All Anderson et al (1996) Wilmington wares are included on this page. He did not offer a separate definition for Hanover. See Herbert (2003) and South (1976).
Sorting criteria: Haphazardly applied parallel brushed or combed impressions over the exterior vessel surface; occasional cross-brushing. Impressions are typically shallow (0.5-1.0 mm) and narrow (1.0-2.0 mm), with striations and smearing common. The paste is characterized by crushed sherds or grog from 3 to 5 mm in maximum dimension, although larger inclusions up to 10.0 mm are sometimes noted. The finish is sometimes observed on the bases of Wilmington Heavy Cord Marked vessels.
Distribution: Poorly documented. The finish is a minority type at the mouth of the Savannah River, and likely occurs with Wilmington assemblages on the southwestern South Carolina coast.
Chronological position: Late Woodland period (ca. A.D. 500-1000). DePratter (1979:130) suggests it may postdate A.D. 600 at the mouth of the Savannah.
Background: The type Wilmington Brushed was formally defined by DePratter (1979:130-131) based on materials in collections from WPA-era excavations at the mouth of the Savannah, and using information from Caldwell and McCann's unpublished Deptford site manuscript. The Wilmington Brushed type from the Georgia coast is characterized by a clay/grog tempered paste, and may be related to the Deptford Brushed type. The type materials are thought to be decorated with bundles of sticks or grass, with tempering consisting of crushed sherds or crushed, low-fired clay fragments (DePratter 1979:129-130) Brushing is reported over the entire body, and is also occasionally noted on the bases of Wilmington Heavy Cord Marked vessels. In the latter cases the brushed appearance may derive from vegetation the pot may have been placed on prior to firing.
The temporal and taxonomic relationships of southeast Atlantic coastal clay/grog tempered wares are currently somewhat ambiguously perceived, and appear to depend as much on geography or absolute dates as on distinctive attributes of the wares themselves. Two major series of clay/grog tempered wares are currently established in the literature for the region (if we view, for the sake of convenience, the Wilmington and St. Catherines series as sequential parts of a local tradition). These are: (1) the Hanover series from coastal North Carolina and northern coastal South Carolina and, (2) the Wilmington/St. Catherines series from central coastal South Carolina south into the sea island area of Georgia. Sherd (clay/grog) tempered ceramics are, therefore, documented throughout most of the area from central coastal North Carolina to the sea islands of Georgia. The northern (Hanover) wares are earlier, and are dominated by fabric impressed surface finish; they are additionally found well into the interior of the coastal plain (e.g., South 1960, Anderson 1975a, Loftfield 1976). The southern wares (Wilmington, St. Catherines), in contrast, are dominated by cord marking, and appear restricted to the coast in the area south of Charleston Harbor, occurring only rarely in the interior (e.g., Caldwell 1952:317; Anderson 1975a:189). The southern wares occur later, although continuity through time and over space is apparent. A number of radiocarbon dates from the central South Carolina coast, in particular, document the length of this tradition, and the temporal overlap between the northern and southern margins (e.g. South 1971; South and Widmer 1976; Dorian and Logan 1979). A north to south movement, or adoption, of this distinctive tempering/manufacturing technology is indicated.
The clay/grog tempered ceramics from the southeastern Atlantic coast thus comprise a distinctive local tradition whose geographic and temporal extent is only now becoming known. The similarities over this area appear to greatly outweigh the differences. While the incidence of specific finishes differs over the area, and assemblages can be sorted, individual sherds typically cannot:
Material from the Savannah River area called Wilmington is generally thicker, sandier, and somewhat more poorly made than material to the north (e.g., Hanover). The variation is slight, however, and can be detected only in assemblages from the northern and southern areas and not from the individual sherds; within these assemblages individual sherd-tempered sherds may be readily substituted in assemblages over the area (Anderson 1975a:189).
Separation of these wares into discrete ceramic series does not make good sense taxonomically. Nowhere is this more evident than in the central coastal area of South Carolina, where it could be argued that a major criteria used to classify wares as either Hanover or Wilmington appears to be the age of associated radiocarbon dates.
For this reason, most post-Refuge/pre-St. Catherines clay/grog tempered ceramic types in the Carolinas are subsumed under the Wilmington series, with variants acknowledged as necessary to accommodate perceived variability in the ware. This would reduce (or at least acknowledge) the ambiguity inherent in attempting to sort the various types now in use (i.e., Hanover Cord Marked from Wilmington Heavy Cord Marked), while simultaneously providing a realistic and accurate method for accommodating the variability that does exist. Such a procedure would greatly streamline local typology (by eliminating redundant ceramic series) and help establish a much needed regional analytical perspective.
