Archaeological Remains of a Mid Nineteenth Century Brick Clamp: A First Look at Brick Clamps in Kentucky

Derek M. Wingfield, Michael D. Richmond and Henry S. McKelway, Ph.D.
Cultural Resource Analysts, Inc.

---

Abstract

Site 15SH50 represents the first archaeological investigation of a temporary brick kiln, or "clamp," in Kentucky. Bricks and brickmaking have received little attention in archaeological literature considering the almost universal presence of brick artifacts on most historic archaeological sites. The information generated from the excavations at site 15SH50 includes data on the size, design, and operation of the clamp and associated features. The analysis of bricks manufactured in the facility provides details on handmade brick. The clamp is compared to examples of similar facilities in Virginia and Tennessee to begin an examination of regional variations in brick clamps. The excavations at site 15SH50 provide an important component for this research.

Introduction

This paper presents details in the size, design, and operation of the first archaeologically documented brick clamp and associated features in Kentucky. The discussion begins with an overview of early brick manufacturing techniques (McKee 1973, 1976 and Gurke 1987) to place site 15SH50 in context with these early industrial activities. A description of the clamp, associated bricks and regional comparisons follows.

The Process of Brick Making

The use of sun-dried bricks is documented at a variety of prehistoric archaeological sites and "with the extensive use of kilns for firing pottery and furnaces for making metal objects, it is hardly surprising that by 3500 B.C., in Mesopotamia at least, experiments were made with the firing of bricks" (Hodges 1970:77). Although sun-dried mud bricks allowed the building of large construction projects, they tended to disintegrate due to weather, stress, and wear. The use of fired brick added strength and durability to buildings and allowed even larger projects to take place. For reasons of economy, however, sun-dried bricks continued to be used for the majority of construction projects for the next several centuries.

The English brought the knowledge to fire bricks to the New World, and several early kilns are known from the settlement of Williamsburg. The process of firing bricks in small kilns, or clamps, continued until the later nineteenth century when the use of machines to manufacture brick became widespread. Several steps were required in the process of turning clay into fired bricks. A review of these steps provides a background for the description of the archeological remains of the brick making facilities at 15SH50.

The process begins with locating and extracting clay. Clay sources vary in regard to the minerals and "impurities" contained within it. Early local brick making operations had little concern for impurities in the clay. By the early to mid twentieth century, advances in brick making technology allowed a greater degree of control in the clay used for brick manufacture. The clay was dug from borrow pits usually located near the kiln.

Once the clay had been dug, the material was further processed. The clay usually was weathered, left to dry in the open and then crushed again before adding water and any tempering material believed necessary. Temper could be added with a ring pit or a mechanical pug mill. A ring pit was a simple hole excavated in the earth in which clay, water, and tempering agent were mixed. In a pug mill, the clay mixture was placed into one end of a cylindrical or cone shaped conduit. The clay was mixed with blades on the interior of the device and then extruded at the other end. Pug mills often used horses to provide the power to turn the blades until engines became available (Figure 1).

After the clay was mixed, the clay was formed into individual bricks. There were two distinct methods for this process. The first method was by hand. Clay was placed into either metal or wooden molds that may have been sanded lightly to facilitate an easier removal of the formed clay or "green" brick. The clay had to be malleable enough to take shape within the mold. The wetter mixture of clay was termed "soft mud." The second method of forming brick was by machine.

Brick making machines became available in the early nineteenth century; however, few were in operation until the late nineteenth century. Early brick making machines used soft mud, compressing the mixture into molds and automatically removing the excess clay from the top of the mold.

After the green bricks were removed from the mold, they were left out to dry. Moisture in the clay will cause the brick to crack as it is fired. Consequently, as the wet clay bricks were removed from the mold, they were dried in the sun or at times, underneath shelters. Bricks would be stacked, leaving room for air to pass, once the bricks were dry enough to withstand the pressure of stacking.

When the bricks had dried, they were ready to be fired in a brick clamp. The clamp was constructed of unfired brick (Figure 1). Initially, on a level ground surface, brick was stacked in several parallel rows about three bricks thick, leaving gaps between the bricks for air to pass through. Fuel, usually wood, was placed in trenches or flues between the rows of brick. After the height of each row had reached about two feet, arches of brick, creating tunnels, connected the rows of brick. A rectangular structure of brick was built over the arches. The exterior of the clamp would have bricks stacked closely together to maintain the necessary heat within the clamp. The bricks on the interior of the structure would be stacked with gaps to allow the air to circulate within the facility.

