ehs logo4

The Maine Archaeological Society Bulletin Volume 27 Number 1 Spring 1987, pp. 31-50.
"Archaeological Excavations At Fort Edgecomb, Summer 1985"

Robert L. Bradley
Maine Historic Preservation Commision

Norman L. Buttrick
Freeport High School Faculty

In the spring of 1985 limited Historic Preservation funds through the Department of the Interior, National Park Service, were secured to enable the Maine Historic Preservation Commission and the Maine Bureau of Parks and Recreation to undertake archaeological excavations at Fort Edgecomb under the direction of Bradley, with Buttrick as assistant director.

History of the site

In 1808, as America watched the Napoleonic Wars consume the great nations of Europe, a complex system of fortifications was established to protect our nation's Atlantic coast (1). For Maine this meant new forts in Kittery, South Portland, Portland, Phippsburg, Edgecomb, Boothbay, St. George, Castine, Machias, and Eastport, designed to protect key ports and estuaries. This network of fortifications is now known as the "Second System," the "First System" having been built in the period 1794-1798.

MaineArchSoc1987 32 Top

Fig. 1. Oblique aerial view of Fort Edgecomb, from the southeast, with Wiscasset in the backgriound.

Fort Edgecomb, located on the eastern side of the Sheepscot River, was built to protect Wiscasset, one of Maine's most important ports of the early 19th century (figure 1). It is the best-preserved "Second System" fort in America, with its impressive octagonal blockhouse and intact earthworks. There are three reasons why the fort has survived. First, because Wiscasset declined in economic importance after the War of 1812, no large granite fort was needed there during the Civil War (1861-1865), when many of America's "Third System" defenses were constructed. Second, the site became pasture after 1820 and was never developed for other purposes. And third, in the 1870s and 1890s local citizens raised funds to repair the deteriorated blockhouse (2). See figure 2 for a view of the blockhouse prior to the restoration of the 1870s, graphic evidence of its poor state of repair, figure 3 for an 1875 fund-raising poster, and figure 4 showing post-l875 work completed.










Fig. 2. Fort Edgecomb Blockhouse (pre-1875).











Fig. 3. 1985 fund-raising poster.

However, not all of Fort Edgecomb survives above the ground. An 1820 survey map (figure 5) , produced by the U.S. Army Corps of Engineers in 1820, clearly shows the locations and rough dimensions of various additional buildings to service the fort--a storehouse, two enlisted men's barracks, and an officers' quarters. These lost buildings are otherwise unrecorded (3).

Ft. Edgecomb

Survey plan









Fig. 4. Fort Edgecomb Blockhouse (post-1875).



Fig. 5. 1820 Corps of Engineers Survey Map (re-drafted).

Purposes of the Project

There were several purposes for the Fort Edgecomb project. Little archaeology had ever been conducted on military sites of this period in Maine (4), and gaining an understanding of the structural nature of lost components would improve the Bureau of Parks and Recreation's on-site interpretation for the visitor. In addition, most Maine archaeological projects must , for security reasons , be conducted discreetly; however, it is important for public benefit and education to provide the public with regular opportunities to view scientific archaeology in progress. Protected, publicly-owned sites like Fort Edgecomb are ideal contexts for such demonstration projects.

Public Education: Results

In advance of the project, Sheila McDonald, Interpretive Specialist for the Maine Bureau of Parks and Recreation, delivered lectures in the spring of 1985 to the Edgecomb Historical Society and the 5th and 6th grades of the Edgecomb Elementary School, detailing the history of Fort Edgecomb and giving notice of the forthcoming fieldwork. In addition the Bureau issued a press release which was published in newspapers statewide and led to subsequent feature articles in the Lewiston Daily Sun, the Wiscasset Newspaper, the Boothbay Register, the Portland Press Herald, and the Brunswick Times-Record . Television stations WMTW (Channel 8) and WGME (Channel 13) provided broadcast news coverage on, respectively, the first and second days of the project. In addition to this extensive publicity, the project was scheduled for work from Tuesday through Saturday each week to enable working people as well as vacationers the opportunity to see the crew in action; and the four weeks of work straddled the principal summer months, running from July 16th to August 10th. The Bureau of Parks and Recreation also developed a special project brochure for free distribution to site visitors, not only to give them material to supplement the standard site brochure, but also (it was hoped) to answer the questions so routinely asked of archaeologists: "What are you looking for?. . . What are you finding? . . . How do you know where to dig? . . . How deep do you have to dig?" Unfortunately, visitors generally pocketed the special project brochure on their way into the site and asked the questions anyway. Printed materials, after all, can be read later at one's leisure. What is needed for such projects is a temporary interpretive panel at the entrance to the site which answers these basic questions, since panels are not stuffed unread into glove compartments.

