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

Methods

A grid system of adjacent squares was used for horizontal control. The intersection of the primary axes was placed near the center of the site to avoid the use of four-figure designations. The grid was to have been oriented with the cardinal directions but, due to an error, the system was actually rotated 18° east of north. This was not discovered until a number of sacks and photographs had already been labeled. A changeover would have been unnecessarily confusing, and we continued to use the north-south, east-west designations.

Vertical control was exercised by reference to elevations of assumed datum points at each mound and near each excavation in the putative habitation area.

Excavation strategy was based on two concepts. First, it seemed probable that the mounds had been erected primarily to cover submound facilities. Surface collections made in the course of the Natchez Trace Parkway Archeological Survey (Jennings, 1940) suggested the site's relationship to the Bynum Site, where submound features were the rule. Secondly, the large size of the mounds made it likely that power equipment would be required at some stage if we were to accomplish more than limited explorations of these features.

Mound H was selected for our initial work. Cultivation had already removed most of the fill, making it possible to test the "submound concept" with a minimum of effort. Subsequent work was planned at Mound A, one of the smallest undisturbed mounds, to check the significance of the fill. Some caution was indicated until we had hard information on the nature of the mounds, and both were tested by manually dug trenches.

The work at Mounds H and A did indeed uncover submound features but, in the latter instance, evidence of a core structure also turned up. Subsequent excavation thus had to combine hand digging and power equipment, hopefully to provide a balance between volume of information and completeness of data.

Mounds D and E were chosen for further excavation. A 15-foot wide trench was manually dug into each mound to test for any structural data which might be contained in the fill. At Mound D, the bulldozer then stripped all but the bottommost 2 feet of fill from an area 60 feet wide all the way across the mound. At Mound E, the cut was 80 feet wide and, since such a small portion of the mound had been involved in the test trench, a 3-foot-wide balk was left down the center to provide a complete profile. At both mounds, the remaining 2 feet of fill and a shallow portion of the native soil were then removed by hand digging.

The approach-trench method of mound excavation was used at Mounds H and A, and in the test trenches at Mounds D and E. Work began with a 5-foot-wide trench on one side and the digging was carried toward the center of the structure. Fill was peeled from the vertical wall of the excavation in thin slices, and profiles were recorded every 5 feet. About 1.5 to 2 feet of the underlying native soil was included in the advancing profile to insure that no submound features were overlooked. This was checked further by cleaning the floor of the excavation after completing each 5-foot segment.

Two modifications in the technique were necessary because of the large size of the mounds. First, the manual excavations did not encompass the entire width of the mound and, second, in the case of three of the mounds, the test trenches were "stepped" due to the height of the profile.

In addition to the work on the mounds, a series of 3-foot-wide trenches were dug to test the "habitation area" along the western side of the site. Over most of the area, material was confined to the plow zone, and, accordingly, the trenches were quite shallow. Advantage was taken of a small area west of Mound H, where the deposit was slightly deeper (1.5 feet), to carry out a series of stratigraphic tests. Here, a number of 5-foot squares were excavated in one-half-foot levels, and the soil was screened.

Mound H (fig. 3)

Continued cultivation had reduced this earthwork to a low rise, 1.5 to 2 feet high, which measured 110 feet wide by 130 feet long.

The mound fill which remained was a brown sandy loam. It closely resembled the site's present topsoil in texture and composition but was slightly darker in color. The native soil beneath the mound was a dark, humus-stained clay which faded into the underlying reddish-brown clay.

Excavation began near the north edge of the mound so that as the work progressed toward the center the profile would be shaded and protected from rapid drying. A 20-foot-wide trench was dug a distance of 60 feet, to a point just beyond the center of the mound. The paucity of information obtained discouraged further excavation.

Feature 5, a fired basin in the old ground surface beneath the mound, was the only significant feature uncovered in Mound H. It was rectangular, measuring 3.7 feet wide by 5.2 feet long, and was 0.7 foot deep. The bottom and sides of this depression and an area of the adjacent mound base had been baked brick-hard and was bright orange in color. The basin had been cleaned of fire debris, leaving only a thin lens of charcoal-stained soil in the bottom, and had been filled with yellow clay.

