Colonial Williamsburg Digital Library

The Climate

Weather on "the Peninsula," as the area between the York and James Rivers is known locally, is characterized by relatively mild winters and warm summers, but not without extremes on both ends of the thermometer. John Smith accurately described the area's climate in 1624:

The Sommer is hot as in Spaine; the Winter cold as in France or England. The heat of sommer is in June, July, and August, but commonly the coole breeses asswage the vehemency of the heat. The chiefe of winter is half December, January, February, and halfe March. The cold is extreme sharp, but here the Proverb is true, that no extreme long continueth.

The winds here are variable, but like the thunder and lightning to purifie the ayre, I have seldome either seen or heard in Europe. From the Southwest came the greatest gusts with thunder and heat. The Northwest winde is commonly coole and bringeth faire weather with it. From the North is greatest cold, and from the East and Southeast as from the Barmudas, fogs and raines.

(Smith 1624:21)

Captain Smith also went on to describe the quick changes characteristic of Tidewater weather, a tendency that anyone who has lived in the area for much time can attest. Statistically, however, the peninsula's weather is rather mild, the average July temperature is 77.4°F, and the average in January is 38.8°F. Annually, the average is 58.5°F. About two years in every ten will have temperatures over 98°F in June and July and below 5°F in January. Annual rainfall averages about 47 inches, with approximately 8.6 inches of snow. The upper peninsula usually enjoys a long growing season with temperatures remaining above freezing for 175 to 200 days each year (U.S. Soil Conservation Service 1985).

The Land

Topographically, this part of the peninsula is characterized by a series of scarps and terraces, or plains ranging from 22 to 30 meters above sea level. Sites 44JC647 and 44JC633 are located on the Grove Plain that extends from the base of the Ruthville scarp eastward to Lee Hall scarp. Formed during the late Middle Pleistocene by the James River, the plain was subsequently severely eroded at its juncture with the river, causing high bluffs on either side of Grices Run (Oaks and Coch 1973).

The Soil

The soil complex around 44JC647 is a Slagel-Uchee with a 0 to 2% slope and is described by the U.S. Department of Agriculture as "Prime Farmland." By definition, such soils "produce the highest yields with minimal inputs of energy and economic resources, and farming these soils results in the least damage to the environment" (U.S.S.C.S. 1985:13). Typically, the soil is about 9 inches deep consisting of a top 4 layer of dark greyish-brown fine sandy loam and a second layer of yellowish-brown fine sandy loam. Of course, where deeply plowed, the layers are mixed. The subsoil below is usually a mottled sandy clay loam to a depth of 25 inches, becoming more clayey to a depth of 60 or so inches. This type of soil usually retains a high water table, and is well suited for agriculture, pasture and growing trees (U.S.S.C.S. 1985).

The Wildlife

Carter's Grove plantation consists of approximately 800 acres of undeveloped land bounded on the north by U.S. Route 60 and on the south by the James River. With both wooded and open land and no agricultural activity, the area is a haven for large numbers of birds, mammals, reptiles, amphibians and insects indigenous to southeastern Virginia. Deer are the largest mammals existing in the wild. So successful are these creatures at coexisting with people that they are quite common even in semi-urban areas like the city of Williamsburg. Also, common in both Williamsburg and nearby rural areas are opossum, raccoon, skunk, red fox, grey fox, grey squirrel, rabbit, groundhog, rat, and mouse. Birds found today at Carter's Grove include several varieties of hawk, falcon, bald eagle, osprey, black vulture, turkey vulture, crow, as well as numerous song birds and water fowl. When colonists first arrived at Martin's Hundred, passenger pigeons and possibly Carolina parakeets were part of the avian fauna. Several varieties of snakes, lizards, toads, salamanders, terrapins, and turtles are also common. The heat of summer brings a bounty of insects, many that sustain themselves by feeding off the local fauna and archaeological excavators.

The Setting

English colonization of the New World was the intellectual brainchild of Richard Hakluyt the Elder and his cousin Richard Hakluyt the Younger. They envisioned a nearly-utopian settlement that would not only supply England with much needed resources but would establish- a harmonious community of Native Americans and Englishmen living in freedom and happiness. The source of these immigrants would largely come from the ranks of the underprivileged in England that Hakluyt the Elder described as "valiant youthes rusting and hurtfull by lacke of employement" (Taylor 1935:319). Queen Elizabeth chose her confidant Sir Walter Raleigh to implement the task. Unfortunately, the first attempt, made at Roanoke Island between 1585 and 1587, ultimately failed. A poor relationship with the local Native American population, lack of proper leadership, and an unwillingness on the part of the majority of the settlers to work have been seen as major reasons for the abortive initial attempt (Morgan 1975).

The Englishmen who sailed up the Powhatan (James) in the spring of 1607, past the marl cliffs at Martin's Hundred on their way to Jamestown Island, had some of the same expectations the Roanoke backers had some 22 years earlier. Unfortunately, they also had some of the same problems. The Jamestown colony nearly failed almost as quickly as Roanoke. Disease and starvation brought on by an unfamiliarity with the environment, fear, laziness, or, as Earle (1979) suggests, sickness from drinking saltwater, caused the first settlers to abandon Jamestown in the spring of 1610. Having retreated partway down the river, the bedraggled lot met reinforcements with fresh supplies coming upriver and subsequently returned to the fort, preserving Jamestown's place in history as the first permanent English settlement in the New World.

The initial settlement of Virginia was the work of private enterprise. Lacking the funds and political backing necessary to underwrite the great effort needed for such an undertaking, King James I granted huge tracts of land to the Virginia Company of London. The Company was composed of private investors who were hoping to make a sizable return on their investment in the colony by sponsoring commodities England needed-silk, iron, and timber-while providing jobs and opportunity for the ever-growing ranks of underemployed English craftspeople and workers (Craven 1949).

Unfortunately, all Company efforts at turning a profit failed miserably. No one reason can be singled out for the disastrous circumstances between 1607 and 1618. Martial law, cruel punishments, lack of incentive for workers, too many gentlemen, too many specialized craftsmen, too many indigents shanghaied from the gutters of London and bars of Bristol, and a really contemptible attitude toward the Native American population combined with the extractive nature of the industries the Company wanted to operate when there was nothing worth extracting to spell trouble. The result was a very high death rate, desertion, and a general lack of productivity (Morgan 1975).

To make coming to America more attractive to potential colonists, and, by that, save the faltering Virginia Company, sweeping reforms were begun in 1618 under Sir Edwin Sandys' leadership. Most important, these reforms included the abolishment of Sir Thomas Dale's extremely harsh Lawes Divine, Morall and Martiall and provided for, among other things, legislative assembly, with representatives elected by the inhabitants of the various plantations. Although the laws they made had to be approved by the company, it was the first representative body in the New World. The Virginia Company also gave 100 acres to all "ancient planters," those in Virginia before Thomas Dale's departure in 1616. The new by-laws instituted the "headright" system by which 50 acres of land would be given to each person paying his own passage to Virginia or to anyone paying passage for another. A man who paid passage for himself, his wife, and two children, for example, would receive 200 acres. Large tracts of land were often granted to wealthy sponsors who would pay passage for many people and have them work off the debt by planting for him. A very modest quitrent of one shilling was charged annually for each 50 acres. Another deal, one that made Martin's Hundred possible, was the establishment of "particular" plantations in which investors would receive 100 acres for each share of stock bought in the company and 50 acres for each transportee. The tenants would pay off their passage to the investors by sharecropping-giving one half their profits to the investors for seven years. After that, they would be free to make their own way. To reduce the tax burden on the settlers, public officials were given tracts of land varying from 50 to 300 acres and several sharecropping tenants to work the land. The new reforms also began what amounted to a company store, headed by the cape merchant, to sell goods to the inhabitants, hoping to preempt trading with privateers (Craven 1949).

Although these reforms must have made Virginia much more attractive to both investors and settlers, the usual graft and corruption associated with the previous eleven years and the devastating effect, both physically and psychologically, of the 1622 war with the Powhatan, caused so much squabbling and infighting within the company that King James revoked the charter in 1624 and made Virginia a royal colony. Although not officially sanctioned by the king, the general assembly and the headright system continued, the latter having some success at populating the land.

Meanwhile, about the same time as the reforms were taking place, an important economic development was taking place. In 1617 the first load of West Indian tobacco was on its way to England. While Europeans had been using tobacco for medicinal purposes since the last part of the sixteenth century, they began smoking it for enjoyment in the early seventeenth century. A major breakthrough in the new industry came when John Rolfe, a Virginia planter and husband of Powhatan's daughter Pocahontas, successfully cultivated the better-tasting West Indies strain of tobacco in Virginia. Tobacco production quickly became the only successful enterprise in Virginia-not the industry the founders had imagined or the kind of industry they wanted, but a quick moneymaker nonetheless. In Jamestown "... the market-place, and streets, and all other spare places [were] planted with Tobacco ... the Colonie dispersed all about, planting Tobacco" (Arber and Bradley 1910:535). The whole colony, despite efforts of the company, and later the crown, to control its production, was consumed in planting the "noxious weed."

