Nine Gal Tavern


Lenville J. Stelle
Parkland College
Champaign, Illinois


©1990 by the Center For Social Research, Parkland College




Note: Parts of this report were presented as a paper at the 1990 Society for Historical Archaeology Conference on Historical and Underwater Archaeology.

Significance of the Site and Research Objectives

The first half of the nineteenth century saw the westward expansion of frontier America into east central Illinois and the upper Sangamon River basin. By the 1830s, where trails and roads intersected the river, pioneer communities began to develop. One of the most common economic expressions in these emerging communities was the tavern. Middletown, located in western Champaign County, was no exception. The first thing that Daniel Porter did after platting the village at the Sangamon ford of the Fort Clark Trail was build a tavern and general store. During the next 30 years at least four more taverns would do business in this community: the Mathew Johnson Tavern, the Rea Tavern or American House, the Nine Gal Tavern, and the Ohio Tavern. The most historically conspicuous of the taverns has been the Nine Gal. It has become the object of an annual historical reenactment, or passion play, by the local Methodist Church. In spite of this interest, local history is surprisingly unclear as to the Nine Gal Tavern's location.

The primary objectives of the present research are therefore exploratory and descriptive: (1) to locate the site of the Nine Gal Tavern; and (2) to recover artifactual material indicative of its function as a tavern. The data bearing on these objectives has significance well beyond the vagaries of local history. The cultural expression of the tavern was an important commercial activity of the pioneer period. The regional form of the tavern is poorly documented in the archeological literature. Achieving these objectives would afford information on this significant historical phenomenon.

Archival Discussion

In the thirty years followings its inception in 1832, Champaign County would develop into a civilized, well populated area complete with towns, schools, churches, roads, and a railroad line. During this period of intense expansion and development, inns or taverns were established along most major roads. The main function of these taverns was to provide food and lodging for man and beast, although beverage alcohol was commonly dispensed. The emergence of the tavern as a differentiated commercial activity reflected the increasing socio-economic complexity of a community.

The signing of the Black Hawk Treaty in 1832 helped trigger immigration to most areas of Central Illinois (Pease 1925:118). The following year, as the last of the Native American societies were relocated beyond the Mississippi River, pioneers were already slowly moving into what had just become Champaign County (Pease 1925:119). Settlements began to appear along the fringes of the corridor forests that were expressed parallel to the banks of the prairie rivers (Cunningham 1905:645). It was there that a plentiful supply of wood, water, and game could be found. The main route transecting Champaign County east to west was the Fort Clark Trail. It connected Danville and Peoria (Fort Clark). Some of the first villages in the county were formed along this trail. They included, from east to west, Homer, Sidney, Urbana, and Middletown (Morgan 1969:18). Middletown was later renamed Mahomet. It was in Middletown Township, in the year 1833, that the history of the Nine Gal Tavern began.

In that year John Bryan (also spelled Bryant and Briant in the historical literature) married Malinda Busey (25 July 1833) in the first solemnized marriage performed with a Champaign County license (Cunningham 1905:684). Both were from Shelby County, Kentucky. On 10 August 1833, Bryan entered a 40 acre tract of land in Section 14 of Middletown Township immediately adjacent to property patented in 1832 by his father-in-law, Issac Busey (Cunningham 1905:684 and State of Illinois 1982:51). The Bryans were the third family to settle in the township.

On May 5, 1834, John Bryan purchased from Champaign County, the west 1/2 of the southeast 1/4 of Section 14, Township 20 North, Range 7 East (State of Illinois 1982:274). This property was located on the south side of the Fort Clark Trail, approximately one and a quarter miles east of the Sangamon ford. In the following year, 1835, Scott Fielding, another early pioneer, plowed a furrow from the river crossing to Urbana thereby improving and shortening the transportation link (Fort Clark Trail) between the county seat and the area referred to as the Sangamon Timber Settlement. Early in 1836, the Urbana to Bloomington State Road was completed along this furrow (Champaign County Commissioners' Record Book A). Later in that year the Champaign County Commissioners ordered a county road to be opened from the State road "...near the house of John Bryan" to the Decatur Road. This became a major traffic corridor from Urbana to Decatur since the prairie sloughs of the Cerro Gordo Till Plain were frequently impassible. The proximity of the homestead to these roads and the river was of central to its function as a tavern.

At this location, later commonly referred to as the Timber Edge Farm, John Bryan lived with his family (six children were listed in the 1850 census) and in the ensuing years prospered as a farmer (Roehm 1986:16). The Agricultural Schedule for the 1850 census indicates that he was a wealthy farmer with 400 acres of improved and 600 of unimproved land and 90 head of mixed livestock. Production for the year ending 1 June 1850 included 400 bushels of Indian corn, 75 pounds of wool, 14 bushels of Irish potatoes, $100 worth of orchard produce, 300 pounds of butter, and 12 tons of hay. As the fourth wealthiest farmer in both the county and township, his real property was assessed at $10,000 (Roehm 1986:16).

Maps, land records, and numerous references indicate that it was Bryan's home that subsequently became the Nine Gal Tavern. The house was a curious structure for the times. Bryan's affluence made possible the sawed lumber (reportedly freighted from Covington, Indiana) of which the structure was fabricated (Abbott 1985:32). The house was a fairly large structure (Abbott 1985:32). The composite of all descriptions and pictures depicts a rambling two-story building, some 40 feet long. Spanning the entire front was a covered porch supported by plain columns. A black-and-white photograph reproduced in Morgan (1969:19) and a sketch signed "Hazen" reveal a New England salt-box type structure (Morgan, in personal communication, indicated that the photograph was supplied to him by unremembered parties and, as a consequence, he could not attest to its veracity; the "Hazen" sketch was apparently produced sometime in this century and is based upon unknown sources, Purnell [1955:79] attributes it to Fred Hazen). In both illustrations the gables were dominated by large, single flue, masonry chimneys. The main entrance was located in the center of the building's front. Each document shows windows on both stories of the front side, although the number and placement differ. A description of the building appearing in the Mahomet newspaper (Abbott 1985:181) indicates that "...a number of small rooms, each with a small window were on the second floor." The exterior of the structure was sheathed with horizontal, lapped siding. The rear of the building was not revealed and one cannot determine if there was a summer kitchen. The documents and descriptions provide no information on out-buildings.

It is not known how long John Bryan and his family resided on this farm. In 1848 Bryan purchased a forty acre parcel of land from the government in Section 23, the next section south (State of Illinois 1982:51) and the 1850 census continued to place the family at this location. The 1860 census reveals that the family had relocated to the town of Champaign with John's occupation listed as livery keeper.

We do not know if the Bryan's kept tavern during their tenure on the homestead. The Champaign County Commissioners had begun regulating and licensing taverns in 1836 and Daniel Porter was issued a license for a store and tavern in Middletown village in the same year. Significantly, the County Commissioners Record through 1850 gives no indication of a tavern license having ever been issued to John Bryan. However, an 1878 county history (Brink, McDonough and Co. 1878:125) asserts that John "Bryant" kept "...the first tavern." in Middletown Township. If we are to believe the Brink, McDonough and Company account, then Bryan would have begun keeping tavern prior to 1836. It is not clear why the fourth wealthiest farmer in the county would have opened his house to this business. Assuming that Bryan did keep tavern, it was most likely conducted in a casual fashion, incidental to his farming activities, and with rewards more social than financial. The location at the intersection of the Urbana-to-Bloomington and Urbana-to-Decatur trails, as well as the proximity of the Sangamon ford would have made the Bryan farm a natural stopping point. During wet weather, when the river was high and the prairies were full, the concentration of travelers would have been even greater.

Sometime after 1850 the Bryan family left the homestead. In 1853 the building was leased to Thomas A. Davidson and family, new arrivals from Ohio (Mathews and McLean 1979:34). The Davidsons transformed the dwelling into the "Ohio Tavern" (Abbott 1985:32). The site now functioned as a commercial tavern. The name Ohio Tavern likely derived from the measurable presence of local settlers from Ohio, as well as from the nativity of the family. Davidson operated the tavern from 1853 through 1856 (Mathews and McLean 1979:34) and it was during this period that Abraham Lincoln was alleged to have stayed there while traveling the Eighth Judicial Circuit (Abbott 1985:32). Folk history suggests that Lincoln would stop at the tavern as he traveled from Decatur to Urbana and that he enjoyed bouncing little Jimmy Davidson on his knee. The tavern business may have been quite lucrative. After purchasing several hundred acres of land, the Davidsons moved off the property in 1856 and began farming (Champaign County Gazette 1880). The 1860 census indicates a real property assessment of $12,000 and personal property of $1,490. He and his wife, Elizabeth, had seven children and one farm laborer living in the household at that time.

