related to weapons repair (Seidel 1987, 2012b). Note also that pane glass represents over 40% of the Architecture group. The simplest explanation for its presence is that it was used in the structure. Thus the size of this group reflects usage in construction rather than being inflated by nail manufacture or some other specialized activity. The Arms group is clearly significant in the 11G37 pattern, being almost five times the highest frequency seen in the Carolina Pattern. Skeptics might argue that 45 gunparts do not a gunsmithy make, but two points should be made. First, broken gun parts might have been recycled in an effort to maximize output, so that we should expect a much higher proportion of unrecognizable scrap than recognizable parts.
22
Second, it is highly likely that a forge room would have been cleaned out on a regular basis to remove both slag and waste metal. If this practice was not followed over a six month period, the accumulation of debris should have been much greater than observed. It is entirely possible that the material excavated from within the confines of the room only represents refuse from the last days of occupation. This also suggests that refuse areas exist away from the buildings that might greatly increase our data base if found and excavated.
Table 3. 11G37 pattern compared with SII6 surface collection Artifact Group Kitchen Architecture Furniture Arms Clothing Personal Tobacco Activities Total Carolina Pattern % Range 51.8 69.2 19.7 31.4 0.1 0.6 0.1 1.2 0.6 5.4 0.1 0.5 1.8 13.9 0.9 2.7 SII6 % 1.2 0.8 0.2 97.9 100.1 11G37 % 2.1 21.9 5.5 0.6 0.1 69.7 99.9
The pattern found in 11G37 is compared with that from the surface in Table 3. Aside from very small changes in the Clothing and Personal groups, and the presence of bottle glass in the Kitchen group, the biggest differences lie in the Architecture, Arms, and Activities Groups. The reason for the under-representation of nails and pane glass in the surface collection is hard to explain, but surface collection similarly underestimated the frequency of excavated architectural materials in the artificers quarters (Seidel 2012d). Surface collection again did not accurately predict the subsurface frequencies, but was nevertheless useful in distinguishing a difference in function form living quarters and even predicting the nature of that activity. The other differences are somewhat easier to explain. The decrease in the excavated Activities Group occurred at the expense of both architectural items and arms. Part of this is understandable in terms of disposal and erosion patterns. Scrap was more prevalent in the erosion-prone areas to the front of the building and behind it, so that more was exposed on the surface. Gun parts occurred in larger number towards the rear of the room itself, and were thus more likely to have been covered by erosion than exposed. The patterns seen in surface collection can therefore be expected to exhibit some variance from the excavated patterns, particularly where groups are disposed of differently.
23
Although by no means perfect, there does appear to be a correlation between artifact patterns found on the surface and those revealed by excavation. Particularly noticeable is the lack of ceramics in both collections and the absence or low frequency of furniture, personal, and tobacco items. More significant than the exact degree of correlation is the indication that surface patterns can have great predictive value. As demonstrated in the analysis of SII6 surface collection data by Seidel (1987, 2012b), this is particularly true if one adjusts the frequencies to remove large numbers of uninformative artifacts, such as slag and scrap metal, which may make more useful patterns. Also important is the observation that it is not so much a large percentage increase in a certain category which is informative, but the deviation from the prevalent pattern. Thus even small numbers of artifacts not seen elsewhere, such as the five gun parts collected from the surface in SII6, can be significant.
Types of Weapons in Use During the War Early in the war, British muskets were most common in the American army. Many had been taken from royal armouries in the colonies or had been obtained during early wars with the French and the Indians. Others were manufactured by American gunsmiths and copied from British guns. These were often made on contract for local Committees of Safety, organizations set up in part to coordinate colonial efforts early in the war, or by manufactories set up by states such as Virginia.
24
One of the biggest problems in the early years must have been a lack of standardization, particularly in bore sizes. So many different types of weapons were in use within a given regiment that it must have been impossible to supply them with adequate amounts of properly sized shot.
25
could be wet, in which case nothing at all happened. Worst of all, one of the springs or another part of the mechanism could break. Broken examples of all three spring types (main, sear, and frizzen) have been found. With hard usage, a variety of other parts might also break or wear out. Experience with old weapons and modern reproductions has shown that with use, the cock can become too loose to function effectively. The method of attaching a cock to the lock was relatively simple. A square projection from the lock (part of the tumbler "g" in Figure 5) was fit into a square hole in cock ("h" in Figure 4) and a small bolt held the assembly in place. With use, the cock is continually pulled back and then propelled forward. The problem arises when either the square hole or the tumbler shaft wears and loosens up. An example of this is shown in Figure 12.