Martin's Hundred

As mentioned earlier, one of the Virginia Company's reforms provided for what were called "particular" plantations, granted to groups of investors and operated directly by the investors. They were like little colonies within the colony. Martin's Hundred was one of them.

The particular plantation of Martin's Hundred was probably chartered in 1618 by the Virginia Company of London for a group of investors known as the Society for Martin's Hundred. Granted 20,000 acres, the Society could administer the plantation any way they pleased to attempt to make money on the initial investment. The next year, some 220 men and women, presumably hoping to make money for themselves and the Society, arrived at Martin's Hundred. Any optimism they may have had was dashed four years later, at the time of the first Anglo-Native American War. Seventy-eight of the 140 inhabitants were killed by the Native Americans on March 22, 1622 and the remaining 62 were captured or fled the hundred, seeking safety in Jamestown. Martin's Hundred sustained a substantial 22% of the 347 fatalities recorded throughout the colony (Noël Hume 1982).

By 1623 about 50 settlers had returned to Martin's Hundred, but by the time a census was taken of the whole colony in February 1624/5, nearly half the 50 returnees had died of disease and only 27 people inhabited Martin's Hundred.¹ The Society's investment was apparently a poor one.

To better assess the situation in Virginia, the crown ordered the census (or "muster") to be taken. Those living in Martin's Hundred are shown in Table 1 (Jester and Hiden 1956).

Fortunately for historians, besides the names and often ages of the individual settlers and their servants, an inventory was taken of their provisions, arms, buildings, and other personal property. Although the muster-takers were inconsistent in what was counted from place to place, the Martin's Hundred muster seems to be fairly complete.

The nineteen men, five women, and three children counted at Martin's Hundred on February 4, 1624/25 appeared to be well-provisioned. Only two years earlier, one of the unfortunate 23 who died over the winter, Richard Frethorne, had complained in a letter to his parents in England of a shortage of food and other supplies: "... I have nothing at all, no not a shirt to my backe, but two Ragges nor no Clothes, but one poore suite, nor but one pair of shooes, but one paire of stockins, but one Capp ... I am not halfe a quarter so strong as I was in England, and all is for want of victualls, for I do protest unto you, that I have eaten more in one day at home then I have allowed me here for a Weeke..." (Kingsbury 1935:58).

Perhaps in anticipation of another attack from the locals, the little community at Martin's Hundred in 1625 was well-armed, with a full set of armor for just about each man, 26 matchlocks, 27 fixed pieces, 29 swords, a cannon, 91 pounds of powder, and 361 pounds of shot. The ordnance and most of the other weaponry were under the control of Henry Harwood, the "governor" of the settlement.

By the time of the dissolution of the London Company, it appears that the focal point of the hundred had been moved from the pre-war location at Wolstenholme Town near the river, to Site A up on the bluff, near the present-day Carter's Grove Mansion. During the 1970s, Ivor Noël Hume located Site A. The three houses listed in under Harwood's muster may have been at this site, along with most of the arsenal. The other four or five houses listed for Emerson, Addams, Jackson, and March were probably dispersed around this administrative center. Unfortunately, it is impossible to determine whether site 44JC647 was the site of one of the houses described in the muster, but both artifacts and documents indicate a slightly later construction date. The presence of a locally-made smoking pipe bowl fragment in the fill of one of the structural posts for the little house would tend to support a second quarter of the seventeenth-century construction date. The domestic pipe phenomenon, for various reasons discussed elsewhere in this report, probably did not occur until the beginning of the second quarter of the seventeenth century. Although it is not usually a sound practice to base conclusions on negative evidence, it does seem peculiar that no armor or gun parts were recovered from the site, although all households listed in the muster had both. There was, however, a fragment of a sword found at 44JC647 and all musterees had swords as well.

Very little specific information is available about the individuals that lived at Martin's Hundred after the 1625 muster. James City County records were destroyed during the Civil War. Land patents (records of the headright system), which did survive the burning of Richmond, are somewhat helpful in at least naming some individuals who patented land at the hundred during the remainder of the century, but again, sufficient detail in location is wanting.

Judging from the site identification and site examination evidence, there are at least eleven sites dating from the second quarter of the seventeenth century to the beginning of the eighteenth century. There appears also to be a community at Martin's Hundred during the seventeenth century with house sites chosen that were neither too near nor too far from the next neighbor. This development is traced carefully in David Muraca's 1993 thesis "Martin's Hundred: A Settlement Study."

It is indeed too bad that a more precise history of site 44JC647 cannot be written. Perhaps knowledge of the individuals who lived in the little house could shed light on why they lived there and what appears to be for a very short period. Not having the privilege of knowing such details, we must rely on other sources for information.

James Deetz (1987), while studying pipe stem bore diameters from Flowerdew Hundred, a contemporary particular plantation in Prince George County, found that the abandonment of small homesteads in the seventeenth century may have been associated with a marked decline in tobacco prices in the 1640s. Repeating Deetz's experiment at Martin's Hundred has shown that a group of sites-Sites A, B, CG-2, CG-11, and CG-8-were abandoned at approximately the same time in the mid seventeenth century as similar sites at Flowerdew.

Clustered farm sites at both Flowerdew and Martin's Hundred may have used the easily accessible land between 1625 and 1650, making it necessary for farmers to move to greener pastures. This spent land, combined with falling tobacco prices, may have led some tobacco cultivators to quit and return to England or find new, richer 10 land in northern counties where larger tracts were being patented in the second half of the seventeenth century.

Martin's Hundred seemed to cease to be an entity by the early eighteenth century. Martin's Hundred Parish joined York-Hampton Parish in 1713 (McIlwaine 1925:IV:316).

Summary

This discussion of world events in the sixteenth and early seventeenth centuries is important if we are to understand some of the processes going on at a smaller scale at plantations such as Martin's Hundred. It has been argued (Craven 1949; Morgan 1975; Braudel 1982) that England needed to colonize in order to rid itself of a portion of a growing population of underworked poor who were becoming a social problem in the late sixteenth century. Colonization is, of course, almost always accomplished or led by the elite of the homeland and it almost always requires the exploitation of someone-usually the native population. In the case of Virginia, however, it became evident very quickly that the native population was better annihilated than exploited on a long-term basis. It was necessary, then, to exploit the surplus lower-class English men and women. Until servitude by virtue of race began in the late seventeenth century, lower class English men and women would have to do. Indentured servants were treated like property and were often traded among the masters, beaten, poorly housed, and over-worked, especially in the Company and near-post-Company period (Morgan 1975).

Plowzone

More often than not, early seventeenth-century archaeological sites seem to be located in rural areas that have at one time or another been used for agricultural purposes. This usually results in the site having been extensively plowed, an activity that, as one would expect, has the effect of destroying a major portion of the site's stratigraphic record, especially the midden layer.² This is indeed unfortunate because the impermanent nature of such sites often means they frequently were not used long enough to develop much of a midden deposit. The plow, designed to overturn soil, mixes this fragile sheet refuse layer with more modern soil deposits as well as the top portions of deeper features such as trash pits, postholes, slot-fence trenches and wells. Until the late 1970s and 1980s many historical archaeologists felt that plowed sites were so devastated by the process that the resultant plowzone (6 to 18 inches of the top of a site) was devoid of any useful information and subsequently stripped it away with little or no analysis of the artifacts it contained, or its chemical constituents (Noël Hume 1982; Edwards et al. 1989).

Other archaeologists disagree with this practice. The most vocal advocates of analyzing the plowzone on seventeenth-century sites have been the archaeologists at St. Mary's City, Maryland. As early as 1978 Robert Keeler examined artifact and chemical distribution of the St. John's homelot in southeastern Maryland (Keeler 1978). More recent articles by Tim Riordan, Julia King, Dennis Pogue and others have examined how areas around a house were used by carefully analyzing artifact distributions throughout the yard (Riordan 1988; King 1988; Pogue 1989). Based on case studies regarding the horizontal and vertical movement of artifacts subjected to plowing (Ammerman and Redman 1978; Lewarch and O'Brien 1981), the St. Mary's archaeologists were able to discern changes in activity areas of a houselot over time. An excellent example of this process was work done at the van Sweringen site, where conclusions about yard use were drawn on analysis of the quantity and type of artifacts found in each area of the houselot (King and Miller 1987). Although plowing a site certainly does destroy a great deal of stratigraphic and provenience information, the Ammerman/ Redman and Lewarch/O'Brien studies show that the artifacts in fact are not moved more than several feet from their original source of deposition, retaining some of the integrity of their original horizontal distribution.