A greenstone platform pipe (fig. 22) was found on the old ground surface at the north edge of Feature 5. A number of postholes were scattered around Feature 5 but no pattern was discernible (fig. 3).

Two other features were observed, both of undetermined significance. The first was a small burned patch, 1.5 feet in diameter, on the old ground surface located on the S195 line. The second was a wedge-shaped deposit of yellow clay uncovered south of Feature 5. Only a portion of it was exposed by the excavations, but it appeared to be oval in outline and had steeply sloping sides. It was 8 feet wide and 15 feet of its length extended into the trench. The upper part of the feature had been plowed away.

No intact skeletal remains were found at Mound H (or in any of the other mounds). A number of small bits of decayed bone, which might represent the remains of inhumations, were found on the old ground surface beneath the mound.

It should also be noted that several tiny flakes of mica were found on the old ground surface beneath the mound.

FIGURE 3.—Mound H.

Mound A (fig. 4)

This mound was 7 feet high and approximately 60 feet in diameter. Cultivation had sliced a small portion from the southern edge but this was the only damage. We began excavation on the south side because there were fewer tree stumps here, and a 25-foot-wide trench was dug 30 feet into the mound.

Mound A was built largely of brown silty clay which was similar to the native clay of the area. The fill was generally homogeneous in appearance, although the general direction of the loading could occasionally be seen. The only feature in the fill was 0.6-foot-thick layer of yellow clay on the old ground surface at the west side of the large submound grave (Feature 7) with a dome-shaped shell of the same material in the fill above it. A thin wedge, 0.2 to 0.4 foot thick, of water-sorted soil was found on the flanks of this dome, indicating that it had been exposed briefly to weathering.

The significance of the above structure was not recognized while excavations were being carried out. I regarded the mound as a single-stage earthwork built to cover the submound grave. The upper layer of yellow clay, which was matched by sloping load lines on the opposite side of the grave, was interpreted as indicating that mound construction was begun by heaping the first loads of earth around the sides of the grave. A subsequent look at the profiles showed, however, that the grave had been dug through the water-sorted material on the edge of the dome. This is an obvious indication that the dome was built before the grave was dug, and puts a new light on the whole matter. The dome of yellow clay would seem to represent the capped surface of a primary mound built to cover the low platform, or prepared mound base, of yellow clay on the old ground surface. The function of the primary mound was not determined, since no burials or other features were uncovered in the small portion of it exposed in our trench. It may be that the yellow clay base would have yielded mortuary facilities (a similar situation was to be later discovered at Mound D) but this, of course, is merely conjecture.

After a brief period--and the thinness of the eroded material on the flank of the mound suggests that it was brief--the grave was dug into the ground at the side of the primary mound and covered by an addition to the earlier structure.

Mound A features included a single submound grave, a pit intrusive in the top of the mound, and two cremations in the mound fill.

Feature 7 was a rectangular pit which measured 6.5 feet wide by 11 feet long. The flat bottom of the feature was 2.2 feet below the old ground surface. Three thin lenses consisting of bits of calcined bone mixed with charcoal-stained soil, presumably cremations, lay on the floor of the pit. A small, shapeless fragment of wood, covered on one side with a film of copper salts, lay near one of the cremations and was the only artifact. Three postholes were located on the margin of the pit, but no other structural features were found. There was no evidence of fire or burning within the pit.

Feature 9, a 3.5-foot-square pit, had been dug from the surface of the mound 4 feet into the fill. No artifacts or human remains were found in it, and it may well have been a looter's pit.

Two cremations, both represented by a thin lens of bits of calcined and charred bone, were uncovered in the mound fill. The cremation near the east side of the trench (fig. 4) was two feet above the mound base, and the one north of Feature 7 was five feet above the mound base. They had both been included in the mound as it was built.

FIGURE 4.—Mound A.

Mound D (fig. 5)

This was one of the larger of the earthworks, being about 15 feet high and approximately 90 feet in mean diameter. It had not been damaged by cultivation.