Figure 12. Cock modified to narrow the tumbler hole exterior surface on the left; reverse on the right
In this instance, an expedient used by some modern gunsmiths (William Kennedy, Mathias Koppinger, personal communications) was apparently attempted to remedy the situation without the use of new parts. The solution was to remove the cock, flip it over onto and anvil or other hard surface, place a chisel next to the hole on each side, and strike it a number of times. The blows forced the metal out, effectively narrowing the hole, so that it would fit more the tumbler more tightly and could be reused. A variety of other parts could break as well. Frizzens and sears occasionally broke, necessitating replacement. The threads on the top jaw screw or cock pin sometimes stripped, so that a variety of these have been found. In addition, ramrods bent or broke, screws and bolts sheared off, and cocks broke in two.
26
Not all of the pieces recovered were broken. A number of pieces were apparently new when lost or discarded. They show no evidence of ever having been used or put on guns. These pieces, such as the 144 cocks, cock pins, hammers, and springs shipped to the Pluckemin armoury, must have been turned out as blanks in armouries such as Easton and Springfield for shipment to armourers in the field. Inventories of such parts for British or French flintlocks could be kept on hand by field armourers and quickly fitted to individual guns when needed. The cock illustrated in Figure 13 is probably an American-made piece designed to fit on a British "Brown Bess." It may have been formed in a die, rather than forged "free-hand," a technique not adequately documented for pieces of American manufacture (Brumfield 1985). A small hole was started on the inside face of the piece, but was never completely drilled through. Armourers in the field would have completed the hole and squared it off to fit the tumbler. This could not be done before-hand because the position of the tumbler to the frizzen might be slightly different from lock to lock. A proper fit was crucial to ensuring a good spark and ignition. Unmodified pieces, particularly in combination with repaired and unrepairable items and experimental archaeology such as that done by smiths at Colonial Williamsburg (Brumfield 1985), promise a much better understanding of smithing techniques. Some of these blanks must have been lost, while others were unsatisfactory for one reason or another. A clip, for example, designed to hold in place a band around the barrel of a French musket was recovered. As shown in a schematic drawing Figure 14, the band was designed to slide down the barrel to the left. A hole on the band (marked 1) was designed to be retained by a spring clip fit into the stock (2). A clip found in 11G37 (Figure 15) was perfect in all respects but one. Perhaps because he was tired after a day 27
of turning out large numbers of these clips, the gunsmith simply made a mistake - he put a pin on the wrong side of the clip so that it became useless (Figures 15 & 16). Disgusted, no doubt, he simply threw it out. Barrel bands from French weapons have also been recovered (Figure 16).
Figure 15. Retaining clip for barrel band (French style) recovered from 11G37 the tang is on the wrong side
28
In addition to blanks, a number of modified pieces have been recovered. The reasons behind these modifications are sometimes obscure and difficult to explain. One peculiar artifact which provides an example is a small iron ring or portion of a tube (Figure 17). Several of these were found during surface collection, and more turned up in excavation at Pluckemin (as well as in the collections at Morristown National Historical Park). They were identified rather quickly as coming from the end of the sockets on British bayonets. Although the identification was relatively straight-forward, the reason for finding them was somewhat more difficult to determine. This was not a part of a bayonet which would seem prone to breakage, and is, in fact, one of the stronger portions of the piece. More easily broken would be the blade and shank.
The answers to such puzzles can occasionally come from unexpected quarters. Like some other archaeologists, the author took to shooting replicas of Revolutionary war guns in order to become familiar with the weapons, and how they operated, broke, and were repaired. A plausible answer to the bayonet puzzle was supplied when trying to find a bayonet to fit a French Charleville. Unable to find the proper bayonet, the only solution appeared to lie in modifying a more readily available British bayonet to fit the gun. To make it fit, it was necessary to shorten the socket by cutting off the ring at the base (see Figure 18). The piece left over from the modification was identical to the excavated fragments. A logical explanation for this peculiar find is that armourers at Pluckemin were modifying British bayonets to fit the newly arrived French muskets.
29
Figure 17. Collar or base of bayonet socket (British type (scale in centimeters and millimeters)
30
The excavations described here provide only a glimpse at the activities of the armourers at Plucekimn. The contents of one room, at most, were recovered during the 11G37 operations. It is difficult to imagine the twenty-five officers and men of Austin's company of armourers all operating within the confines of a single 18 by 18 foot room. There must be other rooms which were also used by the gunsmiths; perhaps individual rooms were used for specific tasks, such as forging, filing, stock inletting, and bayonet repair. In addition to these resources, it is likely that additional discards are deposited elsewhere on the site, perhaps in refuse pits. Although both the excavation and the analysis presented are in some ways preliminary, they serve to illustrate the potential of such study. More detailed examination, especially in consultation with weapons experts and modern craftsmen who practice the older trade of gunsmithing at sites such as Colonial Williamsburg, would almost certainly provide additional insights into the breakage and repair of weapons, and no doubt bring to light additional field expedients used to keep the army armed and accoutred. In spite of the long standing interest of collectors and military historians in firearms, our understanding of the raw materials, tools, and technology used by 18th century armourers is in its early stages. Closely dated assemblages such as that from Pluckemin may prove highly valuable to the study of American gun manufacture in general.