Extensive excavation of site 44JC647 would afford Colonial Williamsburg archaeologists the opportunity to experiment with plowzone artifact distribution in a 16 fashion previously untried. Plowzone archaeology usually consists of sieving plowzone from one meter square to five-foot square test units placed at random or regular intervals over a whole site, then using a computer plotting program such as SYMAP or SURFER to approximate the overall distribution of various types of artifacts on the site. Rather than just digging plowzone test units in various locations, artifacts within a 103.5 square meter area were piece-plotted three-dimensionally using a laser theodolite. The procedure was as follows: when an excavator discovered an artifact by careful trowelling, it was removed and replaced by a roofing nail with a white plastic tag attached, having a string of five-digit numbers was printed on the tag, using indelible ink. The number on the tag was written on a polyethylene bag in which the artifact was placed. This tag number would follow the artifact throughout its analysis. Using a prism rod, the exact three-dimensional location of the tag would be read by the laser theodolite and recorded in a small computer attached to the instrument. At the end of the day, the coordinates of each artifact were transferred from the site computer to a larger PC in the Archaeological Research offices. Later, the artifacts were inventoried using a FoxPro database program. Besides the Material, Type and other information explained later in the report, the artifact's tag number and subsequently its three-dimensional coordinates were included in the inventory database.

Using GEOSYS, now G-Sys, a geographical information system program developed specifically for archaeology by English archaeologist Dominic Powlesland, the artifact inventory was integrated with AutoCAD, the industry-standard mapping program, that produced maps of the site showing the exact location of each artifact tagged. Artifacts could be displayed by Type, Material, or any other attribute included in the inventory program. After the plowzone was fully removed, scale drawings of the features intruding the subsoil were made, digitized, and integrated into G-Sys so that the artifact information could be superimposed on the overall site plan (Figure 3).

In addition to piece-plotting artifacts in part of the plowzone, over 140 half-meter test pits were also dug approximately two meters apart (Figure 4). The soil was screened through quarter-inch hardware cloth and the artifacts inventoried as usual. The location and elevations of both the top and bottom of each test pit were recorded with the laser theodolite so a topographical map of the plowzone and subsoil could be constructed. Piece-plotting and digging test pits provided an opportunity to compare both types of data. An analysis of the plowzone layer at 44JC647 may be found in Chapter 5.

Features

After spending approximately two months testing the plowzone on 44JC633 and 44JC647, both sites were mechanically graded with a Drott 40 backhoe, a process that took about three days. An 1163-square-meter area at 44JC647 and an 898-square-meter area at 44JC633 were stripped to subsoil (Figure 5). After stripping, the areas were shovel-smoothed and hand-trowelled to reveal subsurface features. At 44JC647 several historic features were found besides the five structural postholes located during the piece-plotting. These include three additional postholes, three pits, 17 Figure 3. Distribution of artifacts in the plowzone. Figure 4. Site 44JC647 Plowzone test units. 18 Figure 5. Location of sites. and a slot fence trench. All of the features were drawn to 1:10 scale on metric graph paper.

Since a laser theodolite was in use, establishing a finer-grained grid system was unnecessary. As with the piece-plotting, the theodolite was always set up over the same point on the grid established during the previous archaeological examinations. The locations of features were recorded by placing two nails in the ground near each feature, recording the position of those nails on the feature drawing and recording the three?dimensional location of each nail with the laser theodolite. The nail coordinates were recorded on the scale drawing and could then be located on the grid, generating a composite drawing when the individual feature drawings were digitized.

After the features were drawn to scale, each was photographed with 35 mm color slide film and 60 mm black and white print film. All features were photographically recorded prior, during, and after excavation.

Features were also recorded using the Department of Archaeological Research's standard Context Record Form (Figure 6) that is fairly self-explanatory. Context numbers were assigned in consecutive order as dug, except in the case of meter squares in the plowzone.

Although carefully excavated, all soil from the features was screened through quarter-inch hardware cloth. Varying amounts of soil were saved from important contexts for later analysis, especially flotation, wet-screening, and chemical composition.

Figure 6. Context from recording form.

Artifacts

As described above, all piece-plotted artifacts were bagged separately with their tag number, context number, and the date written on the bag. Artifacts from features were not piece-plotted, so they were placed in bags bearing the context number and date only. The artifact bags were logged into the D.A.R. laboratory at the end of each day. The previous day's artifacts were washed and sorted daily by the Lab Technician assigned to the project. Since, at the Project Director's request, the artifacts were not to be returned to their bags, they were also labelled with their respective context numbers. Each artifact was entered into the artifact inventory system according to the attributes described in Figure 7.

Faunal material recovered from the features was sent to the Zooarchaeological Laboratory for analysis. A full report on the faunal material recovered is included as Appendix A. Selected botanical material recovered from flotation was sent to Dr. Stephen A. Mrozowski of the University of Massachusetts, Boston for analysis. Soil samples were analyzed by the Virginia Department of Agriculture at Blacksburg. Reports on the findings from these analyses are included in Chapter 5 of this report.

Domestic and imported smoking pipes as well as ceramics recovered from the site were cross-mended in accordance with standard archaeological procedure.

Figure 7. Artifact inventory record.

The Plowzone

Like most sites of this period, 44JC647 had been subjected to the plow, probably long after its demise in the middle of the seventeenth century. The ground surface that existed around the house when it was occupied was churned and mixed by plowing, and the artifacts and other archaeologically significant material that had been strewn about the site were displaced. In spite of these obvious drawbacks, much can be learned from the contents of the plowzone layer. How this layer was excavated at JC647 was explained in Chapter 4, but the results of the soil chemistry and artifact analysis are presented here.

Soil Chemistry

Four elemental chemicals are most important to archaeologists: phosphorus, calcium, potassium, and magnesium. Phosphorus anomalies seem to indicate the presence of human excrement as phosphorus is a basic constituent of human DNA. Calcium is found in bone, shell, and to a lesser extent in wood, and potassium appears to be related to the burning of wood, especially in an oxidizing atmosphere. The presence of high levels of magnesium has ambiguous archaeological meaning, but may be associated with very intense burning. Concurrent high concentrations of phosphates, calcium, and potassium may be characteristic of discrete refuse deposits. Fortunately, these elements are rather stable in the ground and may last for quite some time (Pogue 1988b; Stone et al. 1987).

Thirty-five random soil samples were taken from plowzone contexts at JC647. The samples were sent for analysis to the Virginia Cooperative Extension Service at Virginia Polytechnic Institute in Blacksburg, Virginia. A printout for each sample (Figure 8) was returned in a short time. The phosphorus, potassium, calcium and magnesium levels were given in pounds per acre and the zinc, manganese, copper, and iron values were reported in parts-per-million. (In order to convert lbs/acre into ppm, one simply divides the lbs/acre value by two.) Values of all the elements are given in Table 2. All of the values given in the chart, except pH, are parts-per-million. The last three columns are the grid coordinates of the sample.

The distribution of the chemical values can be more useful, however, if a graphic format is used. SURFER was used to plot varying concentrations of the elements.

22

Figure 8. Soil testing report, Virginia Cooperative Extension Service.

This topographical map of potassium values in the plowzone (Figure 9) shows a very high concentration of the element just outside what was probably the only entrance of the house, indicating possible dumping of fireplace ash right outside the door. Lesser but significant concentrations of potassium were found off the northeast corner, in the "back yard," suggesting that the inhabitants did occasionally bother taking the ashes a bit further away from the house.

The distribution of phosphorus values in the plowzone may indicate where human waste was being deposited around the house. Although the highest levels of phosphorus concentrations were in pit 50AJ-1151,³ there were lesser "peaks" around the yard as shown in Figure 10. Interestingly, phosphorus concentrations show a similar pattern to those of potassium, suggesting that the inhabitants may have dumped refuse other than fireplace ash in the area off the northeast corner of the house, perhaps after the pit mentioned above was completely filled.

Concentrations of the element calcium were very high (1200+ ppm.) in all three of the features tested. It is difficult to say exactly what lower concentrations in the plowzone samples really indicate. All of the samples tested had values of between 132 and 420 ppm., far below the 1200 ppm readings in the pits. There were no bones and very few shells within the plowzone, so it is unlikely that the anomalies in readings could have been determined by the presence of either. The "concentrations" of calcium at the southwest corner of the house may be associated with material plowed-up from nearby pits 50AJ-1151 and 50AJ-1152. The light concentration of calcium in the northwest 23