With the departure of the Davidsons the documentary evidence grows thin. There is no primary evidence of subsequent tavern proprietors. What is known is that the property remained in the Bryan family until 1864, a year after John's death (6 July 1863), at which time it was sold by Malinda to the wealthy neighbor, B. F. Harris (Champaign County 1979). Harris was the wealthiest man in the county and resided in a "mansion" several miles to the southwest. He presumably used the Bryan building to house a series of his tenant families. However, insofar as the structure was referred to as the "Nine Gal Tavern" by twentieth century historians, there may have been at least one other tavern keeper on the property after the Davidsons. Popular history has it that the tavern was so named to honor the later tenant's nine red-headed daughters (Morgan 1969:19 and Abbott 1985:32). If this is true then the occupation would likely have occurred between 1856 and 1859.

The 1850 census records six men in the county indicating innkeeper as their occupation (Roehms 1986:11). Two of them were from Middletown Township. One of these was Mathew Johnson, associated with the tavern of the same name, and the other was LeG. D. Robertson. Robertson was from Kentucky, married, and had two young sons. His worth is listed as $60. The place of Mr. Robertson's business remains unclear, although our best guess was that he managed or worked in the establishment originated by Daniel Porter. While Mathew Johnson's occupation is listed as "farmer/innkeeper", John Bryan's is recorded as simply "farmer." In the 1860 census the occupation of Mathew Johnson is recorded as miller and Robertson is gone. In this census no one from the township reported their occupation as that of innkeeper.

Arguing from the 1860 census data, the site had by 1860 returned to its previous function as the residence of a single farm family. However, the social difference between the Bryan occupation and the post-1860 occupation was that between an owner and a tenant (The Breeder's Gazette 1911). While the specific identities of these latter families remain unknown, they were only middle class farm managers. The Bryan construction survived until 1891 when it was razed and replaced by the present building (Isabelle Purnell, personal communication).

The present structure was completed in 1892 by the Harris family and was a substantial structure in its own right. It was obviously intended for an important family in their system of agri-business. By 1905 the farm was in the possession of a grandchild, also named B. F. Harris. The farm now consisted of 320 acres. This Harris's approach was that of the aggressive businessman operating a highly specialized production unit. A detailed description of agricultural strategy and production, as well as a photo of the "superintendent's residence" appears in a May 1911 lead article of The Breeder's Gazette. A net profit of $6,500.70 for 1910 derived from the sale of 50 steers, 294 sheep, and 410 hogs, as well as 1,404 bushels of oats and 3,500 bushels of corn. The article describes the administration of the farm in the following way:

Although he keeps in close personal touch with the affairs of the farm by telephone and visit, Mr. Harris employs a superintendent, who is paid to board and is in charge of the laborers required.

With a total farm labor cost of $1,860.92 for 1910, the superintendent's salary must have been quite modest. The families of superintendents and farm managers continued to occupy the building through the 1960s.

To summarize the historical chronology of the site, the land was purchased from the government in 1834. In that same year a large frame home was constructed. The wealth that it symbolized and its proximity to the Sangamon ford and a major road intersection made it a natural stopping point for travelers. It is likely that the Bryan family provided assistance to travelers but never became formally involved with the business of tavern keeping. This condition changed by 1853 when the Davidsons opened the Ohio Tavern. There was probably at least one other tavern keeper (Nine Gal Tavern) after the Davidsons. However, by 1860 the property had reverted back to its original function as a single family residence associated with the farm. In spite of a change in ownership in 1864, this condition obtained until 1891 when the house was demolished and replaced by the existing structure. After 1891 the new residence was reserved for a series of farm superintendents and managers.

Site Description

As previously indicated, the location of the Nine Gal Tavern is an issue of some debate. Locations offered by local historians number four: (1) just east of the Sangamon ford and north of Bloomington Road; (2) east of the river but on the south side of Bloomington Road; (3) one-half kilometer further east on Bloomington Road also on the south side; and (4) the location on the Timber Edge Farm, approximately 1.1 km east of the ford. Our research centers on this last locale because our reading of the period literature and the government land sales records to John Bryan suggest it to be the most likely site.

The buildings of the farmstead (Figure 1) are situated upon a small loess covered knoll lying down slope of the crest of the Champaign Moraine. The Champaign Moraine is a late Wisconsinin event dating from perhaps 16,000 BP. While the knoll may only be an erosional expression, its origin may have been as a small ice contact feature. The Sangamon River breaches the Moraine some 1.1 kilometers to the west northwest. This is an area of relatively bold topographic relief. (For a detailed analysis and reconstruction of the natural environment as it existed in Middletown Township in 1820 see Cultural Resources Survey: An Environmental Summary Of The Lake Of The Woods [Stelle 1986]).




Figure 1. The Nine Gal Tavern Site.


Research Design and Data Collection

As was previously indicated, the objectives of the present study are exploratory and descriptive. In its higher registers it falls within that range of investigation labeled culture-historical. While our interests are driven by a model of community development and processual concerns, the dearth of regional archeological data on this time period, specifically economic expressions and their attendant material assemblages, makes it difficult to test meaningful hypotheses. The research questions addressed by the present study are:

The data collection strategy consequently focused on activities that would afford systematic site testing and description. Specifically, it involved six elements.

1. An examination of the interior of the existing house and out-buildings. Of particular interest were exposed structural elements. The goal was to identify materials salvaged from the Bryan constructions.

2. Surface collection of large tracts of the farm. Eighty acres stood as relict timber. All of this tract, as well as some cultivated ground was undergoing development as an upper class subdivision by the fall of 1986. The main drainage ways were being reorganized for the construction of two flood control impoundments. Exploiting these disturbances, every exposed patch of earth was examined. Consequently, much of the land surface that had supported forest and savanna was accessed. From the beginning of the survey, the ground that had supported prairie revealed extremely low densities of human detritus; consequently, it received less attention.

3. A systematic shovel probe survey of the house site was conducted. The survey was accomplished using a grid based upon 2 m centers. At each grid intersect a shovel probe 30 cm in diameter was excavated to culturally sterile soil. The grid was anchored at site datum. The shovel probe grid extended beyond the limits of the primary residential scatter.

4. The north-south and east-west lines of the grid were swept with a metal detector. Where readings indicative of a metallic object were encountered, a shovel probe was excavated.

5. To acquire information about what might lie beneath the existing structure, two 1 m by 2 m test units were excavated. The units were positioned parallel to the foundation walls and on either side of the front door. The units were dug to sterile soil.

6. Where subsurface features were encountered test trenches were excavated to sterile soil. They were 30 cm wide and extended beyond the limits of the feature.

The collection of data occurred over a five week period. The work was conducted by various groups of college and seventh grade students enrolled in Parkland College field schools. The seventh graders were only involved with the surface collecting and the shovel probe survey.

Findings



1. Excavations and feature descriptions:

Seven test units were excavated. They included two one meter by one meter units around the in situ foundation stones, two one meter by two meter units directly in front of the existing structure, and three thirty centimeter wide trenches across subsurface features. Two additional subsurface deposits were only explored by shovel probes.

Test Unit #1 was located west of the front door of the Harris structure and parallel to the foundation wall (unit datum S0.43,W2). The unit contained a large, partially exposed glacial erratic (Figure 2) that may have been part of the foundation system of the original Bryan construction. If so, it would seem to have been removed to its present location during or subsequent to the erection of the existing house (a region of disturbance extended completely under the boulder). Earthenwares included redware, hand painted pearlware, undecorated whiteware, black transfer printed whiteware, and nine sherds of indeterminant whiteware. The mean ceramic date for the pearlware and whiteware was 1854. Other artifactual material included: porcelain (1), bottle glass (4), flat glass, pressed glass (1), milk glass (1), metal objects (25), both early and late brick, brushes (2), a shell button, and a bone button.

Test Unit #2 (Figure 3) was located just east of the front entrance to the Harris building and parallel to the foundation wall (unit datum S0.67,E6). Artifactual material included: salt-glazed stoneware (2), bottle glass (26 sherds including an element of a scroll flask), flat glass, pressed glass (3 including a paneled tumbler), a glass button and bead, metal (32), both early and late brick, and a piece of plastic. The earthenware recovery was composed of redware, yellowware, undecorated pearlware, undecorated whiteware, hand painted polychrome whiteware, spongeware, flow blue, and indeterminant whiteware. The mean ceramic date for the pearlware and whiteware was 1845.






Figure 2. Test Unit #1 of the Nine Gal Tavern Site.







Figure 3. Test Unit #2 of the Nine Gal Tavern Site.