31
References Cited
Barna, Frank Carl 1984 A Most Intricate Department: The Commissary General of Military Stores or Ordnance Department Under Benjamin Flower and Samuel Hodgdon, 1777-1782. Master's Thesis, University of Maryland.
Brown, M. L. 1980 Firearms in Colonial America: The Impact on History and Technology, 1492-1792. Smithsonian Institution Press, Washington D.C.
Fitzpatrick, John C. 1931-44 The Writings of George Washington. 39 volumes. Washington, D.C. Hodgdon Letterbooks 2011 James, Charles 1816 An Universal Military Dictionary. T. Egerton. London. this grand supply The Samuel Hodgdon Letterbooks, 1778-1784. Vol. 1, July 19, 1778 March 31, 1781. Joseph Lee Boyle, editor. Heritage Books.
Neumann, George C., and Frank J. Kravic 1975 Collectors Illustrated Encyclopedia of the American Revolution. Stackpole Books, Harrisburg, Pennsylvania.
McDougall Papers n.d. Alexander McDougall Papers. New York Historical Society, New York.
Parrington, Schenck & Thibaut 1984 The Material World of the Revolutionary War Soldier at Valley Forge. The Scope of Historical Archaeology. Edited by David G. Orr and Daniel G. Crozier. The Laboratory of Anthropology, Temple University, Philadelphia.
Peterson, Harold 1968 Seidel, John L. 1983 "Archaeological Research at the 1778-1779 Winter Cantonment of the Continental Artillery, Pluckemin, New Jersey," Northeast Historical Archaeology, Vol. 12: 7-14. 32 The Book of the Continental Soldier. Stackpole Books, Harrisburg, Pennsylvania.
1987
The Archaeology of the American Revolution: A Reappraisal & Case Study at the Continental Artillery Cantonment of 1778-1779, Pluckemin, New Jersey. Ph.D. dissertation. University of Pennsylvania, Philadelphia. "'China Glaze' Wares on Sites from the American Revolution: Pearlware Before Wedgwood?" in Historical Archaeology Vol. 24, No. 1:82-95. "The Winter of 1778-1779 at Pluckemin" in Flintlock and Powderhorn, Vol. 11, No. 1:4-11.
1990 1993
1995a "Military Industry in the New Nation" in Invisible America, ed. by Mark Leone & Neil Silberman. Henry Holt & Company, New York. 1995b " 'Class Warfare': The American Militia System" in Invisible America, ed. by Mark Leone & Neil Silberman. Henry Holt & Company, New York. 2012a The Continental Artillery at Pluckemin & Middlebrook, 1778-1779. Report submitted to the Friends of the Jacobus Vanderveer House, Bedminster, New Jersey. Washington College Public Archaeology Laboratory, Chestertown, Maryland. 2012b Patterns on the Ground: Surface Archaeologt at the 1778-1779 Pluckemin Cantonment. Report submitted to the Friends of the Jacobus Vanderveer House, Bedminster, New Jersey. Washington College Public Archaeology Laboratory, Chestertown, Maryland. 2012c Trash & Tea: Gleanings from a Revolutionary War Refuse Dump. Report submitted to the Friends of the Jacobus Vanderveer House, Bedminster, New Jersey. Washington College Public Archaeology Laboratory, Chestertown, Maryland. 2012d Barracks on an Elegant Plan: Excavations in the Artificers Quarters, Southeast Line, Knoxs Artillery Cantonment of 1778-1779. Report submitted to the Friends of the Jacobus Vanderveer House, Bedminster, New Jersey. Washington College Public Archaeology Laboratory, Chestertown, Maryland. Sekel, Clifford, Jr. 1972 The Continental Artillery in Winter Encampment at Pluckemin, New Jersey, December, 1778 June 1779. Master of Arts thesis. Wagner College, Staten Island, New York.
Showman, Richard K. (editor) 1979 South, Stanley 1977 Method & Theory in Historical Archaeology. Academic Press, New York. The Papers of Nathanael Greene. Volume III. Chapel Hill.
33