Table 2.
Soil Chemistry Values-Plowzone

Sample #	pH	P	K	Ca	Mg	Zn	Mn	Cu	Fe	B	XC	YC	ZC
291	5.5	12.5	42.0	252.0	42.0	2.5	16.1	1.1	11.6	0.1	22455	26041	1050
293	5.6	10.0	51.5	252.0	42.0	2.2	16.1	1.1	11.0	0.1	22453	26246	1014
295	5.4	9.0	55.0	204.0	32.5	1.9	16.1	0.9	11.3	0.1	22448	26446	1063
297	5.0	13.5	48.5	192.0	31.0	2.1	12.9	1.0	15.1	0.1	22452	26658	1045
312	5.5	12.5	55.0	228.0	37.0	2.3	16.1	1.0	12.5	0.1	22550	26152	1004
329	5.4	15.0	59.5	312.0	45.5	2.4	16.1	1.0	9.8	0.1	26739	27480	1012
331	5.9	15.0	117.5	312.0	61.0	2.7	16.1	0.9	10.8	0.1	26488	27102	1005
352	5.3	14.0	55.0	240.0	42.0	2.3	16.1	0.9	11.6	0.1	22752	26150	1001
369	5.4	15.0	69.0	348.0	61.0	2.6	16.1	1.0	9.6	0.1	27581	26562	977
371	5.6	15.5	67.5	252.0	41.0	2.3	16.1	1.0	11.2	0.1	27601	26479	971
388	5.3	8.0	44.0	192.0	29.0	2.0	10.4	1.2	12.9	0.1	22911	25799	973
392	5.0	14.5	47.0	216.0	33.5	2.3	16.1	1.0	12.4	0.1	22944	26144	1002
407	5.1	19.0	61.0	288.0	50.5	2.5	16.1	1.1	14.5	0.1	23106	25609	998
409	5.1	9.5	48.5	197.0	26.5	2.3	12.8	1.5	14.0	0.1	23729	24985	971
428	4.9	16.0	58.0	252.0	44.5	2.5	16.1	0.9	13.7	0.1	23245	25750	1000
447	5.0	11.0	40.5	192.0	29.0	2.0	16.1	0.9	13.8	0.1	23300	25650	1000
449	4.9	14.0	47.0	192.0	33.5	2.1	16.1	1.0	14.2	0.1	23300	25900	1000
468	5.2	17.0	67.5	336.0	63.5	2.8	15.1	0.7	10.6	0.1	23445	25750	1000
487	5.1	11.0	42.0	168.0	25.0	2.0	13.9	0.9	14.1	0.1	23500	25650	1000
489	5.3	8.0	39.0	156.0	18.0	2.5	9.1	1.3	11.3	0.1	23500	25850	1000
720	5.6	11.0	59.5	252.0	36.0	2.3	16.1	0.9	10.5	0.1	22562	25946	1003
760	5.5	16.0	58.0	252.0	38.5	2.4	16.1	1.2	11.7	0.1	22754	25949	1006
800	5.7	13.0	61.0	312.0	50.5	2.3	16.1	0.9	10.8	0.1	22954	25945	1002
1016	5.1	17.5	47.0	132.0	23.0	1.8	3.9	0.8	13.4	0.2	23638	25503	993
1027	5.4	11.0	45.5	168.0	21.5	1.7	8.6	0.8	11.3	0.1	23768	25886	1003
1028	5.7	12.5	58.0	240.0	33.5	2.2	4.3	1.2	9.8	0.1	22112	25334	993

Note: pH=pH value; P=phosphorus; K=potassiuim; Ca=calcium; Mg=magnesium; Zn=zinc; Mn=manganese; Cu=copper; Fe=iron; B=boron; XC= X coordinate (easting); YC=Y coordinate (northing); ZC=Z coordinate (elevation).

24

Table 2 (cont'd).
Soil Chemistry Values-Plowzone

Sample	pH	P	K	Ca	Mg	Zn	Mn	Cu	Fe	B	XC	YC	ZC
1030	5.5	5.5	58.0	216.0	41.0	1.6	4.6	0.8	9.6	0.1	22898	26933	1010
1033	5.5	12.0	46.0	192.0	27.5	1.7	6.5	0.7	10.6	0.1	21986	26583	1006
1035	5.4	10.5	45.5	192.0	32.5	2.2	6.1	0.8	10.8	0.2	23200	26892	1015
1038	5.4	11.0	45.5	180.0	20.5	1.7	6.7	1.1	11.6	0.1	21870	26389	1002
1040	5.7	8.0	47.0	204.0	31.0	1.9	5.9	0.9	12.6	0.2	23648	26426	1016
1047	5.6	10.0	37.5	204.0	19.0	2.1	7.2	1.5	13.4	0.1	23071	25301	987
1049	5.1	7.0	48.5	168.0	21.5	1.9	6.4	0.9	11.9	0.1	23462	26095	1007
1052	5.3	10.5	39.0	168.0	21.5	2.1	9.7	1.2	14.0	0.1	22316	26472	1007
1141	6.1	3.5	51.5	420.0	47.0	0.7	4.2	0.5	8.8	0.1	21488	26076	991

25 Figure 9. Plowzone potassium values. Figure 10. Plowzone phosphorus values. 26 area of the back yard may be associated with material dumped from the fireplaces that has dissolved into the soil over the years (Figure 11).

Feature 50AJ-1150

Feature 50AJ-1150, uncovered during the mechanical grading of the site, consisted of three distinct layers. The top layer was 50AJ-1150, a black, friable sandy loam with a great deal of charcoal, daub, 76 nails, 11 unidentifiable animal bones, an imported clay pipe bowl, a North Devon lead-glazed coarseware pipkin fragment, two case bottle fragments, a fragment of English flint, charred seeds, and charred fruit pits. The second layer was 50AJ-1154, a dark reddish brown silty loam containing only some daub and a single fragment of domestic lead-glazed coarseware from an unidentifiable vessel form. The third layer was 50AJ-1176, a reddish yellow sticky clay with a few charcoal inclusions and no artifacts. This layer appears to have been subjected to intense heat.

All of the soil from the three layers within the feature was saved and brought back to the department for wet-screening, flotation, and chemical analysis. Nine soil samples from the three layers were sent to the Virginia Cooperative Extension Service for Figure 11. Plowzone calcium values. 27 Figure 12. Site plan. chemical profiles. The results, listed in Table 3, are measured in parts per million. Letters with plus (+) and minus (-) signs indicate the Extension Service's interpretation of Low, Medium, High, and Very High content of the tested element.

As mentioned earlier, high amounts of phosphorus (P) usually indicate presence of human or animal excrement, high amounts of potassium (K) usually correspond with the wood ash deposition, and high calcium (Ca) readings signal the presence of bone or oyster shell. High magnesium (Mg) readings may indicate burning but correlations with the other elements listed above have not been firmly established (Pogue 1988b). Concurrent high concentrations of phosphates, calcium and potassium are usually characteristic of discrete refuse pit deposition (Stone et al. 1987).

The results of soil analysis for feature 50AJ-1150 do indicate very high values for calcium and magnesium and a corresponding high pH value, but medium to low readings in potassium and phosphorus. Considering the large amount of charcoal and charred material recovered from the feature, the relatively low potassium figures may seem surprising. However, potassium appears to be associated with ash, created by burning wood in an oxidizing atmosphere, perhaps on the surface or in a fireplace (Custer et al. 1986). While large amounts of charcoal (created by burning wood in a reducing atmosphere) were present, very little ash was recovered.

Little or no shell was recovered from feature 50AJ-1150, so the high calcium readings probably result from a combination of the small amount of bone found and the wood charcoal. Dennis Pogue's (1988b) consultation with Maryland soil scientists revealed that high calcium values may result from intense burning of wood as well as from the decay of oyster shell and bone.

The low-medium to low-high phosphorus values average 11.9 ppm for all nine samples. This is only 2.3 ppm above the control sample of 9.6 ppm, also a "medium" reading, and far below the 60 ppm values found in feature 50AJ-1151. Since phosphants 28

Note: pH=pH value; P=phosphorus; K=potassium; Ca=calcium; Mg=magnesium; Zn=zinc; Mn=manganese; Cu=copper; Fe=iron; B=boron. Indications next to the number (Low, Medium, High, Very High) are assigned by the soil testing lab.

Besides the chemical testing of feature 50AJ-1150, part of the fill was closely examined for surviving botanical remains. Wet-screen and flotation samples from contexts 50AJ-1150, 50AJ-1154, 50AJ-1176, 50AJ-1151, and 50AJ-1152 were sent to Dr. Stephen Mrozowski at the University of Massachusetts, Boston for identification. Along with obviously modern intrusive uncharred seeds from local non-cultivated plant-types, Dr. Mrozowski also observed charred grass seeds in association with the charcoal and charred peach pits recovered from the feature. He reports that "the charred grass, portulaca and chenopodium seeds may be significant discoveries. If the grasses could be identified as English species, they could represent some of the earliest known such examples unearthed in North America" (Mrozowski 1992:3). Unfortunately, identification of grass seed to the species level is difficult and as of this writing no one has been found to be willing to attempt the classification.

Feature 50AJ-1151

As with features 50AJ-1150 and 50AJ-1152, all of the fill from 50AJ-1151 was bagged and taken to the department for wet-screening, flotation, and chemical testing. Five soil samples were from the feature were sent to V.P.I. for chemical profiles (Table 4).

Immediately obvious from the table above are the very high phosphorus values obtained from feature 50AJ-1151. Since high phosphorus levels are usually indicative of human or animal wastes, because phosphorus is one of the basic constituents of the DNA molecule and therefore present in animal tissue and wastes (Eidt 1985), it appears that this feature may have been a receptacle for either human excrement or animal tissue.

The high calcium values found in 50AJ-1151 may be due to different factors rendering similar results in feature 50AJ-1150. Feature 50AJ-1151 contained about 35 oyster shells and over 140 animal bones, the slow decay of which probably caused both the high calcium and pH values. Unlike feature 50AJ-1150, magnesium levels in 50AJ-1151 were only slightly higher than the control sample, indicating that little or none of the soil had been subjected to intense burning.

Iron, copper, manganese, and zinc values for 50AJ-1151 were also notably higher than those found in 50AJ-1150. Although the reason for high or low values for these elements on archaeological sites has not been determined, anomalies among features, or between plowzone and features, may depend upon the kinds of materials discarded in the pits.