Test Unit #3 centered on the first of the foundation stones found in primary context. This stone served as site datum. It was located directly out from the front door to the existing house. The field stone was hidden beneath a thin vegetational mat (Figure 4). Artifactual material included bottle glass (8 fragments one of which was from an historic flask), flat glass, metal (13 items), structural materials (both early and late brick), shell buttons (2), and a piece of plastic bottle. Earthenwares were represented by undecorated whiteware, spatter, green transfer printed whiteware, and mid-blue transfer printed whiteware. The mean ceramic date for the whiteware was 1855.

Test Unit #4 centered on the second foundation stone still in primary context (Figure 5). It was located at N1.7,E2.2. Recovered from this unit were a sherd of flat glass, modern machine cut nails (16), wire nails (2), strap metal (2), early brick, mortar, a Rockingham door knob, a short length of iron wire, and a green glass marble.

Feature #1 (Figure 6) was located at S0.34,W4.25. A 30 cm trench was extended north across the area of subsurface disturbance. The feature, along this axis, was 192 cm wide and extended down to a maximum of 80 cm. The artifactual material recovered (N=196) included fragments of clay smoking pipes, bottle glass (six items, three of which were from historic flasks), flat glass, pressed glass tumblers (23), pressed glass, milk glass, metal (notable among these 64 items were barrel/cask hoops, metal buttons, harness buckles, a wrought iron hook, a clasp knife, a hand wrought nail, machine cut nails, and a .32 cal. rim fire casing), and both early (6) and late brick (2). Earthenwares were represented by redware, Rockingham yellowware, undecorated pearlware (1), undecorated whiteware (3), hand painted polychrome whiteware (6), spatter (2), blue transfer printed whiteware (9), purple transfer printed whiteware (8), green transfer printed whiteware (1), red transfer printed whiteware (6), edge decorated whiteware with a scalloped rim and impressed bud (1), and indeterminant whiteware (31). The mean ceramic date for the pearlware and whiteware was 1851. Excluding the indeterminant whiteware from the calculation yields a corrected value of 1843. Feature #1 is interpreted as an ash pit associated with the fireplace on the west side of the Bryan house. The deposition is assigned to the Bryan occupation (1833-1853).

Feature #2 was located at N9.8,E6.5. A 30 cm wide trench was extended north 50 cm to a point beyond the margin of the subsurface disturbance. The south margin of the feature was not identified. The exposed portion of the feature had a maximum depth of 59 cm. The undisturbed horizon was found from 29 to 59 cm below surface. The matrix of this deposit was almost pure ash with small quantities of charcoal. The artifactual recovery (N=50) included hand painted monochrome porcelain (3), hand painted polychrome




Figure 4. Test Unit #3 of the Nine Gal Tavern Site.






Figure 5. Test Unit #4 of the Nine Gal Tavern Site.

Figure 6. Feature #1 of the Nine Gal Tavern Site.

porcelain (1), bottle glass (1), flat glass (2), cut glass (1), pressed glass (2 objects, one of which was a fragment of a paneled tumbler), metal (12 items, notable of which were a hand wrought sprig, machine cut nails, a small spring, and part of a small pair of scissors), and early brick (2). Earthenwares were represented by redware (5), undecorated whiteware (6), hand painted polychrome whiteware (4), brown transfer printed whiteware (2), red transfer printed whiteware (6), and edge decorated whiteware (one item with a scalloped rim and impressed curved lines and one item with a scalloped rim and impressed bud designs). The mean ceramic date was 1844. The feature is interpreted as an ash pit associated with the fireplace on the east side of the Bryan structure. The deposition is from the Bryan occupation (1833-1853).

Feature #3 was initially located by a shovel probe at N9,E3.3. A 30 cm wide trench was extended north-south past the feature boundaries (Figure 7). The area of disturbance had a maximum depth of 64 cm below surface. Artifacts recovered (N=169) included Albany slip stoneware (2), luster porcelain (2), bottle glass (4 items, including two of an embossed peppermint bottle with a rough pontil scar), flat glass (16 objects, of which eight were of a mirror), pressed glass (7 sherds, with six from tumblers), metal (36 items, notable of which were a hand wrought nail, machine cut nails, a hollow iron key, a case knife, a bone handled clasp knife, and a hand wrought chain link), and structural materials (11 objects, including five early brick and two late brick). The earthenwares were represented by redware (3), undecorated pearlware (2), edge decorated pearlware with embossed patterns (2), undecorated whiteware (14), annular whiteware (3), hand painted blue whiteware (1), hand painted polychrome whiteware (3), blue transfer printed whiteware (1), black transfer printed whiteware (3), brown transfer printed whiteware (4), green transfer printed whiteware (9), red transfer printed whiteware (39), polychrome transfer printed whiteware (5), and edge decorated whiteware with a scalloped rim and impressed curved lines (4). The mean ceramic date was 1842. Feature #3 is best interpreted as an ash pit; although, the thick lens of in situ fired clay remains anomalous. The deposition is associated with the Bryan occupation (1833-1853).

Feature #4 was only investigated by three shovel probes. It was located at S12,E18 and displayed a significant concentration of artifactual material. The depth of the deposit was 40 cm below surface. The recovery included redware, stoneware, bottle glass, flat glass, pressed glass, metal, both early and late brick, and the heel of a ladies leather shoe. All of the remaining ceramic was whiteware (N=68): undecorated (38%), annular (3%), blue hand painted (3%), spatter (1.5%), blue transfer printed (6%), purple transfer printed (1.5%), green transfer printed (3%), red transfer printed (1.5%), mid-blue transfer printed (1.5%), flow blue (1.5%), and indeterminant (40%). The mean ceramic date was 1857. If the indeterminant whiteware is removed from this calculation, the mean




Figure 7. Feature #3 of the Nine Gal Tavern Site.


ceramic date is 1855. The bottle glass was represented by 37 items. Most of this was indeterminant except for color; although, there were two rough pontil bases. The pressed glass included five sherds, three of which were from paneled tumblers. The feature is interpreted as a trash pile associated with the Davidson occupation (1853-1856) and the operation of the Ohio Tavern.

Feature #5 was also only explored by three shovel probes. It was identified at S22,E12 and like Feature #4 displayed a significant concentration of artifactual material. The maximum depth was 35 cm below surface. The inventory included salt glazed stoneware, plain white porcelain, bottle glass, flat glass, machine cut nails, parts of two barrel hoops, and both early and late brick. All of the remaining ceramic was whiteware (N=174): undecorated (94%), spatter (1%), plain embossed (1%), and indeterminant (4%). The mean ceramic date for this assemblage was 1860 (this value holds even if the indeterminant category is deleted). The bottle glass included 32 sherds of which 30 were of whiskey bottles. No pressed tumblers were recovered. Feature #5 is interpreted as a trash pile associated with the unknown proprietor of the Nine Gal Tavern (1857-1859).

2. Artifacts



Operational definitions of the various artifacts and their associated technologies can be found in An Archeological Guide to the Historic Artifacts of the Upper Sangamon Basin (Stelle 1989). The present discussion provides a summary description of materials relevant to an interpretation of the site.

A. Earthen Wares

Earthenware was one of the most commonly recovered debris categories for this site (N=809). Four types were recognized: redware, yellowware, pearlware, and whiteware.

Redware

Redware was a common recovery item (N=55), representing 6.8 percent of the earthenwares. Table 1 indicates the distribution of redware by surface treatment. The most common surface treatment was a lead glaze, of which 20% was glazed only on the interior and 80% on both surfaces. The salt glazed vessel had an unglazed interior. The two vessels with an Albany slip were so treated on both surfaces. The base of a large mixing bowl had a white slip interior and a clear glaze exterior. Much of the redware was recovered from Features #1 (N=10) and #4 (N=13). Redware was produced in the United States throughout the nineteenth century although its recovery may be more common on sites predating the Civil War.

Table 1. Redware Recovered From the Nine Gal Tavern Site.
Surface Treatment
 Frequency
 Percent
Unglazed
 3
 5.5
Lead Glaze
 15
 27.3
Salt Glaze
 1
 1.8
Albany slip
 2
 3.6
White slip interior
clear glaze exterior
 1
 1.8
Indeterminant
 33
 60
TOTAL
 55
 100

Yellowware

Three forms of surface decoration were identified: clear glaze, Rockingham, and mocha. There was no annular decorated yellowware other than the sherd of mocha. The mocha body sherd was the familiar dark brown. The Rockingham included a body sherd and a molded basal element. Table 2 provides the frequency of each form. Representing only 2.3 percent of the earthenwares, yellowware is an uncommon ceramic from this site. Because of its long period of production, yellowware is not a particularly sensitive temporal indicator, but could have been present from the beginning of the site's occupation.