The tunnels that were created served as flues for the clamp, channeling the heat throughout the structure once the fuel was lit. The bricks would then be fired for several days. After the firing period ceased, the structure was left to cool for several days. Once the cooling period had ceased the structure of newly fired brick was dismantled and the individual bricks were ready for use elsewhere.

The Clamp

Site 15SH50 is located in southern Shelby County, Kentucky, approximately 3.35 km north of the community of Shelbyville along KY 55 (McKelway et. al.1996) (Figure 2). The land on which site 15SH50 resides is rich in history and was originally part of Benjamin Pope’s Virginia preemption of 400 acres surveyed in the 1770s. Over time the site was absorbed as Painted Stone Farm, that had its nucleus in Squire Boone’s Station, a 1780 pioneer settlement (O’Malley 1996).

Systematic shovel testing identified site 15SH50. The shovel tests from the site yielded high concentrations of brick debris and flecking in a roughly 200 square meter area. The lack of domestic debris suggested that a special use facility had been encountered. Fourteen hand excavated test units were placed within the surface brick concentrations. A thin layer of silt loam with small inclusions of brick was underlain by a thick layer of brick rubble and a zone of fired red to yellowish red clay. This fired soil was overlying a thin lens of what appeared to be a light gray ash. Although the color of this zone suggested ash, the soil was sandy and appeared cemented and hardened. Beneath the brick rubble and ash, the soil was black and granular. This stratum was initially interpreted as a lense of charcoal. The dense brick rubble suggested a structure of some sort; possibly, a brick clamp. The size of the structure was estimated at 10 m by 20 m, oriented generally southwest to northeast (Figure 3).

Trenching across the structure revealed several important characteristics. The stratigraphy revealed brick rubble overlying several ashy semi-circular features and intensely baked, burned soil comparable to the stratigraphy revealed in the units. The semi-circular features observed within the profile wall were interpreted as the archeological remains of brick clamp flues. In profile, the flues have an ashy residue on the interior, which is underlain by fired and baked earth. The soil closest to the fire had been baked and mixed with ashy residue. Beneath the baked ashy lens and at the periphery of the flue channel a blackened burned soil was evident that turned a dark reddish brown to brown as the distance from the center of the flue increased. This soil discoloration was directly related to the variable distribution of heat from the center of the flue. The distance between flues appeared to be about 60 cm and the depth was variable. To expose the kiln in planview, a backhoe was used to strip off the brick overburden from two opposing quadrants.

The edge of the structure at the eastern end was a zone of baked clay. The edge was not straight but rather curved in a northwestern direction. There were sixteen flues detected in plan view (Figure 4 & 5). The flues were separated by areas of baked clay. Alternating bands of gray ash and yellow/red burned flue soils contrasted with the baked red clay on either side of the flue, creating a striped pattern in planview (Figure 6) and a layered pattern in profile (Figure 7).

In planview, the heads of the flues were semi-circles of flecks of charcoal and two zones of soil reflecting the heat with which the soil was subjected. A reddish dark brown soil was at the outer edge of the flue, but a yellowish red soil was on the inner edge of the ring around the flue entrance. Burned limestone appeared to have been displaced from around the opening of one flue. Within the flue channel, patches of gray ash and yellowish brown soil were evident. An attempt was made at removing the fill from one of the flues. The gray ash-like remains of the flue were so shallow that little was learned. There was burned and baked soil beneath this ash lens indicative of the extreme heat generated.

A depression was located 35 m north of the site. This appears to be the borrow pit where the clay was extracted for the brick manufacture. A spring pond to the north may have provided necessary water to process the clay. Shovel tests south of the kiln noted silt loam soils with flecks of clay. It is possible this area was used to dry and process clay and possibly mold the clay into bricks. The archeological remains at site 15SH50 appear to represent the full array of brick manufacturing; from processing the clay to firing the bricks. As important as a description of the kiln facility is a description of the associated produce–the bricks.