Primary references: DePratter 1979:130-131.
WILMINGTON (WALTHOUR) CHECK STAMPED
Sorting criteria: Check and rarely linear check stamping over the exterior vessel surface; occasionally smoothed somewhat after stamping. The paste is characterized by crushed sherds or grog from 3 to 5 mm in maximum dimension, although larger inclusions up to 10.0 mm are sometimes noted.
Distribution: Poorly documented. Check and linear check stamped pottery characterized by sherd or clay/grog tempering appears restricted to the Sea Island area from the mouth of the Savannah River to Charleston Harbor, beyond which it occurs both along the coast and in the interior of the Coastal Plain from the Ashley-Cooper to the Cape Fear/New River drainages in North Carolina. The ware becomes increasingly uncommon from south to north in North Carolina.
Chronological position: Early/Middle Woodland periods (ca. 300 B.C. - A.D. 600). North of Charleston Harbor the finish appears to occur throughout the Middle Woodland period, where it is roughly contemporaneous with Deptford finishes, from ca. 500 B.C. to A.D. 500. On the southeastern coast of South Carolina the ware appears to date later, and only occur at the end of this interval; DePratter (1979:111, 130) places this material (which he typed Walthour Check Stamped) to between A.D. 500 and 600 at the mouth of the Savannah.
Background: This category subsumes earlier types developed for clay/grog tempered check and linear check stamped wares locally, notably Caldwell's (1952:316, see also Waring 1968:220) Wilmington Check Stamped, renamed Walthour Check Stamped by DePratter (1979:130, 1991:176), and Hanover Check Stamped, after the series defined by South (1960, 1973). Walthour Check Stamped, found at the mouth of the Savannah, is assumed to date to the initial part of the Late Woodland, from ca. A.D. 500-600, and is thought to be a direct development from Deptford (DePratter 1979:130). The northern Hanover type is assumed to have been contemporaneous with Deptford, differing only in the selection of paste. A north-to-south movement, or adoption, of clay/grog tempering/manufacturing technology appears indicated by the distribution and dating of assemblages recovered to date. Wilmington materials south of Charleston Harbor appear to postdate ca. A.D. 500, while those to the north (formerly called Hanover) occur much earlier, and are in fact replaced after ca. A.D. 500 by other series.
Wilmington Check Stamped appears to be relatively uncommon in the coastal plain of South Carolina; a survey of ceramics on 313 sites recorded only 49 clay/grog tempered check and linear check stamped sherds, on 17 sites (Anderson 1975b). A total of 15 sherds of Wilmington Check Stamped were recovered at Mattassee Lake. Except for the exterior surface finish, the ware was virtually identical to the fabric impressed clay/grog tempered ware from the site, Wilmington Fabric Impressed. Both check and linear check stamped finishes were present, although most (N=11, 73.3 percent) exhibited linear check stamping.
Primary references: Caldwell (1952, 1971); Waring 1968c, Williams 1968; Anderson 1975b; DePratter (1979:130); Anderson et al. (1982:276).
WILMINGTON (WALTHOUR) COMPLICATED STAMPED
Sorting Criteria: Complicated stamping characterized by concentric circles, figure eights, and other designs over the exterior vessel surface. The paste is characterized by crushed sherds or grog from 3 to 5 mm in maximum dimension, although larger inclusions up to 10.0 mm are sometimes noted.
Distribution: Poorly documented. Complicated stamped pottery characterized by sherd or clay/grog tempering appears largely restricted to the mouth of the Savannah River.
Chronological position: Late Woodland period (ca. A.D. 500 - 600).
Background: This category subsumes earlier types developed for clay/grog tempered complicated stamped ceramics, notably Wilmington Complicated Stamped (Caldwell 1952:316, see also Waring 1968:220), renamed Walthour Complicated Stamped by DePratter (1979:130, 1991:176). Walthour Complicated Stamped, found at the mouth of the Savannah, is assumed to date to the initial part of the Late Woodland, from ca. A.D. 500-600, and appears related to Deptford Complicated Stamped, which it closely resembles save for the paste differences (c.f., DePratter 1979:7, 126, 130). Excavations conducted on Wamassee Neck on St. Catherines Island by Caldwell in 1969 and 1970 demonstrated that the type appeared at the very end of the Deptford phase (DePratter 1991:7). The stamp designs strongly suggest that the type (and the related Deptford Complicated Stamped) is a local variant of Swift Creek Complicated Stamped.
Primary References: DePratter (1979:130, 1992:176); Caldwell 1952:316; Waring 1968:220.