Pearlware

Pearlware is the earliest table ware horizon for the site. Twenty four sherds, representing 3 percent of the earthenware, were recovered (see Tables 3 and 4). Objects with no obvious surface decoration were classified as undecorated. It is the most common recovery category and included elements of three cups.

Table 4 presents the data on shell edge decorated ware. The pearlware component included five rim elements: two with scalloped rims and impressed curved lines and three with embossed patterns.

Table 2. Yellowware Recovered From The Nine Gal Tavern Site.
Type
 Frequency(%)
 Production Range
Clear glaze
 13(68.4)
 1830 to present
Rockingham
 2(10.5)
 1840 to 1900
Mocha
 1(5.3)
 1840-1900
Indeterminant
 3(15.8)
TOTAL
19(100.0)




Table 3. Pearlwares Recovered From The Nine Gal Tavern Site (see Table 4 for shell edge treatments).
Type
 Frequency(%)
 Median
Production Date
Undecorated
 17 (89.5)
 1805
Blue hand painted
 1 (5.3)
 1800
Polychrome hand painted
 1 (5.3)
 1800
TOTAL 19 (100)


Whiteware:

Whiteware is by far the most frequently recovered ceramic from the site (N=711) and represents 87.9 percent of the earthenwares (Tables 4 and 5). This category also includes the several variations collectively referred to as ironstone.

The stylistic elements of the annular decoration included bands of blue, white, brown, and green, as well as a roulette pattern. The spongeware is virtually all blue. Hand painted polychromes are of the sprig variety.

The transfer printed wares (N=181) defined 26% of the total whiteware sample. Blue, red, and flow blue were the most common treatments (Table 5).

Table 4. Edge Decorated Wares From The Nine Gal Tavern Site.
Type
 Frequency(%)
 Median
Production Date
Pearlware (N=5)
Scalloped rim

Impressed curved lines
 2 (13.3)
 1817
Embossed patterns
 3 (20.0)
 1829
Whiteware (N=10)
Scalloped rim


Impressed curved lines
5 (33.3)
 1817
Impressed straight lines
 2 (13.3)
 1820
Impressed bud
 1 (6.7)
 1823
Unscalloped rim
Unmolded
 2 (13.3)
 1879
TOTAL 15 (99.9)

The shell edge treatments (Table 4), both pearlware and whiteware, produced a mean ceramic date value of 1828. The remaining pearlware and whiteware had a mean ceramic date of 1853.

The marker's marks found include: "Venables", "La Belle China", "W..."[W. Adams or Woods?] , and "Pasto[ral?]...".

Table 5. Whitewares Recovered From The Nine Gal Tavern Site.
Type
 Frequency(%)
 Median
Production Date
Undecorated, Plain
 270 (38.5)
 1860
Annular
 10 (1.4)
 1850
Blue hand painted
 13 (1.9)
 1850
Polychrome, hand painted
 26 (3.7)
 1840
Sponge/spatter ware
 16 (2.3)
 1850
Transfer printed:


Blue
 29 (4.1)
 1845
Black
 5 (0.7)
 1840
Brown
 8 (1.1)
 1840
Purple
 11 (1.6)
 1845
Green
 19 (2.7)
 1840
Red
 66 (9.4)
 1840
Mid-blue
 16 (2.3)
 1829
Polychrome
 6 (0.9)
 1850
Flow Blue
 21 (3.0)
 1850
Plain, embossed
 3 (0.4)
 1860
Decalcomania
 3 (0.4)
 1930
Luster
 2 (0.3)
 1870
Indeterminant
 177 (25.3)
 1860
TOTAL 701 (100.0)

B. Stoneware

Stoneware is a ceramic fired at higher temperatures than earthenware. It is associated with relatively thick, heavy vessel forms like jugs and crocks. The variations in surface treatment recognized in this study include the following types: unglazed, salt glaze, Albany slip, Bristol, and American blue and gray. The distribution of stonewares from the Nine Gal Site is displayed in Table 6. Salt glazed forms and Albany slip were most commonly recovered. Stoneware types were produced over long periods of time and are relatively insensitive horizon markers for the archeologist.

Table 6. Stoneware Recovered From The Nine Gal Tavern Site.
Type
 Frequency
 Percent
Unglazed
 8
 7.7
Salt glazed exterior
Unglazed interior
 8
 7.7
Albany slip interior
 13
 12.5
Albany slip, both sides
 35
 33.7
Bristol
 1
 1.0
Blue and gray
 3
 2.9
Indeterminant
 18
 17.3
TOTAL
 104
 100.1

C. Porcelain

The porcelain recovery numbered 23 objects, including a leg and a foot element from a doll. Table 7 delineates the distribution of table wares. The majority (67%) of this material consisted of undecorated white body sherds. Surface decorations included hand painted monochromes (14%), hand painted polychromes (5%), over-glaze luster (10%), and decals (5%). These surface treatments are not considered temporally sensitive.

D. Clay pipes

Three stem elements and a bowl fragment were identified in the recovery. Manufacturers could not be identified.

E. Bottle glass

No intact bottles were recovered. Fragments were classified according to the construction characteristics they displayed. Broadly grouped, the fabrication characteristics included formation process, finish, glass color and surface texture, and embossing and labeling.

Table 7. Porcelain Table Wares Recovered From the Nine Gal Tavern Site.
Surface Treatment
 Frequency
 Percent
Plain white
 14
 67
Hand painted monochrome
 3
 14
Hand painted polychrome
 1
 5
Luster
 2
 10
Decal
 1
 5
TOTAL 21
 101

A total of 238 bottle glass fragments were recovered. A frequency distribution of the identified characteristics are displayed in Table 8. Each sherd was classified with regard to its most temporally sensitive characteristic. This approach was facilitated by the fact that 136 objects could be classified by color alone, 99 more by only one characteristic other than color (always aqua or clear), and only three by multiple characteristics other than clear or aqua color. Of these last three, two sherds were from a small bottle embossed with "..permint" (peppermint), showing a rough pontil scar, and cast in a two piece mold. The remaining sherd displayed a rough pontil scar and was cast in a two piece mold.

Portions of nine scroll or violin flasks were identified. These flasks were popular from the 1830s through the 1850s (McKearin and Wilson 1978:423). Embossed motifs included both stars and raised rib swirls.

Embossing on vessels other than the aforementioned peppermint bottle and scroll flasks included the letters E, NE, R, ROC/CHE, ST, and DAVIS.

Mason jars were surprisingly uncommon in the recovery. Elements of only four vessels were identified. They are frequently encountered on farmsteads post-dating 1858.

F. Pressed glass

A total of 88 pressed glass objects were identified. Tumblers accounted for 73% (N=64) of the sample. Treatments included

Table 8. Frequency Distribution of Construction Characteristics Identified on Bottle Glass Fragments From the Nine Gal Tavern Site.
Characteristic
 Frequency
 Production Range*
Molds:

Three piece
 1
 1810-1880
Two piece
 7
 1845-1890
Turn/Paste
 1
 1880-1905
Automatic bottling machine
 7
 1903-present
Rough pontil
 7
 1810-1860
Snap case
 12
 1855-1903
3/4 round base
 1
 1900-1903
Laid on lip ring
 1
 to 1845
Applied lip ring
 1
 1850 - 1870
Glass color black or opaque
 2
 1815 - 1875
Glass color purple
 3
 1880 - 1918
Glass color clear or aqua**
 131
 to present
Pre-chilled iron mold
 2
 to 1870
Slug plate
 53
 1850 - 1905
Embossing
 11
 1867 - 1905
TOTAL 238
* Adapted from Deiss (1981:92-96) and Newman (1970).
** Color was the only observable characteristic on the fragment.

swirled patterns (9%), paneled or fluted patterns (23%), and plain (67%). Fluted tumblers were both hexagonal and octagonal. The other pressed table wares (N=24) included elements of sugar bowls, creamers, jars, plates, and bowls. Design features included ribbing, ray patterns, flower patterns, acid etching, and a single element of cut glass. The vast majority of this material was clear; although, blue, lavender, green, and amber were encountered.

Milk glass was represented by eight buttons (both two and four hole), canning jar seals (N=5), and such incidentals as an "Old Spice" shaving lotion bottle fragment, a marble, and a bead.

Other glass objects in the recovery included two lamp chimney sherds, two light bulb fragments, an opaque green marble, the base to a mercury thermometer, and the lens from a carriage lamp. G. Flat glass and melted glass

Flat or window glass included 1335 sherds with a total mass of 2221.5 grams. The majority had a greenish hue. Eight mirror fragments (8.2 g) were identified. Twenty eight objects (44.3 g) of melted glass were also recovered.