The Bricks

The bricks from site 15SH50 are valuable because they are directly associated with a handmade brick manufacturing facility. A description of these bricks contributes to a list of criteria used to classify handmade versus machine made bricks. The methodology used to analyze bricks from site 15SH50 generally followed the suggestions of Hockensmith (1996:27). The bricks recovered from site 15SH50 certainly warranted special attention, because of the direct association with a crude brick kiln or "clamp." The bricks were inspected for a variety of marks and attributes that Hockensmith (1996) used in the analysis of bricks from the Kentucky History Center.

Attributes recorded by Hockensmith (1996: 27) included the type of brick, whether the brick appeared to have been manufactured from soft or stiff mud, manufacturing marks, such as strike lines, firing marks, brick size and color. The bricks from 15SH50 were examined for any attributes usually associated with machine manufacture, such as perforations through the brick. Pressed brick brand names were recorded as well as any apparent flaws in the manufacture of the bricks, e.g., cracks and warps. Lastly, any imprints, such as animal tracks or marks made on the bricks, were recorded.

Other researchers who have described Kentucky bricks include Black (1987) and Deiss (1987). Black described various marks found on bricks, including chicken tracks and human thumbprints. Several marks and numerals described by Black (1977:101) are interpreted to indicate counts by the brickmaker inscribed into soft brick. Black suggests that some marks found at different house sites are so similar as to indicate the bricks were manufactured at the same brickyard.

The assemblage from site 15SH50 included 14 bricks and brick pieces large enough to warrant additional analysis. General characteristics of the bricks are described below, as well as specific measurements when appropriate. All the bricks recovered at site 15SH50 were common bricks, which had no frogs or impressions made into the clay when wet (Figures 8 & 9). The bricks recovered from site 15SH50 also lacked inscribed maker’s marks or brand names.

Table 1. Attributes of brick recovered from site 15SH50.

* Brick measurements on two bricks recovered from backdirt piles at 15SH50 (Charles Hockensmith, personal communication, 1997).

Measurements of the bricks included length, height, and width recorded in English measurements for ease of comparison with other works. The number of surfaces present on the brick and the number of vitrified surfaces, on both exterior and interior surfaces, were noted.. The color of the interior of the brick was recorded. A subjective assessment of brick hardness was made ranging from 1 (soft) to 3 (hard). Table 1 presents this data. Note that dimensions of two bricks recovered from the backdirt pile were provided by Charles Hockensmith of the Kentucky Heritage Council.

Unfortunately, the sample size for whole bricks at site 15SH50 was extremely small and comparative statistics would not be fruitful. Hockensmith (1996:61) notes that the size of bricks found at the Kentucky History Center site exhibited a wide range of sizes. Early soft mud bricks generally exhibite a range of sizes. McKee (1973: Table 7) provides brick measurements from a variety of sites dating from Roman times through the early twentieth century when brick sizes became standardized. He (1973:53) states that English bricks of the early nineteenth century had an average size of 9 inches in length, 4 ½ inches in width and 2 ¾ inches in thickness. Generally, there is a great deal of variation in brick sizes shown.

There was a standardized measure adopted for brick sizes at the advent of the twentieth century. The dimensions of the National Brickmaker’s Association were 8 ¼ inches in length, 4 inches in width and 2 ¼ inches in thickness. The bricks from site 15SH50, as shown in Table 1, are smaller in every dimension than the twentieth century standards for brick. The bricks from site 15SH50 appear to be small in comparison to bricks recovered by Hockensmith (1996:62) at the Frankfort History Center and at the Maysville brickyard (Hockensmith and Stottman 1996). The bricks from site 15SH50 also are smaller than the common bricks being manufactured by the Shaker community in the early nineteenth century as described by Deiss (1987). There has been a considerable degree of variation in brick sizes through time, and the dimensions of the bricks from site 15SH50 are yet another example.

As expected, there are some general comments on the bricks from site 15SH50 that are considered diagnostic of hand made bricks. First, the bricks exhibit irregularity in shape and form. Table 1 shows the marked degree of variation in brick length, width and height. The differences in brick width are as much as one half of an inch. These size irregularities may result from differences in brick molds, the amount of clay shrinkage during drying or differential heat within the kiln.