The public response to all of these efforts was gratifying, as 1985 visitation to the Fort Edgecomb Historic Site roughly doubled the figure of the previous season (1984) and the year following the project (1986):

1984: 2,115
1985: 4,174
1986: 2,067


Finally on October 24, 1985, the author (Bradley) addressed a well-attended public meeting of the Edgecomb Historical Society, a meeting which also included in the audience the entire 5th and 6th grades of the local elementary school. This slide presentation gave residents local to the site the first summary findings of the project. Subsequently, the project has been a prominent part of Bradley's regularly-updated survey lecture on historical archaeology in Maine.

Archaeology: Methodology

Aerial view

Fig. 6. Vertical aerial view of Fort Edgecomb.

The project crew was small but highly experienced, consisting of the director, assistant director, a surveyor, two full-time excavators, three part-time excavators, and a volunteer photographer. With regard to photography, oblique and vertical aerial phonographs, both black and white and aerochrome infrared, were taken prior to and during excavation (figure 6). All test units were photographed in black and white and color during excavation and upon completion. Horizontal and vertical control related to a datum point on the foundation of the blockhouse. Measure was in English, rather than metric, given the procedures of the original builders, the basic test unit being a 5-foot square, often excavated by half or quarter. Excavation was by stratum, the stratigraphy being simple on this single-component site and consisting of post-fort fill (20th-century), demolition (post-1820), fort (1808-1820), and pre-fort sterile or bedrock. The surveying was facilitated by an existing complete plan of the park (scale 1:360) with 2-foot contour intervals, drawn by Wadsworth and Boston, Architects, in 1961. Excavation proceeded by trowelling, with all back-dirt passed through quarter-inch screen. A plan was drawn for each stratum in each test unit, as well as at least one profile (these are available for study in the files of the Maine Historic Preservation Commission).

Area plan

Fig. 7. General Plan of Excavated Areas.

Archaeology: Goals

The principal focus of the project was the investigation of two of the fort's lost buildings, the officers' quarters and the western enlisted men's barracks (figure 7). Several hypotheses were to be tested: 1) the Federal survey map of 1820 would be found to be accurate; 2) the two buildings would be represented by simple fieldstone footings of two or three courses encountered at a shallow depth below surface (less than a foot); and 3) few artifacts (other than architectural/structural) would be encountered, given the short documented occupation of a few months in 1814 and in 1815 by a small militia garrison, and an even more limited and casual use in 1864. It was, however, expected that the usual high volume of clay pipe fragments would be encountered, typical of virtually all Maine military sites, almost regardless of length and size of occupation. The authors guessed that at least one hundred of these would be recovered. There were to be several big surprises.

Archaeology: Structural Features

In terms of the three hypotheses to be tested, the U. S. Army Corps of Engineers survey map of 1820 (figure 5) proved to be very accurate in pinpointing the locations and general size in plan of the two buildings in question. However, the map gives no hint of construction materials or any other details. On the other hand, with regard to the second hypothesis (the nature of the footings of the two buildings), both the officers' quarters and enlisted men's barracks confounded our hypothesis, and in entirely different ways.

Testing excavation

Fig. 8. Testing the Officers' Quarters Adjacent to the Blockhouse.

Quarters plan

Fig. 9. Plan of Officers' Quarters and Test Units.