H. Metal

As might be expected, a large number of metallic objects were recovered (N=1302). Of these objects 70.4% were nails. Table 9 presents the distribution of nails by type categories defined by Nelson (1968). The modal category is of early machine headed cut nails. With a production range of 1818 to 1840, they correspond to the initial construction of the Bryan house. The measurable presence of modern machine cut and wire nails leaves open the possibility of later additions or remodeling of the original structure, as well as other building projects. Other fasteners included horseshoe nails, wood screws, staples, machine bolts, nuts, concrete anchors, and a pin.

Cutlery included a two-tined fork, a case knife with a pointed blade and a flat tang, the bowl of an ovid shaped table spoon, and three handle elements from either spoons or forks, one with a bone grip still attached. Two clasp knives were recovered. Both knives were four bladed. One displayed bone grips.

Brass items included eleven .22 caliber rim fire casings (maker's marks "H", "U-HighSpeed", and "Super X"), two .32 caliber rim fire cartridge casing, the base of two Winchester 12 gauge shot shells, elements of a pocket watch, elements of a clock, strap/banding material, and a brass covered button. An unfired lead musket ball was also identified.

Three coins were recovered: a 1920 Wheat penny, a 1914 Liberty dime, and a 1936 Buffalo nickel.

Domestic items included small elements from cast iron cooking vessels, small stove elements, short sections of twelve barrel or cask hoops, a wrought iron pot hook, rim elements of a bucket, elements of tin cans, four jar lids, bottle caps, a lamp bracket, a collar for a #3 kerosene lamp, cast iron hinges, a Rockingham door knob, two hand-worked exterior door plates (interesting because they could be from the Bryan structure), cabinet hardware, a furniture caster, a pull from a blanket chest, a hollow iron key, a curtain ring, storm window hardware, and links of chain.

Tool or tool elements recovered consisted of a well worn plastering trowel or float, a single bit axe head, an alligator wrench, a pair of pliers, an iron file, a screw driver, part of a pair of scissors, a pivot cap, two end caps for wooden handled tools, miscellaneous farm machinery parts, and a valve from an

Table 9. Nail Types Recovered From the Nine Gal Tavern Site (after Nelson [1968]).
Nail Type
 Frequency
 Percent
 Production Range
Hand wrought
 6
 0.7
 19th century
Early machine cut
hand made head
 4
 0.4
 1780-1820
Completely machine cut
sprigs and brads
 7
 0.8
 1805 - present
Early machine headed
cut nails
 400
 43.6
 1818 - 1840
Modern machine cut nails
 256
 27.9
 1840 - present
Modern wire nails
 183
 20.0
 1850 - present
Unidentified
 61
 6.7
TOTAL
917
 100.1

internal combustion engine. Harness buckles and clips, as well as a horse shoe were also identified.

Personal items consisted of a toy cap pistol, a small element from a toy steam locomotive, belt buckles, clothing clips, a boot eyelet, and seven metal overall buttons or studs (one with the embossed word "TEST").

Several short lengths of wire were identified. Both iron wire and insulated copper wire are represented. Part of an electrical box was also recognized.

I. Structural materials

The structural materials category is represented by 234 items. It includes brick (83.3%), stone (0.4%), mortar (9.4%), plaster (6.0%), and chalking compound (0.9%).

The bricks were divided into two categories: early and late. The early brick was hand struck in a five sided mold, bright orange in color, soft (it could be scratched with a finger nail), lacked temper (although there are occasional inclusions of varying size), contained irregularly sized voids, and sometimes displayed glazed surfaces. The late brick was formed in a press mold, red in color, hard, lacked voids, and displayed tempering material of regular dimension.

Eighty three percent of the brick recovery was of the early type, 13.3% was of the late variety, and 3.6% was indeterminant. The early type is associated with the structure prior to 1860. The frequency with which they were encountered lends support to the historical interpretations which describe masonry fireplaces at opposite gables of the Bryan structure. "Clamps" may yet be located on the site. The late brick became the object of local industry after 1860. Much of its recovery was related to the foundation and cellar applications in the existing 1890s house.

Also recovered were two pieces of three-in-one tab asphalt shingles, a glob of roofing cement, and a fragment of a red clay field tile.

J. Other

Nine plastic objects were identified. They included a plastic flower, a bottle brush with plastic bristles, and a small pencil sharpener with an iron blade.

Artifacts rendered from mineral included six pieces of slate from a child's slate-board, a piece of chalk, a wooden pencil head, a fragment from a modern grinding wheel, a cream-colored clay marble, and 37 small lumps of coal.

Personal items included three shell buttons (two-hole), a two-hole bone button, two shoe heels, a piece of shoe leather, a ladies leather dress glove, a piece of cloth, and a hair brush.

Discussion and Implications

The Nine Gal Tavern locality is a complex series of archeological events spanning several thousand years of occupation. It is, however, the cultural phenomenon of the tavern that is of central interest to the present study. The tavern is here considered an index of community development. The theoretic interests driving the present study frame a model of frontier community development common to the study area. As a time transient factor of economic differentiation, the tavern's expressions serve as horizon markers of Euro-American social and economic diversification. The form of the tavern changed over time in response to a changing cultural milieu and the redefined needs it placed upon the bio-physical matrix. Of importance to this work is the prospect that the varying tavern forms have archeological identities observable from their attendant techno-material assemblages. Historical archeology is important to anthropology because by adding historical evidence to the archeological record we are better able to bridge the relationships between material culture and the enveloping cultural system. The tavern presents an opportunity to observe and measure just such relationships.

Archeological studies of taverns from within the region are quite limited. From the lower Sangamon valley we find some work at the Lindsey Tavern in New Salem (Petersburg) and Broadwell's Inn near Pleasant Plains. Current research from Illinois includes Phillippe's (1987) work at Hutsonville with the Cox Tavern and Wagner's (1988) work with the "Old Landmark" Tavern in Marion County. The more recent research is as yet mostly unpublished. Work from other regions demonstrates that taverns have distinct archeological expressions relative to other cultural forms. The material assemblage of the tavern can be differentiated on the basis of such variables as region and an urban-rural dichotomy (Rockman and Rothschild 1984). If synchronic variation exists then so also might diachronic variation. As one begins to investigate the tavern phenomenon one quickly discovers that it varies widely in its form and that a particular site may display different forms over time. The common appellation tavern would seem too generic and quite likely subsumes several distinct types. The purpose of this section of the study is to attempt the isolation of some of these types on the basis of historic reconstruction and archeological investigation.

The tavern is a commercial phenomenon signifying particular levels of structural complexity and economic development in a frontier community. Its evolving expression as a cultural form can be traced along several dimensions: population density, transportation, demand, exchange, regulation, community environment, and material form. The proposed model posits three types or levels of taverns: Incidental Tavern, Incipient Tavern, and Full Tavern. The model will be expiated in the milieu of Middletown Township. Whether these conditions and circumstances obtain to other communities remains an empirical question unanswered by the present study.

Pre-Tavern

Initially, small numbers of travelers moving over varied trails simply bivouacked when the day was done or when they could journey no further. With population densities less than one-half person per square mile for county sized regions, residents were so thinly scattered that it was not possible for the traveler to have nightly contact with habitants. Services, supplies, and civilization were widely dispersed. The traveler was forced to be self-sufficient for extended periods of time.

Tavern Type I: Incidental Tavern

The first stage, Incidental Tavern, in the development of public accommodations occurred when there was an increased number of travelers, but only private cabin sites on the landscape. Within these private residences social custom and frontier hospitality served the needs of the traveler. In a romanticized county history of 1878 (Brink, McDonough and Co. 1878:20) the author observed:

Gone is that free-hearted hospitality which made of every settler's cabin an inn where the belated and weary traveler found entertainment without money and without price.

However, even at this level, the model interprets the offering of public accommodations as an inherently economic act. Moreover, only some of the farmsteads would evidence behavioral patterns warranting the reference "tavern." In Middletown Township this stage was encountered between 1833 and 1836.

1. Population density - Population densities of less than two people per square mile for township sized regions would have supported this stage. Platted villages were separated by more than a day's travel time. Walls (1989 personal communication) considers this to be extremely low and typical of the early pioneer period. Champaign County was formed in 1833. The Bryans homesteaded in 1834 and were the third family to do so in Middletown Township.

2. Transportation system - The transportation system consisted primarily of trails with few wagonable roads. However, it expanded exponentially during this stage as links to homesteads were carved across the landscape. Primary arteries like the Ft. Clark Trail were simply unimproved wagon roads. Poorly maintained, the main line of the trail moved laterally and sinuously during wet weather in response to the notorious "chug holes."