Variations in paste mixture and heat within the kiln create differences in the hardness and the color of the bricks. The variations associated with vitrified, glassy surfaces, for example, are attributed to uneven temperatures within the kiln and the proximity of individual bricks to the fire. Many of the broken bricks displayed interior surfaces that were vitrified. This is caused by bricks cracking and spalling during the firing. The exposed surface on the interior of the brick is subjected to the same degree of heat as the exterior surfaces and these surfaces may become vitrified. Many of the bricks had surfaces that were vitrified; however, there was no pattern in the glazing that provided clues to the specific position of individual bricks within the kiln, or to the manner in which the bricks had been stacked.

The bricks from site 15SH50 had a sandy exterior. Sand was often applied to the brick molds so that the bricks could be removed more easily. Parallel striations or "strike marks" on the surface of brick offer evidence of excess clay being removed from the brick molds. Additionally, there were laminations present in the paste of many of the bricks when viewed in cross-section. These laminations are do not appear to have been manufactured by an early brick making machine that forced and compressed the soft clay into the brick mold.

The broad surfaces of the sides of the brick were concave, especially on the faces of bricks that exhibited strike marks. Lips were evident on the edges of the bricks. The bricks were inspected for any depressions that might suggest how the bricks had been stacked, but none were noted.

Comparative Examples

The early brick manufacturing facilities at site 15SH50 are the first investigated in Kentucky, and as such, the kiln might reflect regional traditions in the construction and operation of the facility. A review of other similar facilities was undertaken. Metz and Russ (1991) provide descriptions of facilities in Virginia. Of the seventeenth and eighteenth century clamps reviewed, they ranged in size from 3.4 x 2.7 m to 30.5 x 5.2 m. The brick clamp at Liberty Hall in Lexington, Va. was 6.7 x 4.9 m in size and was constructed in 1799 (Figure 10). This kiln was composed of five "benches" or rows of brick. Between each bench of bricks were the heat channels or flues that were 0.6 m in width. Flues were located on the exterior of the outside brick benches and had been dug into the surrounding subsoil. The heat generated from these flues heated the subsoil creating a baked clay edge for the structure. The clamp was oriented so that the flues and benches extended in a northwest to southeast direction. Metz and Russ (1991) note that other Virginia clamps reviewed for their study had the same basic design.

The Zimmerle kiln, located in Marshall County, Tennessee, was investigated by Smith (1989). This clamp measured 7.5 by 7.0 m and was oriented north south (Figure 11). This clamp contained seven internal brick benches and two narrow brick benches located at the outer edge of the clamp. The width of the internal benches and the flues was about 0.5 m. The one notable and surprising difference in the Zimmerle kiln was the presence of a central "baffle wall" which probably prevented strong cross drafts within the clamp.

Archaeological investigations noted that the debris overlying the structure of the clamp consisted of "a two to three inch layer of consolidated brick dust, bonded with lime and ash" (Smith 1986:139). Smith also noted that this zone was formed after the clamp was abandoned. Beneath this overlying zone was a baked clay floor, hardened by the firing of the clamp as was seen in the Liberty Hall structure.

An important note by Smith is that he believed the clamp operators also fired limestone to use as mortar in the planned construction of the mansion house. This interpretation was based on a concentration of limestone found near one of the flue openings at the edge of the kiln. Lastly, Smith (1986:138) discusses the expected archeological correlates of these early temporary brick kilns:

Such remains will usually consist of nothing more than one or two bottom courses of the kiln or clamp, which were left behind simply because these bricks were not fired well enough to be useable. In some cases all traces of the brick benches may be gone; yet, the presence of the kiln may be indicated by an alternating charcoal stained and baked clay, stripped pattern (the former representing the fire channels; the latter where the benches stood) (Smith et al. 1977:82).

Despite the fact that the clamp at site 15SH50 is much larger than those discussed in Virginia and Tennessee, there are obvious similarities to the previously described facilities.

The basic component at all the brick clamps is the flues that channel heat throughout the structure. At the Liberty Hall and Zimmerle kilns, walls of unfired brick arched to form these channels, creating a generally rectangular structure with tunnels running from side to side. At each of these clamps, the bottom bricks that formed the sides of the flue remained, most likely because the bricks were not fired well enough. At site 15SH50, there were no bricks forming the sides of the flues, only the bottom of the flue channels that evidenced the grayish ashy soil remained. As with Liberty Hall, the distance between flues at site 15SH50 was about 0.6 m.