The officers' quarters foundation supported a small (20 x 17-foot) building immediately adjacent to the east side of the blockhouse (see figure 9). But instead of a couple or three courses of fieldstone to keep sills above grade, a full cellar, almost 7 feet deep, was encountered, supported by massive dressed granite slabs (figures 10 and 11). Mortar stains on the top of these stone footings marked the positions of bricks which had been toppled into the cellar during demolition sometime after 1820. Structurally, these quarters were therefore equivalent to the blockhouse, which features six courses of bricks above granite foundation walls (see figure 8). Therefore, the officers' quarters confounded our structural hypothesis by being far more substantial and permanent than we had predicted. Certainly, for a building of its size, these quarters were overbuilt.

Testing excavation

Fig. 10. Interior of West Corner, Officers' Quarters.

Quarters plan

Fig. 11. Exterior of East Corner, Officers' Quarters.

The enlisted men's barracks, however, were another matter (figure 12). The 1820 survey map shows a long, rectangular block to the east/southeast of the officers' quarters which clearly was designed to fit a terrace one level down from the blockhouse and officers' quarters. Excavations here (figures 13, 14) yielded initially baffling results, as virtually no fieldstone was encountered. Instead, rotten wood was uncovered, some of the pieces of which were articulated. The first thought was that we had not yet found the stone footings and were looking at parts of the collapsed building's frame. Ultimately, however, it became clear that this wood was in fact the intact sills of the barracks (about 64 x 18 feet), laid directly on grade with occasional cobble shims (figure 15). Therefore, the enlisted men's barracks confounded our structural hypothesis by being far less substantial and permanent than we had predicted.


Fig. 12. Plan of Barracks and Test Units.

Testing west

Fig. 13. Testing of Barracks Area, View from West.

Testing east

Fig. 14. Testing of Barracks Area, View from East.

Barracks sills

Fig. 15. Southwest Corner of Barracks Sills.

This stark dichotomy between the structural nature and quality of the bases of the officers' quarters and enlisted men's barracks has no archaeological parallel in any other Maine fort of any period thus far studied, and the authors are unaware of an equivalent dichotomy anywhere else. Whether this will prove standard at other "Second System" forts in Maine at such time as they may be sufficiently tested remains to be seen. For the moment all one can say is that there was an exaggerated class system in the minds of the builders of 1808!

Archaeology: Structural Artifacts

There was no way to predict the volume of artifacts relating to the structures which would be encountered, since there are no records of when after 1820 the two buildings in question ceased to exist. Obviously, buildings allowed to gradually collapse or which accidentally burn will tend to yield far more hardware than those which are deliberately salvaged or dismantled. As it happens, testing of the officers' quarters recovered no hardware, while the barracks yielded 7 such artifacts: 1 door latch (hasp), 1 strap hinge, 4 butt hinges, and 1 cupboard latch. Because of the limited percentage of each building tested, however, it would be rash to draw conclusions from this. Likewise, the pattern of hardware distribution in the barracks (figure 16) is suggestive of an entrance near the southwest corner (hasp, strap hinge) and possible window shutters in the same area, a well as at the eastern end (butt hinges), but the sample is far too limited for any degree of certainty.

Barracks, artifacts

Fig. 16. Distribution of Barracks Hardware.

Window glass fragments were found in association with both buildings (so at least the barracks were glazed!). The number of pieces from the officers' quarters is very small (59), mostly from the side facing the blockhouse west. Although the sample may well be biased, it is possible that the sash and lights from the quarters were at least in part removed for salvage. The barracks' total of 581 is far greater with a particularly large concentration around the southwest corner, suggestive of a window(s) in that vicinity (figure 17).

Glass fragments

Fig. 17. Distribution of Barracks Window Glass Fragments.

Nails, both hand-wrought, transitional, and early full-machined were found. This is consistent with sites of the Federal Period. For the purposes of this study they have been subdivided into two types of function: structural and lathing. In neither building did nails of either type form any patterns, but the totals are revealing. The officers' quarters yielded 68 structural nails and 279 lathing nails, while the barracks produced 729 structural and not one lathing nail. The officers' house had a finished interior, while the enlisted men, it seems, stared at open framing and the inside face of the barracks' sheathing.

Bricks and fragments thereof also formed no recognizable patterns in either building, except that the cellar of the officers' quarters – at least at the west corner – contained a deep demolition stratum of bricks, many of which had been toppled off the top of the stone footings. The barracks, in contrast, contained a thin scatter of brick fragments in nearly every test unit. At this point the locations of hearths and chimneys are unknown.