During this time, links were also created between the emerging village nuclei. In 1835, the Champaign County Commissioners approved the survey of a road between the Sangamon ford at Middletown and the Urbana-Decatur road passing through Centerville. The following year it was built; although, this may be too strong a term in that much of its course was merely blazed. Improvements in the primary transportation routes also occurred. "In 1835 the road from the Sangamon to Urbana, traveled by the early settlers, was of so circuitous a character, that Mr. Scott, who was compelled to travel it often, concluded to straighten it; and accordingly took his horses and plow, and drew a furrow from the Sangamon to Urbana, a distance of twelve miles, and by this direct line a road was made...."(Lothrop 1870:396).

River crossings were effected at natural fords. In some places improvements may have been made in the fords. If this reach of the Sangamon River system itself was ever an important element in the transportation system of east-central Illinois, it had ceased to be so by this stage. Virtually all travel was overland.

To a considerable extent the relationship of the homestead to the transportation network determined the likelihood of visitors.

3. Demand - Economic demand was greatest at those homes located near the intersection of trails, river fords, or at daily conveyance distances along lines of travel. At these loci, sojourners had an interest in warm food, sleeping indoors, sociability, information, supplies, and equipment repair. To the extent that demand remained low and sporadic, travelers were simply fed and put up for the night.

4. Exchange - A reciprocal form of exchange generally obtained. It involved both social and material transfers. The social component could display varied elements. For instance, conversation and interaction with outsiders was both intrinsically rewarding and prestige granting. Secondly, the possessor of information from the world outside had the power to define and interpret that world to the local community. Thirdly, the distinction as the family with the building large enough or attractive enough to be chosen by the traveler provided an external source of social validation for the family's position within the community. Lastly, the demonstration of food surpluses adequate for use in economic exchange with the traveler affirmed prestige within the community.

The transfer of goods could also take unusual forms. For instance, from a personal history authored by Stephen Conger Abbott (Abbott 1902:5) who described his journey from Peoria to Middletown in July of 1847, "Always staid (sic) at farm houses and paid bills with pictures and books." He had earlier purchased ".. $5.00 worth of novels and charts...." with which to compensate home owners.

5. Regulation - Attempts at regulation rarely extended beyond the weight of public opinion. Negative sanctions were informal and typically diffuse. However, in this stage unsanctioned groups of local citizens, vigilantes or "regulators", may have tried to control those individual proprietors who attempted to immorally exploit (theft, assault, and murder) the traveler (Wagner 1988).

6. Environing community - This level of tavern development demanded a local population capable of responding to the social transfers and rewards of the exchange system. We might best think of these communities as proto-communities, in that they were geographically diffuse. In the present context the community was locally labeled the "Sangamon Timber Settlement." The physical extent was from near the contemporary village of Fisher to the Piatt County line, a linear distance of twenty miles. It signified the region of early homesteading in western Champaign County. But in its definition, the geographic boundaries were in some ways less important than cultural boundaries. The residents evidenced homogeneity in their social attitudes, economic activity, religious orientation, and political interests (Walls 1989).

Due to a homestead's central location, size, and a willing attitude of the owner, some sites functioned as meeting places for social, religious, and political activities. For instance, in 1835 the Bryan home was designated the polling place for all of western Champaign County (Champaign County Commissioners' Record Book A). Homesteads meeting these requirements likely also occasionally functioned as places where the traveler could find lodging and refreshment. Geography, family wealth, and an attitudinal component of the personality determined the potential of such a site's function as an Incidental Tavern - a place for strangers to eat and spend the night. Such sites many times formed the nucleus around which villages later formed (Buley 1950:481).

Economic differentiation was minimal. From 1833 to 1836 it was likely that all habitants of Middletown Township were engaged in agriculture.

7. Material form - The material assemblage would be differentiated from that of a single family farmstead only in its relative wealth, complexity, and magnitude. The residential structure would be substantial by local standards. While early homesites were hewn of logs in this region, by the early 1830s sawed lumber was available and used with increasing frequency. The ceramic assemblage would center on expensive wares and might evidence considerable variation. While glass would be reasonably common, its incidence would be tempered by technological considerations and availability. One would expect to find both expensive and innovative forms (for instance, flint glass, cut glass, or early pressed glass). Liquor glass, both bottles and service, would be incidental to the assemblage.

Discussion: The Bryan occupation of the Nine Gal Tavern Site displays many of the characteristics of the Incidental Tavern. The family was the third such to homestead in the township. They were agriculturalists and never applied for a license to keep tavern. The large home they built was positioned near the intersection of the two major trails transecting the township and near the most important river ford. The homestead was located approximately a days journey west of the county seat and on the east side of the Sangamon. When the river was at flood, west bound travelers may have been stopped for more than a week waiting for the waters to recede. There were no alternative accommodations for travelers to access.

By local standards the Bryans were wealthy. The artifactual materials recovered from Features #1, #2, and #3 attest to this. Particularly conspicuous were both the variety and value of the ceramics. Fifteen different types of surface decoration were identified. Following McCorvie (1987:272-274) and others, an average ceramic value was calculated (Table 10). For these three features it was 1.99. McCorvie (1987:273) considers values above 1.5 indicative of wealthy families. The glass recovery also suggests relative wealth. The vast majority was pressed glass. Unfortunately much of it is in the form of pressed tumblers, which would seem to confuse the issue of liquor service. A last observation regarding glass is that from these features were identified two objects unique to the entire recovery: a sherd of light blue milk glass frosted with fluoride and a specimen of cut glass.

That the Bryan's "kept tavern" at all was not directly discernable from the archeological record. Support for this contention comes from the assertion in the earliest county history (Brink, McDonough and Co. 1878:125) that "...John Bryant (sic) (kept) the first tavern...." The demand for their services would have been greatest between 1833 and 1836. As the social and economic infrastructure of the community continued to develop, demand may have lessened and become increasingly episodic and situational (i.e. in response to a flooded Sangamon River), but in the early years of the township they were apparently willing to put up travelers for the night.

Tavern Type II: Incipient Tavern

The second stage, Incipient Tavern, occurred when increased numbers of travelers created the prospect of commercial activity. Morgan (1969:18) cites the case in Big Grove (Champaign-Urbana) when a host recalled that his cabin, eighteen feet square, furnished cozy accommodations for forty-nine guests one night. He admonished the reader to recall that "It must be remembered that people were smaller in those days...."(Morgan 1969:18). The proprietor of an Incipient Tavern "kept tavern" as a secondary economic activity. For farmers and retailers it could have been an additional source of cash. The women of the household would likely play a critical role in its management. In fact, a wife's employment in this fashion may have provided the major source of hard cash for the family. In Middletown Township this stage was encountered between 1836 and 1849.

Table 10. Mean Ceramic Value of Features #1, #2, and #3.
Type
 Frequency
 1855 Value
 (f) x (Value)
Undecorated
 26
 1.00
 26.00
Minimal
 14
 1.16
 16.24
Hand painted
 14
 1.3
 18.20
Transfer printed
 93
 2.5
 232.50
TOTAL
 147
292.94
Mean Ceramic Value =292.94/147=1.9927

1. Population density - Population densities for township sized areas would have been in the range of 2 to 5 people per square mile. The population density of Middletown Township in 1840 is estimated at 2.5 to 3 persons per square mile (Walls 1989:4).

2. Transportation - The transportation system continued to expand with wagon trails forging links to the increased number of homesteads. County and state government increased their response to the need for road improvements. In 1836, the Champaign County Commissioners authorized Isaac Busey and Jonathan Osborn to initiate a state road from Urbana to Bloomington. It followed the furrow plowed by Fielding Scott to the Sangamon ford the previous year (Purnell 1955:40). A differentiated political structure known in Champaign County as the Township Road Commissioner was created. Road taxes were legislated. The tax could be paid with labor rather than money. The Road Commissioner kept track of the number of hours of labor expended by each citizen.

River crossings included not only the improved fords, but also ferries. In 1836 the Champaign County Commissioners (Record Book A) began regulating ferries through licensing and fee prescriptions. Licensed ferry service at Middletown was initiated in 1836 (Champaign County Commissioners' Record Book A).

Public transportation saw its first clear expression in the form of the stage line. It was likely to have been extended through Middletown by 1846.

All of these changes in the transportation system impacted the flow and localization of travelers.

3. Demand - Although demand for public accommodations increased during this stage, it was still not adequate for the initiation of a fully differentiated business activity. While more customers were on the road, the demand remained intermittent in response to such events as flooded rivers and roads or seasonal as in the market created by migratory farm labor. In a developmental context, when the homesteader felt the need to begin charging for services because of increased demand, signage might have occurred with the informal and generic "keep public" (Buley 1950:481).

During this stage, when a competitive environment existed for a finite level of demand, attempts at market definition through the artifice of business names first emerged. While most of the Middletown Township taverns were known by the name of the proprietor, Sarah Rea chose to name hers "The American House." This strategy became more common during the Full Tavern stage.