Smith noted a baked clay floor at Zimmerle and this was evident at Liberty Hall as well. At site 15SH50, baked areas were evident between the flues. Smith (1986) predicted a striped pattern would be evident in kilns where all the brick had been removed. This was the case at site 15SH50. The flues were apparent in plan view and in profiles as shallow gray ash and yellowish red burned soil. Between the flues, where the stacks of brick presumably rested, the clamp floor was baked and exhibited a red or pinkish red color. The individual bricks recovered from 15SH50 were about 8 in length, which indicates the brick benches between the flues was probably three bricks wide. The bricks resting on the floor of the clamp apparently conducted enough heat to fire the area beneath them.

An important point here is that the flues at site 15SH50 represented only the very bottom of the heat channels of the clamp. The clamp manufacturers did not dig into the subsoil to create troughs for the flues. The semi-circular appearance of the flues in profile is produced by the intensity of the heat on the floor of the fuel-lined flue, and the depth to which the soil underlying the fuel source is baked or burned. The ashy gray lense appears to be the remains of the fuel source. The black zone beneath, which covered the entire structure, might be described as "smoked;" soil subjected to heat but within a low oxygen atmosphere.

One difference concerning the flues between these kilns, was the configuration at the head of the flues. At Liberty Hall and at the Zimmerle kilns, the heads of the flues formed a straight line across the facility. At 15SH50, the heads of the flues were semicircular in plan view and extended beyond the burned clay edge for the interior of the structure. It appears that the flues were lit or burned for a period outside of the actual structure. This configuration may have facilitated refueling of the flue fires.

Devices could be placed at the heads of the kiln flues to regulate airflow and the intensity of the fire. Flues were sealed with unfired brick, or could be sealed with limestone. Possibly, limestone was heat treated at site 15SH50 for use as mortar. The limestone may have been placed at the heads of the flues to seal the fires and heat treat the limestone.

A major distinction between the clamps in Tennessee, Virginia and site 15SH50 relates to the presence of a shallow ditch or trench along the periphery of the structure. At 15SH50, the trench bounded the clamp on three sides. The function of the trench may relate to Smith’s discussion of "scove kilns." According to Smith, mud was used in these kilns to plaster the walls of green brick to help retain heat inside the structure. A ditch or trench probably was dug for clay along the north, south, and western periphery of the kiln to use in plastering the walls. This seems to be the most plausible explanation for the trench at site 15SH50. The limestone and brick debris evident in the trench most likely was deposited in the trench as the bricks were removed from the clamp and debris from poorly fired or cracked bricks accumulated in the open ditch.

Summary

Bricks recovered from site 15SH50 display common characteristics of hand made bricks including irregular dimensions, strike scars, a highly variable clay mixture, variable brick hardness, and compression lines. All the bricks exhibited a sandy exterior surface resulting from the application of sand to lubricate the brick mold. The brick forms could be removed more easily from the mold this way. This research also has contributed samples of brick useful in documenting the characteristics of hand made bricks; before mass production of bricks took hold in the late nineteenth century. An important diagnostic of brick clamps may be fragments of brick that are vitrified on the interior surface of the brick. These brick remains are unlikely to be found at residential sites; they would be considered waste and not transported from the clamp.

The investigations of the clamp at site 15SH50 reflect the first investigations to record a small clamp facility to fire hand made bricks in Kentucky. Smith (1986:142) notes that these local kilns, created when the need arose for new bricks in a planned structure, should be one of the most common historic site features. Yet, there is very little archeological data concerning these facilities. The clamp at 15SH50 is valuable for the comparative material it provides on the characteristics of local clamps.

By establishing and recognizing characteristics common to brick clamps, such as dimension, orientation, associated facilities, and soil patterning, archaeologists may better identify and document variation in brick clamps across regions. The excavations at site 15SH50 provide an important component for this research.