Archaeology: Faunal Remains

Although no detailed analysis of faunal remains has yet been undertaken, these can at least be divided into two groups: vertebrate and invertebrate. The vertebrate presumably represent mostly or all bones of domesticates, while the invertebrate are almost exclusively soft-shell clam. No obvious concentrations were identified in either building, but some preliminary observations can be made. Food was consumed in both buildings, meaning that each doubled as a mess, respectively, for officers and enlisted men. Total counts from the buildings are reflective of building size and number of occupants:

Officers' Quarters: 45 bone, 9 shell
Barracks: 258 bone, 75 shell


The only other observation here would be the test unit which straddles the northern sill of the barracks (20S/147.5E). Here all finds lay outside of the building, suggesting the proximity of a door or window through which garbage was thrown.

Archaeology: Non-Structural Artifacts

It should be recalled that the third hypothesis to be tested predicted that few non-structural artifacts would be encountered, given the fact that Fort Edgecomb is virtually a single-component site with a very short-term and small-scale occupation.

This was certainly born out by the volume of ceramics recovered, which can be summarized as follows:

 Officers' QuartersBarracks
Redware 4   272
Creamware 26   74
Pearlware 49   306
Porcelain 0   3
Yellow-Ware 0   7
Hardwhite 0   59
  ______   ______
Totals 79   721

While the total for the barracks may seem large, in fact dropping a single plate can create dozens of sherds. The total ceramic count for the officers' quarters is particularly scant, reflecting either the small number of vessels in use or a more orderly life-style. At the same time one might assume that officers' meals were prepared elsewhere before being served in the quarters, hence breakage was more likely to occur elsewhere. The only anomalous ceramic types present are the hardwhite and yellow-ware, which date to the later 19th century; all 59 sherds of hardwhite were clustered at the western end of the barracks and may reflect the minimal 1864 use of the fort. Indeed, the barracks ceramics of 1814/15 were most heavily concentrated at this same end of the building, though they were found in varying quantities in all test units. Interestingly, the vast majority of sherds from the barracks were found outside the building, indicating that the wooden floor was swept clean of debris at least periodically.

Archaeology: Firearms-Related Artifacts

Of weapons-related artifacts, few were found, and these only in association with the barracks. These amounted to 4 gunflints, 2 musket balls, and 2 pieces of swan shot. Presumably more of this class of artifacts would be encountered downhill in the vicinity of the fort's principal batteries and magazine.

Archaeology: Recreation-Related Artifacts

As stated above, an exception to hypothesis 3 was expected with regard to clay pipe fragments. It was therefore a complete surprise to recover no such artifacts in the vicinity of the officers' quarters and just 11 in the barracks (5 stem fragments, 6 bowl fragments). Considering that 225 square feet of the residential part of the fort were sampled, this is an insignificant total and could mean one of two things. Either smoking was not allowed in or near these buildings or the garrison was using non-clay pipes, such as corn-cob and reed. The latter is possible, given the relative unavailability of British-made (Scottish) clay pipes during the period from the Embargo of 1807 and the ensuing War of 1812. To be sure, Fort Sullivan yielded 225 pipe fragments (5), but how many of these were used by Americans prior to the 1814 British capture of Eastport is very hard to determine. In any case Eastport was a paradise for smugglers from 1807 on, so no safe comparison can be made. Of the handful of Fort Edgecomb pipes, no makers' marks are present and the only decoration represented is one fragment of a fluted bowl, identical to one of the same period from Fort Sullivan (6).

An equivalent finding relates to fragments of alcoholic beverage containers, as summarized below:

 Officers' QuartersBarracks
Wine Bottle Fragments 19   9
Beer Bottle Fragments 1   6
Wine Glass Fragments 0   1
  ______   ______
Totals 20   22

On balance the recreation-related artifacts are so minuscule in volume that the authors strongly suspect that the garrison was prohibited from smoking and drinking within the precinct of the fort. Perhaps such a policy was enforced precisely because the men were militia, rather than regulars, and therefore suspect in the eyes of their Regular Army officers. Further excavations on the site might show that an as yet untested building, such as the eastern barracks, was set aside for "sins." For the moment this is all that can be said.