4. Exchange - With increased demand, the drain on economic resources was too great for social transfers to provide sufficient reward for the local resident. Exchange became increasingly cash oriented. Signage functioned as a social device legitimating the charging of fees and the utilization of cash as the medium of exchange (Buley 1950:481). Larger local populations and increasing economic differentiation, as well as the need for income producing employment recast the circumstance of the traveler as a commercial opportunity. Exchange at this level followed the market principle; charges were such as the market would bear.

5. Regulation - County level government became involved in the regulation of services and pricing with a likely eye toward potential tax revenues and community image. By 1836 Champaign County had begun licensing taverns and regulating charges (Champaign County Commissioners' Record Book A). The licensing fee was $2.00 per year. The entry regarding the fee structure reads:

For keeping a man and a horse one night, including supper, bed and horse feed..$0.75
For a single meal.................$0.18 3/4
For single horse feed..............0.12 1/2
For one-half pint whiskey......... 0.06 1/4
For one-half pint French Brandy....0.18 3/4
For one-half pint Wine.............0.18 3/4
For one- half pint Gin..............0.12 1/2
For one-half pint Rum..............0.18 3/4
For one half pint Domestic brandy..0.18 3/4

Because transient populations presented a special moral problem for the community, moral entrepreneurs were interested in the affairs of the tavern. For instance, the possibility that the appellation "Nine Gal" derived from the availability of commercial sex (the proprietor reportedly had nine red-headed daughters) suggests the potential interest of the religious community in such establishments.

6. Environing community - The community which supported this stage placed increased emphasis on geo-political boundary definitions. Communities were platted and legal identities established. In Middletown Township commercial differentiation included such enterprises as ferries (1836), saw and grist mills (1837), blacksmiths, and retailing (1836) (Brink, McDonough and Co. 1878:125). In 1836 the village of Middletown was platted and the developer, Daniel Porter, opened a general store and tavern (both were licensed by the county [County Commissioners' Record Book A]). He also served as postmaster. Developers like Porter used these civil and commercial nuclei as magnets for residential populations, village growth, and long range profit from land sales.

7. Material form - The architectural expression that would characterize the Full Tavern surfaced at this stage. A large, two story frame structure was the style most frequently encountered in the study area. These buildings were rarely fabricated with the tavern function in mind; consequently, existing structures that displayed the appropriate design elements were exploited. Critical considerations included buildings that offered reasonably large public areas for food service and conviviality and multiple segregated areas for sleeping. The ceramic assemblage associated with the Incipient Tavern would be dominated by less expensive wares. There might also be an increase in the number of sets of dishes and durability would be an important consideration. Inexpensive glass, again conditioned by technological considerations and availability, would be more common than in the first stage. Liquor glass, both bottles for storage and sale and tumblers for service, would be common.

Discussion: The history of the Nine Gal Site provides no evidence of the Incipient Tavern stage. However, the form is seen in three other tavern sites known to the community: Porter's tavern (1836) in the platted village, the Mathew Johnson Tavern (by 1847), and the Rea Tavern or American House (1848). All three were ancillary economic activities for the proprietors: Porter was a developer and merchant undoubtedly using the tavern as a device for encouraging settlement and secondary to retailing, Johnson was a farmer whose residence was situated so as to make retailing and tavern keeping an opportunistic form of income, and Sarah Rea kept the American House Tavern in the family residence of the farmstead. What these particular sites would look like archeologically remains an unanswered empirical question.

Tavern Type III: Full Tavern

In the Full Tavern stage, the entrepreneur and his/her family occupied a relatively large structure and provided services on a cash basis. The tavern served as the primary occupational activity of the proprietor. In Middletown Township this stage was encountered between 1850 and 1859.

1. Population density - Population densities for a township sized region would be in the range of 5 to 8 people per square mile. Mahomet Township demonstrated a per square mile density of 6.9 persons in the 1850 census (Walls 1989:4). The presence of small villages was noticeable. Some of these were platted, others were not. These villages had residential populations of one to five families and were separated by less than a days travel time.

2. Transportation - The network connecting homesteads to villages and villages to villages underwent further expansion. The interest in the quality and maintenance of highway system continued unabated.

River crossings continued to include fords but the ferry was replaced with the bridge. The plans for the first bridge over the Sangamon at Middletown were launched in 1843 by the Champaign County Commissioners (Record Book A). However, the bridge was not completed until 1850 (Brink, McDonough and Co. 1878:125).

Public transportation still centered upon the stage coach line. In fact, the appearance of the railroad significantly reduced the road traffic upon which taverns were dependent. This reduction in demand created a competitive environment which few establishments were able to survive.

3. Demand - Public transportation in the form of the stage coach was significant to the emergence of this expression. As indicated previously, the stage stop guaranteed a regular and stable market of travelers seeking services. What changed relative to the Incipient Tavern was the volume of stage traffic and number of travelers. In Middletown Township it was this increased volume that enabled the Full Tavern. The increased size and economic complexity of the local community, as well as a sustained flow of travelers, facilitated the creation of multiple taverns. When this expansion of the industry occurred, some establishments worked with specialized markets like wagoner freight haulers, local workers, new arrivals, and stage coach travelers. Others concentrated their business on the local residents who were increasingly using the taverns as places of refreshment and libation. Taverns could now be stratified in terms of the quality and character of the service rendered (i.e., Abraham Lincoln is claimed by local historians to have stayed at the Ohio Tavern).

The targeting of markets is further evidenced by the use of names like "Ohio Tavern" and Dr. Adams' "Hotel". The Davidsons clearly attempted to market to the large local population of immigrants from Ohio and Dr. Adams was apparently interested in boarders and a more "citified" image (the "hotel" was situated in what had been Daniel Porter's tavern). Stephen Abbott (1902:9) indicated in his autobiography that in 1853 he moved "...to a hotel to board in Middletown kept by one Dr. Adams."

The idea of business image took on a new significance during this stage as proprietors attempted to separate their commercial enterprise from the family home. Of the four establishments known to have been operating in the Township during this period only the Mathew Johnson Tavern (still operating at the Incipient Tavern level) continued to employ the family home concept. The Ohio Tavern, Nine Gal Tavern, and Hotel were by their names alone, fully commercial and professional enterprises.

4. Exchange - Cash was the medium of exchange for this stage, and its production was the fundamental motive for the entrepreneurial activity. Proprietors of a Full Tavern relied upon the business as their primary source of income.

5. Regulation - Unfortunately, the Champaign County Commissioner's record book for this period has been irretrievably lost to water damage. Consequently, it is much more difficult to determine who was issued a license, if changes in fee structures were occurring, and how rigorously the requirements were being enforced. A likely scenario is that licenses were still required, but fee prescriptions for services other than lodging were largely abandoned, or, in the case of beverage alcohol, replaced by other forms of governmental regulation.

6. Environing community - The community which supported this tavern stage evidenced a significant increase in commercial diversification. The 1850 census listed 49 farmers and 92 farm laborers, a millwright, a grocer, two inn keepers, a physician, two merchants, four blacksmiths, three carpenters, and a laborer (Walls 1989:32). With the increased economic differentiation, an improved transportation system, an established system of public transportation, institutionalized religious groups (by 1858 there were three churches [Purnell 1955:42-45]), public school districts, and a greatly expanded residential population increasingly interested in leisure time consumptive practices, the frontier community was drawing to a close (Walls 1989). As indicated by the 1860 census, the range of occupational endeavors had increased from 10 to 28. The period from 1850 to 1860 witnessed exponential growth in the occupational infrastructure of the township.

7. Material form - The material assemblage of the Full Tavern would focus on a diversity of inexpensive, durable table wares (for instance, plain whiteware, ironstone, or hotel china), an increase in liquor glass, and an increase in pressed glass tumblers. Phillippe (1987) reports fluted, pressed glass tumblers as the most common glass category in the recovery from the Cox Tavern. One might also expect an increase in the frequency of chamber pots. The architectural style that emerged with the Incipient Tavern would continue. In Middletown Township, Full Taverns were located in buildings that had already functioned as taverns. A large, two story building with adequate area on the first floor for food preparation, serving areas, perhaps a bar, and small, private sleeping quarters on the second were typical. There were three Full Taverns operating within the Township during this period: the Ohio Tavern, the Nine Gal Tavern, and Dr. Adams' Hotel.