A 15-foot-wide test trench was opened on the western side of the mound, where fewer tree stumps existed. No artifacts or evidence of burials were found in the upper fill in the test trench, but there were indications that Mound D was not a homogeneous heap of earth. We felt that enough information had been obtained here to give a good idea of the mound's composition, and we went ahead with plans to use the bulldozer. An area 60 feet wide completely across the mound was stripped of all but the bottommost 2 feet of fill. Subsequent excavation in the stripped area turned up very meager results, and work was halted well short of the eastern side of the mound.

The profiles and plans recorded in the test trench and observations in the later work showed that Mound D had three components. The initial phase in the mound's construction was preparation of the base. Humus had been scraped from a large oval area measuring approximately 35 by 45 feet, and had been replaced by a 0.5-foot-thick layer of mixed topsoil, yellow native clay, and reddish clay. Presumed mortuary facilities, described below as Features 14 and 15, were then placed on this base. Subsequently, a 10-foot-high dome consisting of basket loads of brown loam topsoil and native clay had been heaped over the base, covering its associated features. In the portion of the mound which was excavated, the perimeter of the prepared base coincided with its covering dome, and we assume that this relationship prevailed throughout.

Shortly after completion of the primary mound, at least one burial was laid on the old ground surface nearby. A mantle of pure brown loam topsoil was then added, completely covering the primary mound. No erosional material flanked the primary mound, indicating that no significant time elapsed between the building of the primary mound and the addition of the topsoil mantle.

Feature 14 was a burned area of the prepared mound base. The hard, orange-colored surface which defined the feature was rectangular and measured 3 by 4.5 feet. Near this area were scattered patches of ash and of charcoal staining, perhaps the results of cleaning of the fire area. No ash or charcoal was found directly associated with this feature.

Feature 15 was an oval area, 4 by 6 feet, paved with flat sandstone slabs laid on the prepared mound base. A broken copper spool and a few bone fragments, which has been preserved by the copper salts, were found on the paving.

The other burial, on the original ground surface just beyond the edge of the prepared mound base, consisted of a few crumbling skull fragments accompanied by two copper spools.

FIGURE 5.—Mound D.

Mound E (figs. 6 and 7)

This mound, 8 feet high with a mean diameter of 170 feet, seemed to be a low earthwork relative to its areal dimensions, and we supposed that it had been much reduced by plowing. Local people, however, stated that the mound had been cultivated only recently and that its shape had been changed very little. The profiles in the preliminary test trench tended to bear out this information.

The fill was composed of thoroughly mixed brown loam topsoil and native clay. Despite the general homogeneity, the trend and direction of the loading were discernible.

A test trench was begun on the eastern flank of the mound and extended 40 feet into it. This preliminary work suggested that Mound E was a one-stage structure. However, subsequent excavations, carried out after the bulldozer had stripped the mound, showed it to be more complex. Although some details were lost in the stripping, much of the information was salvaged from the remaining two feet of fill and from the central balk.

Mound E was constructed in two and possibly three stages. The primary mound was heaped over a low burial platform and was itself completely covered by a secondary addition. A third period of use was indicated by pits which penetrated the fill of the secondary mound.

The central feature of the primary mound was a 1-foot-high platform. Its overall dimensions and shape could not be accurately determined because, except for a few feet, the eastern perimeter of the platform was not traced. However, it measured approximately 80 feet east-west, and the configuration of the western edge suggested a round or oval ground plan.

The surface of the platform was marked in the test trench profile by a thin, tenuous band of dark soil. The significance of this 0.2-foot-thick horizon was not recognized while the preliminary tests were underway since no features or artifacts were associated with it, and since the fill above and below it were identical in composition. Later work, after all but 1 foot of fill above the platform had been removed, uncovered a shallow fired basin, a number of artifacts, and a few scraps of decayed bone lying directly on this layer. The platform was obviously a prepared burial repository. As no intact skeletal remains were found, the number of interments, and whether they had been placed simultaneously or over a period of time, could not be determined.