Archaeology: Archaeology: Personal Context Artifacts

Military buttons

Fig. 18. Pewter Military Buttons.

Brass buttons

Fig. 19. Plain Brass Buttons.

The only artifacts of a personal context which were recovered were 44 buttons, 1 cufflink, and 1 small iron chain. These were distributed between the buildings as follows:

 Officers' QuartersBarracks
Buttons: Military, Pewter 0   9
Buttons: Civilian, Pewter 1   6
Buttons: Plain Brass 1   7
Buttons: Decorated Brass 0   2
Buttons: Bone 1   6
Buttons: Iron 0   2
Buttons: Shell 0   1
Buttons: Jet 0   1
Cufflinks 0   1
Chain, Iron 0   1
  ______   ______
Totals 3   43

As the above totals indicate, most of the artifacts of this class were encountered in the area of the barracks. Figure 18 shows the range of pewter military buttons found, most of which simply carry the letters "U.S." These are "General Service" buttons, issued to all military personnel beginning in 1808 and most heavily during the period 1812-15. There were for use on fatigues, but were put on uniforms if regimental buttons were lacking (7). Less common are infantry buttons, first issued in 1812, featuring a script "I" (figure 18, second row, third from left, third row, second from right) (8). One example of the 1798 infantry issue was found (figure 18, second row, third from right). These were struck in anticipation of the raising of twelve new infantry regiments to supplement the four then in existence. Since the new units were never formed, buttons with numbers 5 through 16 were subsequently issued as surplus to militia (9). If is entirely appropriate, therefore, to have found a button for the mythical 7th Regiment at Fort Edgecomb. Another type of military button (figure 18, second row, second and last; third row, second, fifth and last) represent the infantry issue of 1811. One of these (figure 18, second row, last) carries the inscription "4R" beneath the eagle, and thus is solely a regular army button of the period. Perhaps one or two regulars stopped in at the fort to help train the militia. This, however, is as yet undocumented. Significantly, Fort Sullivan has likewise produced at least one button from an existing regiment which is known not to have served there (10). Quartermasters may have been less than exact in issuing button consignments. Finally, figure 18 (second row, first) shows another infantry issue of 1811, with an eagle and surrounding inscription, "INFANTRY REGIMENT"; the name "Richards," a small eagle, the letters "? ? ? ES" appear on the reverse, constituting the maker's mark. Efforts are continuing to trace Mr. Richards.

Civilian buttons

Fig. 20. Civilian Buttons of Bone, Shell and Jet.

Misc. items

Fig. 21. Miscellaneous Personal Items.

Plain brass buttons with flat fronts, common in all periods from the 18th century on, are illustrated in figure 19. They could have been used by civilians or civilian militia. Figure 20 illustrates the 7 flat disc buttons of bone, commonly found on 18th- and 19th-century sites of all types and used primarily as shirt fastenings (collar, front, cuffs). Also illustrated (second from right and last) are a civilian shell (mother-of-pearl) button and a cone-shaped civilian button of jet. Figure 21 illustrates the remaining miscellany of civilian personal items found: buttons of pewter, iron and brass, brass cufflinks of the late 18th or early 19th century and a short length of a small iron chain which may have been part of some military gear. Such a miscellany of apparently civilian items was likewise found at Fort Sullivan (11).

Although most of the buttons, both military and civilian, were found in the area of the barracks, they were fairly evenly scattered about the site; therefore, the military ones should not represent an extra supply left behind in 1815. Given the short duration of the fort's occupation and the small size of the garrison, it was a surprise to find so many buttons lost or discarded by the enlisted men. In the realm of pure speculation, let us imagine troops quartered in a marginal building, cut off from women, tobacco and alcohol for long periods of time, bored by inactivity, and angry about the most unpopular war in New England's history. Perhaps there was a lot of fighting amongst themselves (imagine the buttons popping off shirts and coats), or perhaps when their duty ended they cut off their military buttons, so representative of a hated Federal Government, and cast them away as a defiant gesture. Further excavation might just help to explain this anomaly.