Discussion: The Nine Gal Site provides evidence of the Full Tavern stage. Both the Ohio Tavern and the Nine Gal Tavern fall into this category. From the exploration of the site, two discrete deposits were identified relating to these two occupations: Feature #4 to the Ohio Tavern and Feature #5 to the Nine Gal Tavern. Both features were dominated by plain whiteware. Table 11 contains the data bearing on the calculation of the mean ceramic value of Feature #4 and Table 12 of Feature #5. Note that there are significant differences in both the variety of ceramics (10 types for the Ohio to three types for the Nine Gal) and in their mean value (1.39 for the Ohio to 1.00 for the Nine Gal). A possible interpretation of these differences is that the Ohio Tavern, the one that Abraham Lincoln stayed at, was specializing in the higher status segment of the market. Bottle glass constitutes a much larger fraction of the total assemblage than was the case for the Bryan's Incidental Tavern. Additionally, there is also significant

Table 11. Mean Ceramic Value of Feature #4.
Type
 Frequency
 1855 Value
 (f) x (Value)
Undecorated
 26
 1.00
 26.00
Minimal
 3
 1.16
 3.48
Hand painted
 2
 1.3
 2.60
Transfer printed
 10
 2.5
 25.00
TOTAL
41
57.08
Mean Ceramic Value = 57.08/41=1.3921

Table 12. Mean Ceramic Value of Feature #5.
Type
 Frequency
 1855 Value
 (f) x (Value)
Undecorated
 95
 1.00
 95.00
Minimal
 1
 1.16
 1.16
Hand painted
 0
 1.3
 0.00
Transfer printed
 0
 2.5
 0.00
TOTAL
96
96.16
Mean Ceramic Value = 96.16/96=1.0016

variation between the Ohio Tavern component and the Nine Gal Tavern component with regard to the ratio of dishes to bottle glass. For the Ohio Tavern the ratio is 4.5:1 and for the Nine Gal Tavern the ratio is 2.3:1. Moreover, virtually all (94%) of the bottle glass from the Nine Gal Tavern feature was in the form of liquor bottles. The dispensing of liquor may have been a more important activity at the Nine Gal Tavern. Curiously, no pressed glass tumblers were identified from the Nine Gal feature and only five sherds were identified from the Ohio Tavern feature. Chamber pots were not recovered from either feature. In conclusion, and fully sensitive to the limits imposed by the size of the samples, the data suggests that the Ohio Tavern and the Nine Gal Tavern were directed at specialized segments of the local market.

Post-Tavern

The end of the tavern form came quickly in Middletown. The occupational categories of the 1860 census include no entry of inn keeper for the Township. In the post-tavern phase, the appellation "tavern" would have fallen into disuse for describing places of public accommodation. In contemporary usage it was now informally applied to rural eating and drinking establishments. However, with sustained community growth, enhanced transportation systems, and increased demand, the modern economic forms of the tourist camp/court, boarding house, motel, and hotel became possible.

Conclusion

Historic research on Middletown Township during the period of 1833 to 1860 has suggested the possibility of three different types of taverns: Incidental, Incipient, and Full. Archeological investigation of the Nine Gal Site has provided preliminary support for the first and last of these types. John Bryan certainly was in a position, both geographically and socially, to have operated an Incidental Tavern between 1833 and 1836. The historical evidence indicates that he did so. Later, between 1853 and 1859, the structure he built was employed by at least two Full Taverns, the Ohio Tavern and the Nine Gal Tavern. These last two businesses can be differentiated on the basis of the services they provided.

The primary limitations of the model and the research are twofold. First, the expression of the tavern industry is a complex event. In simplest form, a single site may display the linear, stage specific, evolutionary changes framed by the model. As a cultural form, the site's varying functional expressions will proceed or falter in response to local circumstances. Consequently, the entire trajectory may or may not be present at the site level. For instance, in the present study the Incipient Tavern stage was not identified and probably never occurred at this site. Larger geographic and cultural units must be explored in order to observe all three stages. The geo-political unit displaying the greatest potential for this is the township.

The other critical limitation of the model focuses on the archeological expression of the different stages. Because no comparative data exists, it is not clear that the Incidental Tavern can be distinguished from other area homesteads that display similar wealth. Furthermore, having not effectively examined an Incipient Tavern, it is possible that Incipient Taverns may not be archeologically discernable from Full Taverns. What is known on the basis of the current data set is that Incidental Taverns are different from Full Taverns and that Full Taverns have the potential of being differentiated on the basis of market specialization. These problems certainly suggest a direction for future research.

The model, at best, presents an exploratory framework within which the tavern phenomenon can be observed. At minimum it may have useful descriptive qualities. Its validity is, of course, an empirical question.

REFERENCES CITED

Abbott, Stephen C.
1902 The Autobiography of Stephen Conger Abbott. Unpublished manuscript on file at the Champaign County Early American Museum.
1985 1,000 Facts About Mahomet Published in the 'Mahomet Sucker State' Beginning Issue June 22, 1944. Manuscript on file Urbana Free Library. Urbana, Illinois.

The Breeder's Gazette
1911 May 31 no. 22. Methods of Successful Business Farming. Chicago, Illinois.

Brink, McDonough and Company.
1878 History of Champaign County, Illinois. Philadelphia.

Buley, R. Carlyle
1950 The Old Northwest Pioneer Period 1815-1840, Vol. 1. Indiana University Press. Bloomington, Indiana.

Champaign County Gazette. Obituaries. 4 Feb 1880.

Champaign County, Illinois.
N. D. Champaign County Commissioners' Record Book A. Manuscript on file in the Archives of the Urbana Free Library. Urbana, Illinois.
1979 Recorder of Deeds. Deed Index: Grantor 1833-1927. Vol. A, B1. Manuscript on file in the Archives of the Urbana Free Library. Urbana, Illinois.

Cunningham, Joseph O.
1905 Historical Encyclopedia of Illinois and History of Champaign County. 2 Vols. Munsell Publishing Co.. Chicago, Illinois.

Deiss, Ronald W.
1981 The Development And Application Of A Chronology For American Glass. M.S. Thesis. Illinois State University. Normal, Illinois.

Lothrop, J. S.
1870 Champaign County Directory 1870 - 1871.

Mathews, Milton W. and Lewis A. McLean.
1979 Early History and Pioneers of Champaign County. Unigraphic, Inc. Evansville, Indiana.

McCorvie, Mary R.
1987 The Davis, Baldridge, And Huggins Sites: Three Nineteenth Century Upland South Farmsteads In Perry County, Illinois. Preservation Series 4. American Resources Group, Ltd. Carbondale, Illinois.

McKearin, Helen and Kenneth M. Wilson
1978 American Bottles and Flasks. Crown Publishers, Inc. New York.

Morgan, Richard L.
1969 Cornsilk and Chaff of Champaign County. Sesquicentennial Committee of Champaign County. Champaign, Il.

Nelson, Lee H.
1968 Nail Chronology as an Aid to Dating Old Buildings. American Association for State and Local History, Technical Leaflet No.48. History News 24(11).

Newman, T. Stell
1970 A Dating Key For Post-Eighteenth Century Bottles. Historical Archaeology. 70-75.

Pease, Thomas Calvin.
1925 The Story of Illinois. A. C. McClurg and Co.

Phillippe, Joseph Sidney
1987 Hutsonville on the Wabash: Excavation of a Nineteenth Century River Town. In Donald B. Ball and Philip J. DiBlasi. Proceedings Of The Symposium On Ohio Valley Urban And Historic Archaeology, Vol. V. Department of Anthropology, University of Louisville. Louisville, Kentucky.

Purnell, Isabelle S.
1955 History of Mahomet, Mahomet Methodist Church Centennial. Mahomet Methodist Church. Mahomet, Illinois.

Rockman, Diana D. and Nan A. Rothschild
1984 City Tavern, Country Tavern: An Analysis of Four Colonial Sites. Historical Archaeology 18(2):112-121.

Roehm, Frances.
1986 Champaign County, Illinois, 1850, A Historical Overview. Urbana Free Library. Urbana, Il.

State of Illinois, Archives Division.
1982 Public Domain Sales, Land Tract Record Listing. (Champaign County), Vol.2. Springfield, Illinois.

Stelle, Lenville J.
1986 Cultural Resources Survey: An Environmental Summary Of The Lake Of The Woods. Sangama Archaeology. Mahomet, Illinois.
1989 An Archeological Guide to the Historical Artifacts of the Upper Sangamon Valley. Sangama Archaeology. Mahomet, Illinois.

Wagner, Mark J.
1988 The Old Landmark Tavern. Paper presented at the 1988 annual meeting of the Illinois Archaeological Survey. Lewiston, Illinois.

Walls, Gina Denise
1989 Demographic Reconstruction: Mahomet Township 1850, 1860 and 1880. Manuscript on file Department of Social Sciences, Parkland College. Champaign, Illinois.




Please forward any thoughts or comments to: lstelle@virtual.parkland.edu