WILMINGTON CORD MARKED
Sorting criteria: Cord impressions stamped over the exterior vessel surface when the paste was wet. The paste is characterized by crushed sherds or grog from 3 to 5 mm in maximum dimension, although larger inclusions up to 10.0 mm are sometimes noted. Occasionally materials are tempered with smaller (0.5-2.0 mm) lumps of aplastic clay (grog).
Distribution: Wilmington Cord Marked occurs in the Sea Islands of Georgia and South Carolina north to Charleston Harbor. From Charleston Harbor to the Pamlico River, it is found both along the coast and in the interior to the Fall Line.
Chronological position: North of Charleston Harbor: Early/Middle Woodland periods (500 B.C. - A.D. 500). South of Charleston Harbor: A.D. 500-1000.
Background: The type Wilmington Heavy Cord Marked was initially defined by Caldwell and Waring (1939) based on WPA-era excavations at a number of sites at the mouth of the Savannah River. Originally defined by a single type, Wilmington Heavy Cord Marked, and encompassing both sherd and grit tempering, the series has come to include a range of types, all characterized by clay/grog tempering. The type has been described by DePratter (1979:129, 1991:177), who in his most recent formulation dropped the word “Heavy” from the name. This seems particularly appropriate, since the word refers to the thick, typically parallel stamped cord impressions that are commonly found on Wilmington cord marked pottery along the lower Savannah, yet are uncommon away from this area.
Clay/grog or sherd tempered pottery characterized by cord marked and fabric impressed finish occurs widely in the Coastal Plain of both North and South Carolina, and two major series have been traditionally used to encompass this variation. These are Wilmington, defined from work conducted at the mouth of the Savannah during the late 1930s, and Hanover, defined by South (1960) based on survey work in southern Coastal North Carolina in 1960. In recent years the temporal and spatial distribution of these two series have run together, creating considerable taxonomic confusion. For this reason, use of Wilmington terminology is adopted here for all post-initial Early Woodland Refuge, pre-initial Mississippian St. Catherines clay/grog tempered pottery in the Carolinas.
The clay/grog tempered pottery at the mouth of the Savannah has been used to define three phases, Wilmington I, Wilmington II, and St. Catherines (DePratter 1979:111; 1991:11). The first, Wilmington I/Walthour, dates from A.D. 500 to 600, and is characterized by clay/grog-tempered Wilmington Check stamped, Heavy Cord Marked, Plain, and Walthour Complicated Stamped ceramics, the latter an apparent late Swift Creek variant. Wilmington II phase components, which dates from ca. A.D. 600 to 1000, are identified by the presence of Wilmington Plain, Brushed, Fabric Marked, and Heavy Cord Marked pottery. The St. Catherines phase, which dates from ca. A.D. 1000 to 1150, is characterized by the St. Catherine Plain, St. Catherines Burnished Plain, St. Catherines Fine Cord-Marked, and St. Catherines Net-Marked types.
Wilmington and St. Catherines wares, however, are both clay/grog or sherd tempered, and are differentiated primarily by temper size and quality of manufacture:
St. Catherines phase ceramics are characterized by finer clay tempering than that of preceding Wilmington types and by the increased care with which the ceramics were finished. The lumpy, contorted surface of Wilmington types was replaced by carefully smoothed and often burnished interiors and exteriors. St. Catherines Burnished is characterized by careful exterior burnishing, whereas surfaces of St. Catherines Plain are simply smoothed. St. Catherines Fine Cord Marked has more carefully applied and more consistently spaced crossed cord impressions than did its predecessor, Wilmington Heavy Cord Marked. A new type, St. Catherines Net Marked, is also included in the St. Catherines series, but it is rare at most sites (DePratter 1979:119).
A number of radiocarbon dates for Wilmington/St. Catherines components from the Georgia and South Carolina Sea Island area support the posited time range of roughly A.D. 500 to A.D. 1150 for these wares, and stratigraphically the materials are clearly post-Deptford on the north Georgia coast (Waring 1968c; Caldwell 1971; DePratter 1979; Trinkley 1980a, 1981a). The decision to retain separate Wilmington and St. Catherines series in the present taxonomy was based on the differences in paste and surface finish between the two series. Given the potential for overlap, however, it is possible that some or all of the St. Catherines types may ultimately need to be subsumed within the Wilmington series.
A second clay/grog or sherd tempered ware, the Hanover series, was reported by Stanley South in 1960, based on materials collected from predominantly coastal shell midden sites in southeastern North Carolina and northeastern South Carolina. The paste was described as:
Tempered with large lumps of aplastic clay. The majority of these tempering lumps appear to be crushed sherds. The smoothed interior of the original sherd can be frequently seen on some of the crushed tempering fragments. These large lumps of temper result in a rough, lumpy surface on the interior of the sherd, around which a series of small cracks are frequently seen. Occasionally a rounded quartz pebble can be seen in the paste, but this is more the exception than the rule (South 1976:16).