The burial platform, as mentioned above, was covered by the primary mound. The crown of this low dome had been obliterated and its height at the center could not be determined. Its surface was indicated by a thin layer of gravelly, sandy soil. Deposits of water-sorted material on the old ground surface at the edge of the primary mound indicate the structure was exposed to weathering before being covered by the secondary mound.

On the east side, the primary mound extended a few feet beyond the edge of the burial platform. Here, the edge of the primary mound covered a probable crematory pit and two graves dug into the old ground surface. These excavations had been backfilled with primary mound fill and were plainly associated with the first construction stage.

The second stage of construction is represented by two graves dug into the primary mound and the mantle of fill which covered them and blanketed the earlier stage.

A third period of mound use is indicated by Feature 22, a grave which penetrated the secondary mound. We were unable to tell whether or not this represented a period of mound building. The outlines of this feature's intrusion could be traced to the present surface but, with the disturbance of the mound by cultivation, its surface of origin could not be determined. It should be noted that there were other layers of sand and gravel in the profile in addition to the one identifying the surface of the primary mound (fig. 6, dashed lines). None of the others were associated with erosional material or features, however, and were apparently not of structural significance.

The features uncovered in Mound E are described individually below and have been grouped according to their association.

Primary mound features included Feature 17, associated with the platform, and Features 11, 12, and 16, which lay beyond the platform's edge.

Feature 17 was a shallow, fired depression in the surface of the platform. Aside from a thin patch of charcoal-stained soil on the adjacent surface of the platform, there was no fire refuse. The feature was circular, measuring 5 feet in diameter and 0.2 foot deep. We assume that it was a crematory pit. A Marksville Incised vessel (fig. 12a) was found resting on the platform at the edge of the feature.

The artifacts recovered from the surface of the platform were the silver plating (fig. 20), possibly from a set of conjoined tubes; an untyped, sand tempered, zone stamped vessel (fig. 13c); a Baldwin Plain bowl (fig. 13a); and a rectangular slab of wood covered with sheet copper. As mentioned previously, occasional bits of decayed bone were also found on the platform.

Feature 11 was another possible crematory pit. An intense fire had burned the bottom and sides of the feature and left a 0.2-foot thick layer of wood charcoal. The feature was straight-sided with rounded ends and measured 4.3 by 2.2 feet. Basin-like in cross section, it was 1.4 feet deep. No human remains or artifacts were associated with this feature.

Feature 12 was a rectangular pit which measured 8.1 by 4.2 feet and was 2 feet deep. A Flint River Cordmarked pot (fig. 12b) had been placed in the northeast corner of the pit. No skeletal remains were in evidence.

Feature 16 was the third feature uncovered beneath the edge of the primary mound. It was a rectangular grave measuring 16 by 12 feet and was 0.9 foot deep. Three deposits consisting of bits of calcined bone and charcoal-stained soil indicate that the feature was used as a depository for cremations. There were no signs of burning within the pit itself.

Features 10 and 13 are associated with the second stage, and are presumed graves though no skeletal remains were found in them.

Feature 10 was a 3-by 7-foot rectangle which penetrated 2.5 feet beneath the surface of the primary mound. Two vessels were found on the pit floor, one piled on top of the other; a Flint River Brushed jar (fig. 12c) and an unnamed, sand tempered, zone stamped vessel (fig. 13d).

Feature 13 was a rectangular pit, approximately 9 feet deep. It measured 10.1 by 7.6 feet. There were no artifacts in it.

The third-stage grave, Feature 22, was an 8-by 4-foot rectangle. Its bottom lay 10 feet below the existing mound surface, but its original depth could not be determined due to the truncation of the mound. A large, unworked sheet of mica, a cartridge-shaped labret of greenstone, a sandstone grinding pallet, and a large lump of galena were found in the fill of the pit, 1.5 feet above the floor. No human remains were found in Feature 22.

The association of two features, 18 and 19, could not be determined, since they were located in the middle of the stripped area and did not intersect a profile.

Feature 18 was a rectangular pit, 5 by 10 feet. Two small lumps of galena and a cache of small points and blades were found in the fill of the pit, about 1 foot from the bottom. No skeletal remains were discovered in this feature.