Potassium values for the soil samples taken from 50AJ-1151 are again somewhat surprising given the presence of charcoal and burned wood in the pit. Little surface-burned wood ash was found however, so the reasons contributing to the low to middle readings in feature 50AJ-1150 are probably similar for 50AJ-1151.

As the flotation samples from 50AJ-1150, samples from 50AJ-1151 were also analyzed by Dr. Mrozowski. All of the botanical remains he examined were probably modern weeds and grasses that may have been intruded by earthworms and/or groundhogs (Mrozowski 1992). What Dr. Mrozowski did find interesting, however, was the presence of many tiny snail shells: "these shells, representing some five different species of snail, strongly suggest that these context contained rotting organic material. The fact that so few were recovered from other contexts examined, indicates a pattern that

Note: pH=pH value; P=phosphorus; K=potassiuim; Ca=calcium; Mg=magnesium; Zn=zinc; Mn=manganese; Cu=copper; Fe=iron; B=boron. Indications next to the number (Low, Medium, High, Very High) are assigned by the soil testing lab.

Although the lack of viable botanical evidence was disappointing, the presence of the snails may provide some insights on the origin and seasonality of the feature fill. Besides the snails, several animal species were identified from screened and floated feature fill. The flotation, wet-screening, and species identification was done by Steve Atkins under the general supervision of Dr. Joanne Bowen, Colonial Williamsburg's Staff Zooarchaeologist. Identifiable taxa recovered from feature 50AJ-1151 included cow, pig, partridge, bald eagle, goose, box turtle, frog, black drum, and croaker.

Undoubtedly, the reason the residents or builders of the little farmstead dug pits 50AJ-1151 and 50AJ-1152 was to extract clay for daubing what were likely wattle-and-daub walls and chimney. The holes then became convenient receptacles for whatever needed to be buried. The significant traits revealed by the various analyses of the fill in feature 50AJ-1151 are high phosphate values, low artifact count, consistent fill, snails, and a few bones from several families. From these traits it may be suggested that the feature was filled during one summer immediately after the house was constructed and was used a dump for both potentially smelly kitchen offal and possibly human wastes. The fish and amphibians found in the feature, as well as the snails, are ordinarily only active in the summer months. The consistency of the fill indicates quick filling and the few lenses of clean sand may be the result of heavy rain during summer storms characteristic of the area. High phosphates indicate animal waste or rotting flesh, two things best buried in the warm summer months.

Feature 50AJ-1152

Feature 50AJ-1152 was also uncovered during the mechanical grading of the site, and, like 50AJ-1151 consisted of a single context with no distinct layers of fill. This feature, too, was highly disturbed by a groundhog burrow that was excavated as context 1168. The fill was similar to that of 50AJ-1151 in color and texture, also a friable, dark reddish brown sandy loam with light inclusions of brick, charcoal, shell and marl.

Again, since only three features were found at 44JC647 that were likely to hold a significant amount of artifactual, faunal, and botanical information, all of the fill from 50AJ-1152 was removed to the department for wet-screening, flotation, and soil 31 chemical analysis. Six soil samples were taken from various parts of the feature and sent to V.P.I. for analysis (Table 5).

Very high calcium values are found in all three of the tested pits, probably due to the decay of faunal bone and shell. The striking anomaly in 50AJ-1152 is the consistently high magnesium values, with no apparent intensive burning. Unfortunately, little is known about the cause of high magnesium values in archaeological soils. What appeared to be wood ash was mixed in with the feature fill, but is not reflected in the potassium levels. The phosphate readings are all within or only slightly higher than the control indicating little or no deposition of animal or human wastes.

The botanical analysis of flotation samples from 50AJ-1152 were as disappointing as those from 50AJ-1151, with no identification of seeds dating to the seventeenth century. Snail shells were again present in the samples, probably indicating, as Dr. Mrozowski suggests, rotting organic matter.

The faunal assemblage identified by Atkins and Bowen seemed also to demonstrate a summer filling, as indicated by the presence of blue crab, turtles, sliders, croaker, and drum fish. Additionally the pit contained the remains of partridge, chicken, opossum, squirrel, pig and cow, along with Old World rat and house mouse.

Although essentially similar in composition and contents, pits 50AJ-1151 and 50AJ-1152 have notable differences in some areas. A "glue-fit" cross-mend (in this case a pipe stem) usually indicates that the pits were probably being filled simultaneously. It is also likely, therefore, that they were originally dug at the same time, probably during the construction of the house. A more detailed comparison of the artifacts contained in the two features will be discussed in Chapter 6.

The Slot Fence

A little over 64 square meters (725 square feet) of the yard south of the structure was enclosed by what was probably a low fence no more than four feet tall. The fence was defined by the trench that was dug for the upright palings. The western section of the slot extended from the southwestern posthole of the structure toward the southwest terminating at a length of 7.35 meters (24.1 feet). The 35 to 40-cm (14 to 16-inch) wide trench intrudes both the structural post for the little building and feature 50AJ-

On the eastern side of the enclosure, the trench also begins at the corner structural posthole and extends southwesterly 8.3 meters (27.3 feet) where it makes a right angle, extending another 2.5 meters (8.2 feet) toward the west-northwest. There the trench ends, failing to complete the enclosure by connecting to the western trench. There is a 5.3 meter (17.1 foot) gap between the eastern and western fence lines. A gate or alternate fencing type that apparently left little or no archaeological evidence, must have been used to complete the enclosure.

Similar features were found at Site A (Figure 13), located a quarter-mile west of 44JC647 (Noël Hume 1982), and at a seventeenth-century domestic complex (44HT55) in Hampton, Virginia (Figure 14) (Edwards et al. 1989), both of which also date to the 1630s. The enclosure at 44JC647 was strikingly similar to one surrounding a yard behind Structure A at the Hampton site. The fence at Hampton was also located on the south side of the structure and formed a 26-foot square. The only gap in the Hampton enclosure, however, was at the southeast corner and measured less than a foot wide.

The slot fences uncovered at Hampton and Site A were constructed by digging a trench six to eight inches wide and a foot or so deep, placing upright planks in the ditch, and securing them by packing dirt around their bases and attaching a stretcher near the top. The remains of split planking were found at both sites. Although the fence at 44JC647 appears to have been constructed in a similar way, whole saplings rather 33 Figure 14. Slot fence features at the Hampton site (44HT55). Figure 15. Detail of slot fence feature. 34 than split planks were used, much like a pale or palisade. Remains of varying-sized round wooden logs or stakes were found within the slot trench fill (Figure 15). Similar fences were found by Robert Keeler at St. John's, an early seventeenth-century house site near St. Mary's City, Maryland. His analysis contended that such fences were characteristic of early or poor plantations, whereas post and rail fences suggested permanence, later times, or prosperity (Keeler 1978).

Soil samples taken from the plowzone within the enclosure did not indicate high phosphate values that would have been characteristic of an area used to confine animals. It is likely that the low fence probably served to keep animals out, rather than to restrict them. Most domestic animals seemed to have been allowed to range freely, foraging off the local vegetation. The enclosed area, located on the south, sunny side of the structure may have served as a kitchen garden where the residents would attempt to raise a few vegetables during the long Virginia growing season. Keeping rooting animals such as pigs and cows out of the garden would have been imperative.

The Structure

The remains of only one small, earthfast structure were found at 44JC647 (Figure 16). The main part of the building was erected around six large structural posts with a lean-to off the east end indicated by two additional, smaller postholes. The building, excluding the 3.2 meter (10.4 foot) by 2.1 meter (6.9 foot) lean-to, measured 7.5 meters (24.6 feet) east-west by 5.0 meters (16.4 feet) north-south. The two western postholes were four meters (13.1 feet) from the central postholes, which, in turn, were 3.5 meters (11.5 feet) from the eastern pair of postholes. The building appears to have been traverse-raised (built end-to-end) since the distance between sets of north and south postholes do not vary, but distances between the ends and central posts do Figure 16. Detail of structure. 35 (Edwards et al. 1989). No evidence of a fireplace was found during the excavation of building, but it was most likely located on the same end as the lean-to (east).

The lean-to off the east end of the structure measured 2.1 meters (6.9 feet) east-west and 3.2 meters (10.4 feet) north-south. The two postholes supporting the lean-to (50AJ-1170 and 50AJ-1172) were shallow (about 18 cm or 7.1 inches) and only 1170 contained a distinct post mold (50AJ-1171). Only nails and nail fragments were recovered from 50AJ-1170, -1171, or -1172.

The postholes and postmolds making up the remains of the structure were remarkably similar in some respects and different in others. Table 6 summarizes their physical dimensions.

Postmold size varied little among all of the structural posts, indicating consistency in their diameter. It is also interesting to note that all but one post mold (50AJ-1167) intruded beyond the bottom of the post hole, meaning no fill seemed placed in the holes to "square" the building. The depth of the holes dug for the posts was also fairly regular, although their sizes and shapes varied significantly.

Figure 17. Structure at 44JC647 (looking south).