A sample of 1034 sherds of this ware were collected, from 68 sites, and two finishes were identified, cordmarked (N=251 sherds; 24.3 percent) and fabric impressed (N=783; 75.7 percent). While preparing his report, South contacted Waring and described the sherd tempered ware that he had found. While recognizing the similarity with the Wilmington series, they decided that separate terminology would be appropriate because of the geographic separation, and since the ceramics of the intervening area (i.e. coastal South Carolina) were then unknown (South 1960, personal communication). Two types were identified within the series, Hanover Fabric Impressed and Hanover Cord Marked, and these taxa have been widely adopted in the South Carolina archeological literature, particularly in the Coastal Plain north of Charleston Harbor into North Carolina.
Clay/grog tempering has been reported from other localities in North Carolina, although either the wares were untyped (e.g., Haag 1958:69), or else the names advanced have not been widely adopted, as is the case with the Carteret and Grifton series. In the mid-1970s, for example, Loftfield (1976: 54-157; 175-182) formally defined the clay-grog tempered Carteret series, based on materials from 83 sites in south central coastal North Carolina, predominantly from Onslow and Craven Counties. Three types were recognized in the Carteret series, Carteret Cord Marked (N=415 sherds; 22.2 percent) Carteret Fabric-Marked (N=1384 sherds; 73.8 percent) and Carteret Plain (N=74 sherds; 4.0 percent) (Loftfield 1976: 175-182). An Early/Middle Woodland age for the Carteret series was indicated. A similarity or identity of the Carteret series with Crawford's (1966) Grifton series from Lenoir County, North Carolina, immediately west of Onslow and Craven Counties, was noted (Loftfield 1976:234). Comparison of both the descriptions and type specimens for the Carteret and Hanover types indicates that the differences between the series are minimal. In the present study, the Carteret series is dropped and the types subsumed with Hanover as local geographic variants within the Wilmington series
Primary references: DePratter (1979:129, 1991:177).
WILMINGTON (SAND-TEMPERED) CORD MARKED
Sorting criteria: Closely spaced, carefully applied, wide (>2.0 mm) parallel cord impressions. The cord width may be up to ca. 10 mm wide in extreme cases on this heavy cord marked finish. The impressions are typically closely spaced and carefully applied, although care in execution may vary considerably. Paste characterized by varying amounts of small (0.5-2.0 mm), rounded clear or white quartz inclusions. Interior finish typically slightly sandy or gritty in texture.
Distribution: Observed primarily along Savannah River below the Fall Line, and along the Edisto River.
Chronological position: Late Woodland period (A.D. 500-800). Equivalent to Wilmington Heavy Cord Marked on the coast.
Background: The early Late Woodland in the middle Savannah River ceramic sequence dates to ca. 1500 to 1200 BP, and is characterized by sand-tempered plain, cord marked, and fabric impressed pottery. The cord marked material appears to be an inland equivalent of Wilmington Heavy Cord Marked observed at the river mouth. No phase names have been assigned to the early Late Woodland period, which is provisionally described as an interior Wilmington equivalent. No unambiguous diagnostic indicators exist dating assemblages exclusively to this period, although sand-tempered cord marked pottery characterized by closely spaced, carefully applied wide parallel impressions is common.
Wilmington I/Walthour, dates from A.D. 500 to 600, and is characterized by clay/grog-tempered Wilmington Check stamped, Heavy Cord Marked, Plain, and Walthour Complicated Stamped ceramics, the latter an apparent late Swift Creek variant. No interior equivalent for this phase has been observed, although there was undoubtedly a period when the manufacture of Deptford and interior Wilmington wares overlapped. Wilmington II phase components, which dates from ca. A.D. 600 to 1000, are identified by the presence of Wilmington Plain, Brushed, Fabric Marked, and Heavy Cord Marked pottery. A contemporaneous phase provisionally called “interior Wilmington equivalent” has been advanced for the interior Coastal Plain along the Savannah River, where comparable types, differing only in possessing fine sand/grit instead of clay/grog tempering, are present (Anderson 1988, 1994; Sassaman and Anderson 1990).