The 1985 excavations at Fort Edgecomb yielded diverse information on two of the former buildings in the complex. Until this work took place their exact locations were unknown except on an unverified map, and there was no information on their exact dimensions and how they were built. One could only speculate on the potential associated artifact assemblages. These things are now all known–at least in part. There is, however, much more than can be done. The storehouse, eastern barracks and gun emplacements remain uninvestigated. The locations of a well and privies are unknown. And more could be reveal3ed of the two buildings which were studied: where were their hearths? Where were their doors and windows? What were the internal divisions within each building? Are the 1985 artifacts truly representative, or is this small sample deceptive in one or more ways?

Nonetheless the four weeks in 1985 have been revealing, and not just because hypotheses were shattered. The data will give a whole new dimension to the Maine Bureau of Parks' management plans for the park, and they have and will likewise add a whole new dimension to the interpretation of the site for the visitor. The important by-product of public benefit and education did not suddenly cease when the test units were back-filled and the crew drove away. Fort Edgecomb will never again be the same for the historian, archaeologist, school student, and tourist.


The authors are grateful to Earle G. Shettleworth, Jr., Director of the Maine Historic Preservation Commission and Herbert Hartman, Director of the Maine Bureau of Parks and Recreation for their support of the co-sponsorship of the project. Sheila McDonald, Interpretive Specialist for Parks, was invaluable in helping to launch and administer the work. Edward Beach, Supervisor of Operations and Maintenance for Parks, generously provided equipment. Frank Appleby, then Park Manager of Fort Edgecomb, and his assistants, Robert Kuhn and Mellie Giles, were very supportive and graciously put up with the archaeological invasion. Brian Murray, park Manager of Colonial Pemaquid, was very helpful with logistics and equipment. Local residents Alice Boardway and Laurence Davis provided encouragement and useful information. John Wilson, then Archaeologist with the New England Division, Army Corps of Engineers, generously gave a weekend to help, while Ken Thompson of Portland lent his extensive knowledge of the "Second System" to give the project a broader perspective. Nicholas Dean of North Edgecomb volunteered many hours of photography. Finally, our overworked, underpaid and sun-baked crew deserve many thanks: William Hilbish (surveyor), Sylvanus Doughty (architect and excavator), Lois Clark, Sharon Davis and Charles Rand (excavators).

Photograph Credits

Figures 1, 6: Nicholas Dean

Figures 2, 4, 8, 10, 11, 13-15, 18-21: Maine Historic Preservation Commission

Figures 3: Collection of the Maine Historical Society

Figure 5: Re-draft of the original by Christopher Glass for the Maine Bureau of Parks and Recreation


1. Parkman, Aubrey, Army Engineers in New England (Waltham, Massachusetts, 1978), 14. Bradley, Robert L., The Forts of Maine, 1607-1945: An Archaeological and Historical Survey (Augusta, Maine, 1981). 25-27 Patterson, William Davis, "Old Fort Edgecomb," Sprague's Journal of Maine History, 14:4 (1926), 164-179. Dunnack, Henry E., Maine Forts (Augusta, 1924), 90-102.

2. Industrial Journal, September 30, 1892, 1.

3. The original Corps of Engineers map is in the National Archives.

4. Charles Rand's and Neil DePaoli's limited testing in 1982 and 1983 of a lost building at Fort Sullivan in Eastport was reported in DePaoli, Neill, Beneath the Barracks: Archaeology at Fort Sullivan (Eastport, 1986), subsequent to the Fort Edgecomb project. Fort Sullivan, however, is a complex site, involving an American occupation of 1808-14, a British occupation of 1814-18, and American occupation again thereafter through most of the 19th century. It is particularly difficult in places to differentiate between the early American and British occupations.

5. DePaoli, op. cit., 69.

6. Ibid., 95

7. Campbell, J. Duncan, "Military Buttons," Mackinac History #7 (Mackinac Island, MI, 1965), 5.

8. Calver, William Lewis and Reginald Pelham Bolton, History Written with Pick and Shovel (New York, NY, 1950), 153, 155.

9. Ibid., 153-4. Campbell, op. cit., 2.

10. DePaoli, op. cit., 2.

11. Ibid., 88-90.