NATCHEZ TRACE PARKWAY Excavations at the Pharr Mounds and the Bear Creek Site

Methods

A standard grid system was utilized with the axes rotated 30 degrees east of north to correspond with what was thought to be the orientation of the mound. For the sake of convenience, however, loci on the grid were designated as being south and east of the intersection of the base lines. The 0-0 coordinate was placed northwest of the mound so that the entire site would be in one quadrant. The mound was laid out in 10-foot squares, and each square was designated by the location of its northwest corner.

When a topographic map was made, it was found that the mound was oriented with the cardinal directions. It was decided, nevertheless, to utilize the grid as established despite the inconveniences entailed.

The first excavations took place in a series of 10-foot squares just north of the mound. They were dug to provide preliminary training for the laborers and to test the area as a possible dump for the back dirt from the excavation of the mound. (It was later decided to dump the back dirt over the bank into the creek bottom.) The base of a fluted projectile point and a few flint chips were found in the plow-disturbed soil near the surface, but otherwise the tests were sterile.

Although the mound had been greatly reduced, there was still the chance that some information had survived. Accordingly, a 5-foot-wide test trench was begun on the western side of the mound and carried in successive 10-foot segments across the width of the structure. The fill was removed in horizontal, 6-inch layers, and digging was carried 1.5-2.0 feet into the undisturbed native clay beneath the fill.

No postholes, house midden, or other features suggestive of structural remains were located by the trench. However, the profiles showed the presence of two level and continuous horizons which offered further possibilities. To insure that nothing was overlooked, the surfaces of these layers were uncovered in a number of squares adjacent to the test trench. These broader excavations also failed to turn up any significant structural features.

With the possibilities of the mound exhausted, we turned to the village area. To obtain some idea of the magnitude of the remains, a series of 5-foot-wide trenches, spaced at 50-foot intervals, were dug across the village area (fig. 2). It soon became evident that no midden deposits remained. The artifacts, relatively few in number, all came from the plow zone. Only the area immediately to the southeast of the mound was productive. Trench E250 and the northerly portion of trenches E200 and E150 intersected a number of trash pits and postholes. A fair number of sherds were also recovered here. West and south of this area the features rapidly thinned out and the trenches were all but barren of material.

When the preliminary testing was completed, the next task was to remove the plow-disturbed soil from the more promising areas revealed by the trenches in the hope of uncovering house patterns and other features. Since extensive areas were involved, a bulldozer seemed to be the most economical and rapid means of accomplishing this. A Case Model 750 bulldozer was used to back-drag the disturbed soil from three large areas. Two house patterns, numerous trash pits, two burials, and many scattered postholes were uncovered.

The meager results of the excavations were disappointing and I gave some thought to recommending the abandonment of the site as an interpretive feature. Upon reflection, however, it assumed an importance out of proportion to its scientific value. The excavations showed that the site manifested an unusually long occupation, as will be discussed later, and this in itself would make an interesting interpretive story. The site is, furthermore, the only known Middle Mississippian manifestation at the northern end of the Parkway and will offer the sole opportunity to interpret this aspect of parkway archeology to many visitors.

The mound in its damaged condition was obviously unsuitable as an in-place exhibit, and we decided to reconstruct it.

To accomplish this, a series of narrow trenches was dug into the flanks of the mound to determine its original perimeter. It was then a matter of staking the sides and bulldozing the surrounding disturbed soil back on top of the mound. The reconstruction is, of course, conjectural in part. While the perimeter is accurate, the slope of the sides and the height are estimates. The mound, as reconstructed, is 8 to 10 feet high--somewhat less than the original height recorded by the survey. This was, presumably, an estimate, and a survey photograph (fig. 7) suggests that it may have been in error.

The Mound

The excavations showed that the mound had been roughly square in outline and had measured 85 to 90 feet on a side (fig. 3). It was constructed of a yellowish-brown, silty loam in which the individual, peck-size loads were visible.

FIGURE 3.—Plan and cross section of mound

A number of stages of construction were seen in the test trench profiles. These were indicated by interruptions in the otherwise continuous basket loading and, in one case, a horizontal layer of red clay.

Our interpretation of the building sequence is illustrated in figure 4. Stage A, the earliest, was a low, flat-topped earthwork, capped with a layer of red clay. Its platform was square, measuring about 30 feet, with sides which roughly paralleled the sides of the final stage of the mound. A number of features were associated with Stage A. Five post holes were scattered from the center to the western edge of the platform. A circular pit, 3 feet in diameter and 1.3 feet deep, was also located near the western edge. It had been dug from the surface of the clay cap and contained a small, flat slab of sandstone. The pit was otherwise sterile. A small patch of the surface of the platform near the southwestern corner had been burned and was bright orange in color and brick-hard.

FIGURE 4.—Stages of mound construction (schematic).

Feature 1, the large circular pit in the center of the Stage A surface, seems to date from a later stage and will be discussed subsequently.

There were two shallow depressions in the native clay beneath Stage A. Neither contained artifacts or burials. They may have been small borrow pits where, before construction of the mound began, clay was obtained to plaster house walls. One depression was located directly beneath Feature 1 but seems to have had no connection with it. The loaded soil above it extended unbroken and undisturbed between the bottom of Feature 1 and the depression.

After Stage A had been completed, the mound was increased by adding fill to the eastern and southern sides, thus enlarging the already existing platform. No features were noted on the remnant of Stage B surface uncovered in the excavations. Its shape and size were not determined.