Introduction

The procedures for handling the artifacts and other archaeologically-significant finds recovered from site 44JC647 were explained fully in Chapter 4. Briefly reiterating, most of the artifacts found in the plowzone were located three-dimensionally using a transit. Others were grouped in the contexts from which they came, such as plowzone, test pits or those found through screening soil from specific areas of the plowzone. Artifacts from each feature were bagged together. At the end of each day, all of the finds were taken to the main laboratory for processing. This included the sorting, washing, and labelling of each piece. An inventory of all material was recorded using FoxPro, a database management software, after which artifacts from each context were laid out on tables for analysis.

The first analysis to take place was the measuring of the bore diameters of imported clay smoking pipes. In the 1950s J.C. Harrington, a Park Service archaeologist working in Jamestown, discovered that the diameter of imported clay smoking pipe bores became smaller at a constant rate as time progressed (Harrington 1954). The phenomenon is rather consistent and can be used to date sites as well as provide indications of occupation period and duration (Figure 18). Lewis Binford devised a simple regression formula that provides a mean date for pipe stems of various bore diameters (Binford 1962). Although this formula is useful in dating deposits from the late seventeenth century to the late eighteenth century, its accuracy is questionable for both the early seventeenth and nineteenth centuries (Deetz 1987; Noël Hume 1982). Binford mean dates only furnish a single date, and provide no information regarding Figure 18. Table of stem hole diameters (after Harrington 1954). 38 duration of occupation. Harrington histograms, graphic representations of each bore diameter are much more informative, especially when comparing numerous sites. Both methods were used in the analysis of imported pipes recovered from 44JC647.

Besides measuring the bore diameters of the imported pipe fragments, both ceramics and pipes were mended and cross-mended. First, all ceramics and pipe fragments from the same context that fit together were mended by gluing, then they were pulled from their respective contextual groups and sorted according to ware type or pipe type. As many pieces as possible were subsequently glued together. These are referred to a "contiguous cross-mends," or actual glue-fits from differing contexts. Sherds or pipes from different contexts that appeared to come from the same vessel or whole pipe but do not actually glue together were recorded as "non-contiguous cross-mends." Cross-mending ceramics is a fundamental part of archaeological analysis and must precede most systems of artifact analysis. The tedious and time-consuming procedure has several purposes. It relates contexts to each other that may have nothing else in common except that they are on the same site. It also provides a "minimum vessel count," i.e., the fewest number of ceramic vessels discarded on the site. Mending of ceramic vessels often makes the identification of the vessel form possible by allowing the analyst to see more of the pot, tankard, or plate he or she is attempting to identify. In the case of JC647, another purpose was accomplished by cross-mending ceramics. Since many of the sherds that mend with each other were piece-plotted and therefore their exact location was known, distances between and among cross-mended sherds could be measured in order to determine approximately how far plowing spread ceramics from a common source.

Sherd-Level Analysis

Some analysis of the ceramics and pipes from a site can be accomplished prior to cross-mending and obtaining a minimum vessel count. These procedures analyze the raw numbers of various artifacts, usually to look at intrasite functional areas, site function as a whole, and distributions of various artifact types. This type of analysis, called "sherd-level analysis," was the first step in studying the artifacts from JC647. Analyzing the site from the "vessel level," i.e., after the cross-mending determined minimum vessel count, can be more revealing and gives a more accurate assessment of relative wealth and site duration, the main emphases of this report. The sherd-level and vessel-level identification of the ceramics from JC647 was done by, or under the direct supervision of William E. Pittman, a well-known expert in the identification of seventeenth-century ceramics.

Only three categories of ceramics were recovered from 44JC647: coarseware, stoneware, and tin-enamelled earthenware (delft). The physical distributions of the three categories in the piece-plotted section of the plowzone are depicted in Figures 19-22. Unfortunately, little can be speculated from the apparently random horizontal positions of the 552 sherds. Plowzone test pits turned up an additional 196 ceramic fragments, but the only concentrations were those plowed out of the two refuse pits 50AJ-1151 and 50AJ-1152. Site-wide, coarsewares made up 61% of the ceramics, tin-enamelled wares 22%, and stonewares 17%. Approximately 63% of the coarseware 39 Figure 19. Distribution of coarseware, stoneware, and earthenware. sherds found on the site were of European manufacture, while the remaining 37% were made locally (Figure 20).

Ceramic Vessels

Ware Types

As explained earlier, the principal function of the cross-mending process is to determine the minimum number of individual vessels recovered from a site. Knowing the ware type and functional use of each vessel aids in terms of use-areas, status, and gender roles (King 1988; Pogue 1988a; Riordan 1988; Yentsch 1991a, 1991b). The ceramic assemblage at 44JC647 consisted of eighteen individual vessels identified by D.A.R. Curator of Collections William E. Pittman. Table 8 illustrates the types and Figure 27. Distribution of domestic pipes in plowzone. 49 Figure 28. Percentage of domestic pipe fragments.

Table 7.
Domestic Smoking Pipes

Cat #	Element	Color	Decoration	Remarks
404	bowl	red	rouletted	stars around rim
406	bowl/stem	red	rouletted	stars; marbelized body
408	bowl/stem	red w/white slip	plain	bowl w/ heel
409	bowl/stem	reddish	stippled	triangles; heelless bowl
410	bowl	buff	stippled	stippling around rim
411	bowl/stem	reddish	plain
412	bowl/stem	red	stippled/punct	white slip on bowl
413	bowl/stem	grey/tan	plain
414	bowl/stem	pinkish buff	rouletted/incisd	zig-zag; chevrons; circles
415	stem	grey	none	faceted
416	stem	red	none	marbelized body
417	stem	red	none	marbelized body
418	bowl	red	punctate	vertical line of circles
419	bowl/stem	pinkish-buff	none	heelless; hematite inclusions
420	bowl/stem	buff to grey	none	faceted stem
421	stem/heel	pinkish buff	none	hematite inclusions
422	stem tip	light buff	none	teeth marks(?) in stem
50
423	bowl/stem	light grey	none	fine sandy paste
424	bowl base	reddish	none	with heel
425	stem	reddish	none
426	bowl base	pinkish-buff	none	heel; hematite inclusions
427	bowl rim	buff to pink	rouletted	rouletting like imported
428	bowl/stem	reddish	none
429	heel	red	none
430	stem	red	none	facet along bottom of stem
431	bowl/stem	light red	none	heelless
432	heel/stem	light pink-buff	none	hematite incls.;molded
433	bowl/heel	lt. grey to pink	poss maker mark	molded
456	stem	reddish	none

characteristics of each, while Figure 29 illustrate ceramic percentages by material and type.

The site contained no Westerwald stoneware, a fine quality blue on grey ceramic type frequently found on contemporary Tidewater sites. The presence of North Italian slipware suggests that JC647 was occupied during the 1630s (Noël Hume 1982). It and the Mediterranean sgaffitto slipware point to the cosmopolitan nature of early seventeenth-century artifact assemblages commonly recovered from area sites. Navigation laws passed in the mid-seventeenth century legally restricted trade to goods manufactured in England or shipped on British vessels. Ceramic types other than those produced in Britain, except Chinese export porcelain, are not often recovered from sites dating to the latter part of the seventeenth or eighteenth centuries.

As discussed in Chapter 4, the analysis of the distribution of artifacts in the plowzone can be revealing in terms of determining the use of the space around a dwelling for different activities. On sites occupied for a lengthy period, this analysis can demonstrate how the yard was used differently by various occupants (King 1988). Ceramic vessel locations can be used to help determine these activity areas. For example, finding several milk pans in a specific area could mean that the processing of dairy products was taking place there. Unfortunately, the distribution of vessels at JC647 appears random, similar to the other types of artifacts recovered from the plowzone (Figure 30).

Artifact Displacement

One of the basic tenets of plowzone archaeology is the assertion that, although plowed, artifacts really do not move far from their place of original deposition (King and Miller 1987). Piece-plotting the plowzone and subsequently cross-mending the ceramics 51

Table 8.
Summary of Ceramic Vessels

Cat #	Material	Form Type	Ware Type	Element	Remarks
250	stoneware	bottle	Frechen	neck, body, base	mask on neck, coat-of-arms
251	stoneware	bottle/jug	Frechen	body
252	stoneware	bottle/jug	Frechen	body, neck	mask on neck, coat-of-arms
260	earthenware	mug	Delftware	base	blue/manganese
261	earthenware	hollow ware	Eng. delft	body
262	earthenware	hollow ware	Eng. delft	lateral handle
263	earthenware	dish	Eng. delft	rim, body	green on white
270	coarseware	dish/platter	slipware	rim, base	Mediterranean
271	coarseware	flatware	slipware	rim, body	Mediterranean
272	coarseware	bowl	N. Italian	rim	"Pisan" slipware
273	coarseware	bowl	N. Italian	rim, body	marbelized slip
274	coarseware	hollow ware	N. Italian	body	marbelized slip
280	coarseware	pan	domestic	base, rim, body	hematite incls.
281	coarseware	storage jar	domestic	body	hematite incls.
290	coarseware	storage jar	domestic?	body	micaceous
300	coarseware	bulbous mug	domestic	body, rim	hematite incls.
310	coarseware	pipkin	N. Devon	rim	plain
311	coarseware	bulbous mug	black-glazed redware	foot, handle, base

52 Figure 29. Ceramics by ware type. Figure 30. Distribution of ceramic vessels. 53 and pipe stems afforded an opportunity to closely examine this phenomenon. The drawing of the site and cross-mends in Figure 34 illustrates the distances some sherds moved away from each other during the course of plowing. Vessel C seems to illustrate this point, having a distance of 24 meters (over 78 feet) between two of its fragments. The average distance between cross-mended sherds at JC647 was 7.92 meters, or about 26 feet. This is a considerably longer distance between cross-mends shown in the Ammerman-Redman and the Lewarch-O'Brien studies mentioned earlier. The assumption that the movement of artifacts in the plowzone can be traced by piece-plotting cross-mends tends to de-emphasize the possibility that sherds from a single vessel could have been discarded many meters apart from each other very soon after the vessel was broken.