Feature 19 measured 11.7 by 11 feet. Semicylindrical depressions along three sides of the pit probably represent the casts of decayed logs. The logs had laid on, or slightly in, the burial platform and this suggests a probable association of Feature 19 with that structure. No bones or grave goods were uncovered.

The nature of the old ground surface beneath Mound E requires description. The old humus was light gray in color over most of the mound base, but there were several black midden-stained areas. There were also a number of small burned spots on the old surface. Pottery was plentiful in the old humus here, when compared to other areas of the site where it was generally quite scanty. A number of postholes and Feature 21, a 1.3-foot-deep oval pit filled with midden-stained soil, were also associated with the submound horizon. This evidence suggests that Mound E had been built over an occupation area which had been protected from obliteration by the mound itself. It is noteworthy that the site surface around and near Mound E produced no evidence of habitation or use.

FIGURE 6.—Mound E.

FIGURE 7.—Schematic Profile of Mound E.

Habitation Area

The Pharr Site was originally described as having an extensive village site in the area between the mounds (Jennings, 1941, p. 211). However, no indications of habitation were observed in the freshly plowed fields of this portion of the site. A scant handful of flint chips was the sole product of our surface reconnaissance. It seems remarkable that cultivation could have so thoroughly erased traces of the village. Nevertheless, the situation described above for Mound E suggests that this is probably what happened. Improvements in farming techniques quite possibly could have destroyed what less progressive methods had left relatively undisturbed as late as 1941 when the site was recorded.

The only area which showed surface indication of habitation was along the western edge of the site. In addition to the thickly scattered flint chips, a variety of projectile points and miscellaneous chipped stone artifacts were recovered from the surface here (figs. 14, 15, and 17). Included were several forms of projectile points similar to Archaic types (fig. 15h-k and n). Regrettably, they were not submitted for expert identification.

Three locations along the western side of the site were tested (fig. 2). Only one location, southwest of Mound H, produced features or subsurface artifacts in significant quantity. This was an area of black, midden-stained soil which excavation showed to be 1 foot deep. Aside from three shallow depressions in the native clay, 1.2 to 2.8 feet deep and filled with midden-stained soil, only two features of note were uncovered.

Feature 3 was a circular pit, 3.5 feet in diameter and 2 feet deep. It was filled with the rich black soil and is noteworthy because it contained a large lump of burned daub, the only piece found at the site, a few bits of calcined bone, and over 100 sherds of shell tempered pottery. This is virtually the only shell tempered pottery from the site.

Feature 6 was a trash pit. It was a roughly circular depression, 15 feet in diameter and 1.3 feet deep. The pit contained an extraordinary amount of pottery. Only one-quarter of the pit was excavated and 470 sherds were recovered from it.

The results of the stratigraphic tests carried out in this area were not illuminating. Most of the 1 foot of fill had, of course, been churned by plowing and the percentage distribution of types did not vary significantly from top to bottom (table 4).

The excavations and surface finds in the area of habitation failed to produce conclusive results, and we can only theorize about the nature of the several occupations.

Most of the large quantity of flint scrap was plainly the byproduct of tool making. This indicates a chipping station which may predate the mounds.

It seems reasonable that the midden-stained soil beneath Mound E and southwest of Mound H was the product of an occupation contemporaneous with the mounds. Most of the pottery recovered from these areas dates from this occupation, and Feature 6 seems to have been used as a trash pit during this period. At least one feature, Feature 3, is later, but since virtually no other shell tempered pottery was recovered outside of this pit, it may be viewed as an isolated incident.

The black greasy soil would seem to indicate an intensive, continuing occupation but, aside from a single lump of daub and a few isolated postholes, no structural features were uncovered. There are many reasons which might account for this. Perhaps only the crudest of shelters were built. Even a relatively sturdy wattle-and-daub structure need not have left surviving remains. If the structure had simply been abandoned, without having been burned, no fired daub would have been produced and the decomposing wall posts could readily be absorbed without trace in the rich brown loam.

No evidence was recovered to indicate the settlement pattern, but it seems likely that the area was used intermittently, perhaps during burial rites and mound building, or in connection with seasonal harvests, or both.