Acknowledgements

The authors would like to thank Mr. Charles M. Niquette of Cultural Resource Analysts, Inc. for the use of valuable time, equipment and facilities. Mr. Grant L. Day, also of Cultural Resource Analysts, Inc., should be recognized for suggestions and comments. Cultural Resource Analysts’ archaeological testing of site 15SH50 was a part of the archaeological assessment required for the Kentucky Transportation Cabinet upgrade of KY 55. The authors would like to thank the Kentucky Transportation Cabinet for the use of the data generated from site 15SH50. We also would like to recognize the guidance and help provided by Mr. Charles D. Hockensmith of the Kentucky Heritage Council and Mr. Kurt Fiegal of the Kentucky Transportation Cabinet’s Environmental Division. Without their assistance, the realization of this paper might not have been possible. Illustrations appearing herein as Figure 1, Figure 10 and Figure 11 have been reproduced with the permission of their respective authors and/or publishers.

This paper, with minor revisions, will be published in the Ohio Valley Historical Archaeology: Journal of the Symposium on Ohio Valley Urban and Historic Archaeology Volume 12 (pages 68-88), edited by Mr. Donald B. Ball. The volume should be available in late 1997 or early 1998. For additional information, the reader is directed to the original contract report (McKelway et al. 1996).

References Cited

Black, W. R. Jr. 1987 Anomalous Marks in Nineteenth Century Bricks. Proceedings of the Symposium on Ohio Valley Urban and Historic Archeology 5:96-104. Edited by D. B. Ball and P. J. Diblasi. Louisville.

Deiss, R.W. 1987 Shaker Brick Types from South Union, Kentucky. Proceedings of the Symposium on Ohio Valley Urban and Historic Archeology 5:96-104. Edited by D. B. Ball and P. J. Diblasi. Louisville.

Gurke, K. 1987 Bricks and Brickmaking: A Handbook for Historical Archeology. University of Idaho Press, Moscow.

Hockensmith, C. D. 1996 Bricks and Brickmaking in Frankfort: An Archival and Archaeological Study of the Kentucky History Center Site. Kentucky Heritage Council Occasional Reports in Archeology, No. 1, Kentucky Heritage Council, Frankfort.

Hockensmith, C. D. and M. J. Stottman 1996 The Maysville Brick Company, Mason County, Kentucky. Journal of the International Brick Collectors Association 14 (1):3-15

Hodges, H. 1992 Technology in the Ancient World. Barnes and Noble, New York.

McDaniel, J. M., K. C. Russ, and P. B. Potter 1994 An Archaeological and Historical Assessment of the Liberty Hall Academy Complex 1782-1803. The James G. Leyburn Papers in Anthropology 2. Liberty Hall Press, Lexington.

McKee, H. J. 1973 Introduction to Early American Masonry: Stone, Brick, Mortar and Plaster. The Preservation Press, Washington D.C.

1976 Brick and Stone: Handicraft to Machine. Building Early America: Proceedings of the Symposium Held at Philadelphia to Celebrate the 250th Birthday of the Carpenters’ Company of the City and County of Philadelphia:74-95. Edited by C. E. Peterson. Astragal Press, Mendham.

McKelway, Henry S., Michael D. Richmond and Robert B. Hand 1996 An Archeological Reconnaissance of Proposed Parcels of Kentucky State Highway 15 Upgrade, and Phase II Investigations of Site 15SH50 Shelby County, Kentucky. Contract Publication Series 96-42. Cultural Resource Analysts, Inc., Lexington.

Metz, J. D. and K. C. Russ 1991 Understanding Traditional Brickmaking in Virginia. Quarterly Bulletin of the Archeological Society of Virginia 46: 96-106.

O’Malley, N. 1996 Squire Boone’s Painted Stone Station: A Locational Survey. University of Kentucky Archeological Report 368. University of Kentucky, Lexington.

Smith, S., F. W. Brigance, E. Brietburg, S. D. Cox, and M. Martin 1977 Hermitage Brick Kiln Investigations. In Results of the 1976 Season of the Hermitage Archeological Project. Ms. on file, Ladies Hermitage Association and the Tennessee American Revolution Bicentennial Commission, Nashville.

Smith, S. D. and C. A. Watrin 1986 Zimmerle Brick Kiln. Tennessee Anthropologist, Vol. XI (2):132-144.