However, several potentially important facts were revealed from the distribution of the ceramic vessels at JC647. In their 1993 Society for Historical Archaeology paper, Julia King and James Gibb noted that a significant percentage of the vessels recovered from a seventeenth-century house site in Maryland were found only in plowzone contexts. This was also true at JC647. Thirty-nine percent of the ceramic vessels found at JC647 were recovered exclusively from the plowzone. Only 22% were found exclusively in features, and another 39% were found in both features and the plowzone. This is significant because it reiterates the importance of careful sampling of the plowzone on any site that has been subjected to agricultural activity. If the plowzone had simply been stripped off without sampling, the functional categories of the vessels recovered would have been significantly altered, possibly altering the interpretation of the site as well. In addition to simply changing the percentages of functional types recovered from the site, it was also noted that the three North Italian vessels were recovered exclusively from plowzone contexts. The features contained none of this particular kind of ceramic. The three tin-enamelled earthenware vessels were found only in features or in plowzone contexts very near the features, suggesting they were plowed out of the pits after deposition. As noted in Chapter 5, the refuse pits at JC647 appear to have been filled soon after the house was built, so artifacts recovered from the pits would likely relate to the very earliest years of occupation. The tin-enamelled earthenware could have been brought from England as personal items to facilitate setting up housekeeping. The North Italian slipware, since it was only found in the plowzone or midden layer, was perhaps acquired after the pits were filled, and therefore after the occupants settled in Virginia.

Functional Analysis

Each of the ceramic vessels recovered from 44JC647 performed a certain function for the people who owned them. It is desirable to classify the vessels into functional categories to better understand the various activities that the use of the vessels implies and as a basis for comparing this site with other contemporary sites in the early Chesapeake. Ann Yentsch (1990, 1991a, 1991b) has devised several functional categories for ceramic vessels based on the POTS typologies suggested by Beaudry et al. in their 1983 article in Historical Archaeology. Yentsch's activity categories include Food Preparation and Storage, Bottles, Serving and Dining, and Drinking. In her 1991 54 article in The Archaeology of Inequality (McGuire and Paynter 1991). Yentsch suggests that coarsewares, because of their earth-tone colors, are associated with women and therefore food preparation, whereas the refined wares such as delft and porcelain are associated with men, i.e., dining and drinking activities. She also suggested that "the symbolism of pottery and porcelain vessels is a consequence of the social rank of the people who use the vessels" (Yentsch 1991b:193). Edwards and Brown (1993) have also suggested that the ware types associated with various activities may also be a function of status. The higher percentages of refined wares in dining/serving and drinking activities, the higher the status. They also suggest that the presence of non-food related ceramic vessels (chamber pots and drug pots) also indicates higher status. It must be emphasized at this point that none of this analysis can consider those vessels made of wood or pewter, as those materials do not ordinarily survive in seventeenth-century Tidewater archaeological contexts. Tables 9 and 10 show the ceramic vessels recovered from JC647 by function and ware type.

Table 9.
Functional Categories

	No.	Pct.
Food Preparation and Storage	7	39
Bottles	3	17
Serving and Dining	5	28
Drinking	3	17
TOTALS	18	100

Table 10.
Percentages of Vessels by Function

	Food Prep/Store		Bottles		Serving/Dining		Drinking
	No.	Pct.	No.	Pct.	No.	Pct.	No.	Pct.
Stoneware	-	-	3	17	-	-	-	-
Tin-Enamelled	2	11	-	-	1	6	2	11
Coarseware (local)	3	17	-	-	-	-	1	6
Coarseware (European)	2	11	-	-	4	22	-	-
TOTALS	7	39	3	17	5	28	3	17

Introduction

Site 44JC647 is located approximately one?half mile from the eastern shore of the James River, within close proximity of a ravine system draining into Grice's Run. Faunal remains, dating ca. 1630-1650, were recovered primarily from two trash pits within the site's boundaries. In addition, a small number of bones were recovered from a slot fence (context 50AJ-1153), an area of the plowzone that sealed one of the trash pits, other areas of the plowzone surrounding the site, and a shallow circular pit containing a large quantity of charcoal, ash, and carbonized peach pits and seeds (contexts 50AJ-1150, -1154, and -1175). Faunal remains excavated from this site were analyzed at the zooarchaeology laboratory at Colonial Williamsburg's Department of Archaeological Research, under the direction of Dr. Joanne Bowen.

While faunal remains provide important information for the study of foods consumed by past populations and the system through which they are procured, distributed, prepared, and consumed, methods used to recover them during archaeological excavations influence the range of taxa recovered from a site. Studies have demonstrated (Thomas 1969; Clason and Prummel 1977; Grayson 1981; Cruz-Uribe 1988; Shaffer 1992) that smaller species are not always recovered using the standard ¼-inch mesh. This study attempts to examine the impact of archaeological retrieval methods on interpretations of the diversity and relative importance of taxa in the diet of the occupants of site 44JC647.

The two trash pits (contexts 50AJ-1151 and -1152) were located near a small, impermanent domestic structure with a lean-to to the east side and a slot fence enclosing the southern perimeter. One pit (context 50AJ-1152) was cut by the slot fence and was somewhat circular in shape and filled with brown loam mixed with gray ash; the other pit (context 50AJ-1151) was smaller in shape and approximately 1 meter in depth. It was filled with brown loam and ash, with wood fragments near the bottom.

Recovery methods used to excavate the site included using ¼ inch screens for fill from all features, with the addition of flotation and wet-screening samples from all of the ¼-inch screened matrix from the two trash pits, as well as the shallow circular pit. This testing of approximately 2,300 liters of soil from the three features included flotation and wet-screening through 1 16-inch mesh.

To recover botanical remains and microfauna not generally recovered with ¼-inch mesh screen soil samples were both floated and wet-screened. From the heavy fraction of both the floated and wet-screened material, remains of small animals recovered increased the range of identified species by 25%. Botanical remains were also recovered from both the wet-screened and floated material. Among them were a number of 78 seeds, along with twelve corn cob fragments. These have been identified by Rollin Woolley of Colonial Williamsburg, as gourd seed corn approximately 8?9 inches long with 24?30 rows of kernels. Further study and analysis of the corn will be continued by Bruno Marino and seed identification will be conducted by Stephen Mrozowski.

Analysis of the faunal remains recovered from both the ¼-inch mesh screens shows that the site's occupants relied on domestic animals for their primary food intake, supplementing them with wild species. Analysis of the microfauna recovered through flotation and wet-screening revealed the presence of four additional wild taxa-frog, rat, mouse, and opossum. Although these recovery methods were time consuming, the additional taxa produced by these methods enhanced the picture of the diversity actually present in the assemblage. The presence of an opossum, which is considered a food animal, increases the range of animals consumed. But the presence of the frog, rat, and mouse, all of which are commensal animals, introduces the question of whether the two trash pits had been left open for any length of time. Although no evidence of rodent chewing was found on any of the bones, other evidence indicates the contents of the pits had been exposed. Contained in the macro faunal assemblage were several bones showing evidence of extensive dog-chewing. Also enforcing the interpretation that the pits had been left open is the presence of land snails in the flotation samples (Mrozowski 1992).

Methods and Analytical Techniques

The methods used for identifying and analyzing the faunal assemblage followed the standard procedure of the zooarchaeological lab at Colonial Williamsburg. All bone from the assemblage was sorted into identifiable and unidentifiable elements. The unidentifiable bone was assigned to a taxonomic order by class (fish, bird, mammal, etc.), then counted and weighed.

The bones were identified using a comparative skeletal collection in the zooarchaeology laboratory, created and maintained by Dr. Bowen. One bird species was identified using the larger collection at the Smithsonian Institution. Each element was identified to the lowest taxon possible, along with information on the element, side, location, and weight as well as other characteristics (i.e., fusing stage of the epiphyses, relative size, burned, butchered, or chewed, etc.), then entered into a dBase compatible computer program. Measurements were taken on elements using procedures described by von den Driesch (1976). Minimum number of individuals was determined by visual comparison, using characteristics such as side, size, age, tooth wear and stage of eruption to estimate the numbers of individuals present in the assemblages. Biomass was based on values established by Reitz and Cordier (1983). With these values entered into the computer program, a summary chart was created (Table 1), which provided the percentage of each taxa based on the following quantitative procedures; the number of identified specimens, bone weight, minimum number of individuals, pounds of usable meat and biomass values.