Stage C extended the mound a few feet farther to the east and covered the surface of Stage B and a portion of the surface of Stage A, raising a part of the preceding platform about a foot. Only a small portion of its surface survived and, aside from Feature 1, nothing of note was uncovered. Feature 1 was apparently dug after Stage C had been completed, since its eastern side is adjacent to the edge of the platform of this phase. Alternatively, when Stage C was built, it may have extended only to the edge of the pit, in which case Feature 1 was earlier. Feature 1 was a circular pit, 12 to 14 feet in diameter, and had been dug some 4 feet below the surface of Stage C. It was filled to a depth of 2 feet with water-sorted soil which had obviously washed in from the pit's sides. No skeletal material or significant artifacts were found in the feature, and its function was not determined.

The final construction phase, designated Stage D, is represented by a narrow wedge of fill covering the eastern flank of Stage C and the fill overlying the preceding platform and eastern side of Stage A. This final mantle completely covered the surface and sides of the older earthworks.

The surface of the final stage was obliterated by cultivation. Most of the disturbed soil on the flanks of the mound was from this stage, and its contents provide the only clues to its use. A considerable lens of burned daub (the mud plaster used in house construction) was uncovered on the western flank of the mound and indicates that a house or temple had been built on the final stage.

A number of burials were made in Stage D. Two pits, both intrusive from the fill of this stage, were noted in the test trench profile. Excavation showed both graves to be rectangular, measuring about 6 feet long and 3 feet wide. Feature 2, the pit in the E10 segment of the trench, contained only a few crumbling bone fragments. Feature 3 yielded two restorable vessels and a considerable portion of a third. A few bone fragments were also found near the bottom. Additional evidence of burials in this stage is the splintered and broken human bones found in the disturbed fill on the mound's flanks. They would seem to represent burials destroyed when the mound's summit was pushed off.

One more feature needs to be mentioned. Feature 4 was a deep, trash-filled pit adjacent to the eastern side of the mound. A large quantity of splintered bone, shell, and pottery sherds were removed from it.

The picture that emerges is of a mound which was successively rebuilt in flat-topped stages. The significance of these stages, whether they constituted mere pauses in an essentially continuous building process or were functional structures, could not be determined in most instances. It is certain that the Stage A platform never supported a building--there were too few postholes and no daub or charcoal associated with it. It is equally certain that a portion of this platform in conjunction with the Stage C surface was exposed for a considerable period before the final mantle was added. This is indicated by the depth of washed-in soil in Feature 1. Too little remained of the other surfaces to determine their nature except for Stage D, on which, as already mentioned, a building was constructed and burials were made.

The Village

The two houses, many pits, and numerous postholes which constitute the only village remains uncovered were concentrated in a small area immediately southeast of the mound (fig. 2). This is a considerably smaller area than that defined by the presence of surface material. It may be that the surface material south of the mound, where few features were found, was transported from the mound and strewn over the surface by cultivation. Erosion may have erased some features, but there is no reason to believe that it was more active in this area than to the east, where features survived.

House 1

House 1 was a subrectangular structure, with slightly bulging sides and rounded corners, which measured 22 by 24 feet (figs. 5 and 8). Its walls were supported by posts set in individual holes. The postholes averaged 0.7 foot in diameter and were spaced an average of 1.7 feet apart. They averaged a foot in depth, though doubtless they had been truncated by cultivation. Some of the postholes were oval in outline and were dug with one side slanted to the bottom. This was probably done to allow heavier posts to be slid into place. Two oval-shaped holes on the western side of the house suggest that heavier posts had been placed here, perhaps to to support the door lintel.

Small bits of burned clay, some still bearing the imprint of round twigs or cane, were found in many of the postholes. They indicate that the house walls were wattle-and-daub and that the house had been destroyed by fire.

Since the houses overlapped, it was impossible to assign interior features to one house or the other. The pattern of large postholes within the houses suggests strongly, however, that each roof was supported by four posts.

A shallow, circular hearth, 1.5 feet in diameter, within the house patterns had burned sides and was filled with ash.

House 2

House 2 was rectangular, with rounded corners, and measured 20 X 21 feet. It was oriented very nearly with the cardinal directions. Short, parallel trenches with postholes at either end indicate that an entranceway projected from the western side of the structure (figs. 5 and 8). The size and spacing of the wall posts were the same as in House 1. Burned daub was again found in the postholes, suggesting that House 2 was also destroyed by fire.

Pits

A number of pits of differing sizes and shapes were uncovered in the village area (fig. 8). They varied in depth from less than 1 foot to 2.5 feet. All were filled with black, midden-stained soil but were generally free of debris other than occasional pieces of pottery or bone.

The function of these pits is unknown. It may be that they were borrow pits where clay was taken for plastering the houses.

Burials
(figs. 2 and 6)

Burial 1 was located during the excavation of the mound but should properly be described with the village area. It lay 20 feet west of the mound in a shallow grave, only 1.5 feet below the old ground surface. Much of the skeleton had been absorbed and only the skull and segments of a few of the long bones could be traced. The body apparently lay on its back, with the legs flexed and drawn up to the left side. No artifacts were associated with it.

Burial 2 was uncovered in the E250 trench, some 20 feet north of S300, where it lay only 0.9 foot below the present ground surface. This was the primary interment of an adult. The body lay on its left side with the arms folded on the chest and the legs flexed. The bones were quite soft and crumbly but easily traced. There were no grave goods.

Burial 3 was located approximately 25 feet north of S300E150. It lay 0.4 to 0.8 foot below the surface and had been partially disturbed. The bones were in an advanced state of decay and were difficult to trace, but a flexed inhumation was suggested. Again, there were no grave goods.

Burial 4 was uncovered in the large stripped area east of trench E200. Like the others, it lay in a shallow grave, only 1.2 feet deep. This was the primary interment of an adult who had been placed in a supine position. The bones were readily traceable but extremely crumbly and none were removed intact. No artifacts were found with the interment.