Many natural and cultural variables can modify faunal remains, leaving them so altered from their condition at the time of disposal that the picture of dietary patterns they provide sometimes has little resemblance to reality. Natural factors such soil acidity, 79 climate, carnivore chewing, and rodent chewing each modify the condition of bones once discarded. Humans insert additional sources of modification by trampling discarded bone. Additionally, activities related to the processing of the carcass, followed by the preparation of meat and disposal of bones after eating contribute additional alterations. Even breakage from field excavation techniques alters bone, sometimes leaving them unidentifiable (Grayson 1984). Together, these variables can so significantly alter the condition of the bone that it is difficult, or even impossible, to determine dietary patterns from the faunal assemblage.

In order to determine whether the faunal assemblage could accurately reflect dietary patterns, all bones were carefully examined for any evidence of natural or human-related modifications (Bonnichsen and Will 1980). All evidence of carnivore chewing, rodent chewing, weathering, burning, and butchering were carefully recorded and the data was entered into our dBase program.

Damage to the bones was caused primarily by humans. Through butchering, trampling, and burning, bones were fragmented, often into tiny pieces. Many bone fragments were also burned, probably the result of cracking from heat or fire. Some carnivore chewing was evident on some of the bone, but no evidence of rodent gnawing was seen. Some bones were weathered, indicating they had been exposed to the open environment for some time. But overall, the bones were in relatively good condition, and the data therefore provides a view of the diet of those individuals who lived on the site.

The study of dietary patterns uses quantification as a major tool to interpret the relative abundance of each taxon identified from a faunal assemblage. Four measures of relative abundance are commonly used. The most basic method, the number of identified specimens (NISP), counts the number of elements to measure the relative abundance of a taxon. A second technique used is the minimum number of individuals (MNI), a method which compares right and left sided elements in relation to age, size, etc. to establish the least number of individuals present in the assemblage. Although this measure shows the relative abundance of species, it is figured in terms of the number of individuals, not how much each contributes to the diet. Relative size in this method is disregarded. To determine the dietary importance of a taxon two other measures are used. One is to measure the minimum amount of meat represented per taxon by figuring the average weight of the taxon. The MNIs are then multiplied by the average weight (Grayson 1979). The second method, referred to as biomass, is based the allometric relationship between the weight of the bone and amount of flesh it supports (Reitz et al. 1987; Wing and Brown 1979).

Meat weight estimations are important in documenting the relative contribution of a species to the diet, but they can be applied only to bones excavated using trowels and ¼-inch mesh screen, since those recovered from flotation and wet-screening were only samples from the larger excavation. But even though microfauna found in the fine-grained recovery methods cannot be included in the dietary estimates, they do provide a better measure of richness, or number of taxa found in the assemblage. Thus, we have determined diversity by adding those taxa found in the flotation and wet-screen samples to the species list obtained from the ¼-inch mesh sample. The species identified increased by 25 percent.

Description of Identified Taxa

Including those taxa recovered from the flotation and wet-screen samples with those taxa recovered from the ¼-inch mesh screen sample, a total of sixteen taxa were identified from the site. On the basis of biomass figures, the domestic species represented 64.5 percent of the diet, whereas the wild species contributed a much smaller portion, 8.5 percent. A brief description of each taxon identified follows.

References

Behler, John, and F. Wayne King

1979

The Audubon Society Field Guide to North American Reptiles and Amphibians. Chanticleer Press, New York.

Bonnichsen, Robson, and Richard T. Will

1980

Cultural Modification of Bone: The Experimental Approach in Faunal Analysis. In Mammalian Osteology, by B. Miles Gilbert, pp. 7-26. Modern Printing, Wyoming.

Bowen, Joanne

1989

Faunal Remains and New England Urban Household Subsistence. In The Art and Mystery of Historical Archaeology, ed. by Anne Yentsch and Mary C. Beaudry.

1992

"Foodways in the Eighteenth-Century Chesapeake." Paper presented at the 1992 COVA meeting.

Chapman, Joseph A., J. Gregory Hockman, and William R. Edwards

1982

"Cottontails, Sylvilagus floridanus and Allies." In Wild Mammals of North America, ed. by Joseph A. Chapman and George A. Feldhamer, pp 83-123. The Johns Hopkins University Press, Baltimore.

Clason, Annake T., and Wietske Prummel

1977

Collecting, Sieving and Archaeozoological Research. Journal of Archaeological Science 4: 171-175.

Conant, Roger

1975

A Field Guide to Reptiles and Amphibians of Eastern and Central North America. Houghton Mifflin, Boston.

Cruz-Uribe, Kathryn

1988

The Use and Meaning of Species Diversity and Richness in Archaeological Faunas. Journal of Archaeological Science 15:179-196.

Davis, Simon J.M.

1987

The Archaeology of Animals. Yale University Press, New Haven.

Edwards, Andrew C.

1992

Personal communication, Department of Archaeological Research, Colonial Williamsburg, Virginia. 85

Flyger, Vagn, and J. Edwards Gates

1982

"Fox and Gray Squirrels, Sciurus niger, S. carolinenis, and Allies" In Wild Mammals of North America, ed. by Joseph A. Chapman and George A. Feldhamer, pp 209-229. The Johns Hopkins University Press, Baltimore.

Fowler, Dean

1991

Personal communication, Virginia Commonwealth Game and Inland Fisheries Department.

Frethorne, B.

1623

Letter written to his parents. The Records of the Virginia Company London, ed by S. Kingsbury. Washington, D.C., 1906-1935, IV:58-59,62.

Gardner, Alfred L.

1982

"Virginia Opossum, Didelphis virginiana." In Wild Mammals of North America, ed. by Joseph A. Chapman and George A. Feldhamer, pp 3-36. The Johns Hopkins University Press, Baltimore.

Grayson, Donald

1979

"On the Quantification of Vertebrate Archaeofaunas." In Advances in Archaeological Methods and Theory, Volume 2, ed. by M. Schiffer, pp. 199-233. Academic Press, New York.

1981

The Effects of Sample Size on Some Derived Measures in Vertebrate Faunal Analysis. Journal of Archaeological Science 8: 77-88.

1984

Quantitative Zooarchaeology. Academic Press, New York.

Lippson, Alice Jane, and Robert L. Lippson

1984

Life in the Chesapeake Bay. The Johns Hopkins University Press, Baltimore.

Lyman, R. Lee

1982

"Archaeofaunas and Subsistence Studies." In Advances in Archaeological Method and Theory, Volume 5, ed. by M. Schiffer, pp. 331-393. Academic Press, New York.

Miller, Henry

1984

Colonization and Subsistence Change on the 17th Century Chesapeake Frontier. Ph.D. dissertation, Michigan State University, Lansing.

1986

Transforming a "Splendid and Delightsome Land": Colonists and Ecological Change in the Chesapeake 1607-1820. Journal of the Washington Academy of Science 3: 173-187.

Mrozowski, Stephen

1992

The Results of an Archaeobotanical Analysis of Samples from CG8, Williamsburg, Virginia. Manuscript on file, Department of Archaeological Research. Colonial Williamsburg Foundation, Williamsburg, Virginia.

National Geographic Society

1987

Field Guide to the Birds of North America. National Geographic Society, Washington, DC. 86

Reitz, Elizabeth

1979

Spanish and British Subsistence Strategies at St. Augustine, Florida and Frederica Georgia Between 1565 and 1783. Ph.D dissertation, University of Florida, Gainesville. University Microfilms, Ann Arbor.

Reitz, Elizabeth, and Dan Cordier

1983

"Use of Allometry in Zooarchaeological Analysis." In Animals and Archaeology: 2. Shell Middens, Fishes and Birds, ed. by C. Grigson and J. Clutton-Brock, pp 237-252. B.A.R Series, London.

Reitz, Elizabeth J., Irvy R. Quitmyer, H. Stephen Hale, Sylvia J. Scudder, and Elizabeth S. Wing

1987

Application of Allometry to Zooarchaeology. American Antiquity 52: 304-317.

Shaffer, Brian

1992

Quarter-Inch Screening: Understanding Biases in Recovery of Vertebrate Faunal Remains. American Antiquity 57:129-136.

Thomas, David H.

1969

Great Basin Hunting Patterns: A Quantitative Method for Treating Faunal Remains. American Antiquity 34: 392-401.

von den Driesch, Angela

1976

A Guide to the Measurement of Animal Bones from Archaeological Sites. Bulletin No. 1 Peabody Museum of Archaeology and Ethnology, Harvard University, Cambridge.

Wing, Elizabeth, and Antoinette Brown

1979

Paleonutrition Method and Theory in Prehistoric Foodways. Academic Press, New York.

Note: NISP=number of identified specimens; MNI=Minimum Number of Individuals (Ad=adult size, IM =immature size).