The difference between the inland and coastal early Late Woodland assemblages lies in the type of temper employed in each area. Wilmington and St. Catherines series ceramics along the lower Savannah are characterized by clay/grog-tempering, while assemblages in the interior are sand-tempered. Clay/grog paste has rounded, subrounded, and irregular lumps of sherd, clay, or fired clay ranging in size from ca. 2 to 10 mm mixed into it. These inclusions typically differ appreciably in color and texture from the surrounding body of the sherd. Clay/grog-tempered paste is extremely rare in the middle Savannah River Valley, where only a few sherds have been found (Sassaman and Anderson 1990).
Late Woodland assemblages dominated by sand-tempered cord marked pottery are widespread in the interior Coastal Plain along the Savannah River, both along the floodplain and in the interriverine uplands. Chronological control within local cord marked assemblages is poor, however, with only a rough separation between earlier and later Late Woodland currently possible, made on the basis of cord size and the occurrence of cross-stamping. There is some suggestion that rim and lip treatment has chronological significance, but this remains to be explored locally.
Primary references: Stoltman 1974; Anderson et al.(1979); Sassaman and Anderson 1990).
WILMINGTON FABRIC IMPRESSED
Sorting criteria: Fabric impressions applied over the exterior surface of the vessel while the paste was plastic; occasionally smoothed somewhat after stamping. The paste is characterized by crushed sherds or grog from 3 to 5 mm in maximum dimension, although larger inclusions up to 10.0 mm are sometimes noted. Rims straight to excurvate, typically rounded.
Distribution: Wilmington Fabric Impressed occurs in the Sea Islands of Georgia and South Carolina. From north of Charleston Harbor to the Pamlico River, it is found both along the coast and in the interior to the Fall Line.
Chronological position: North of Charleston Harbor: Early/Middle Woodland periods (500 B.C. - A.D. 500). South of Charleston Harbor: A.D. 500-1000.
Background: The type Wilmington Fabric Impressed was first defined by Anderson et al. (1982:271-275) based on a sample of 235 sherds from the Mattassee Lake sites on the lower Santee River. The presence of fabric marked pottery within the Wilmington series had been previously observed by Caldwell and McCann during their analysis of materials from the Walthour site near Savannah (as reported in DePratter 1991:34-35). In DePratter’s revised sequence for the north Georgia coast (1991:11, 35), the type Wilmington Fabric Marked was placed in the Late Woodland era and dated to between A.D. 600 - 1000, and interpreted as the first introduction of fabric marking into the area. Waring (1968:220) had earlier included a Wilmington Net-Impressed type in his formulation of the mouth-of-the-Savannah sequence, occurring throughout the period Wilmington pottery was made; this type is not included in DePratter’s (1979:111, 1991:11) reformulation, where it appears to have been replaced by Wilmington Fabric Marked. A clay/grog tempered net impressed ware is reported from the mouth-of-the-Savannah sequence, however, St. Catherines Net Marked (DePratter 1979:131-132, 1991:11, 182).
The Mattassee Lake type sample was dominated by poorly defined, or "loose" weave fabric impressions, where both the warp and weft elements were pliable, although about one-quarter of the assemblage was characterized by a rigid warp element. Overstamping was fairly common, and about one-third of the sherds exhibiting a rigid warp element were cross stamped. The stamping is typically parallel, or at low angles to the rim, and is only rarely perpendicular (stamp orientation determined by the alignment of the warp element with the rim). The assemblage was dominated by reddish-yellow and reddish-brown interior and exterior colors. Temper density varies appreciably, and appears to constitute an appreciable portion of the paste (estimated at from 10 to 50 percent by volume). Quartz sand and other minor mineral inclusions are present in many sherds, although majority are virtually temperless (excluding, of course, the grog), with little or no sand evident. Interiors were poorly to well smoothed, and an appreciable minority of the sherds exhibit a lumpy, irregular surface with fine to coarse, wide scraping marks made with a comparatively soft implement while the paste was quite wet. As noted by South (1976:16), the "large lumps of temper result in a rough, lumpy surface" over the interior of many of the less carefully smoothed sherds; it should be stressed, however, that a majority of the interiors were well-smoothed. Rims were invariably straight to excurvate and rounded, unmodified lips were most common, although about one third of the lips were flattened and stamped with the fabric wrapped paddle. Several bases were recovered, and moderate sized conoidal jars roughly 40 cm in diameter at the rim, and having a capacity of from 10 to 15 liters. The Mattassee Lake material more closely resembles Hanover and Carteret assemblages from North Carolina than Wilmington and St. Catherines material from the mouth of the Savannah, particularly over rim form, lip shape, color, and interior finish.
Primary references: (1) Hanover: South (1960, 1976); Loftfield (1976; Carteret series); Phelps (1981); (2) Wilmington: Caldwell and Waring (1939a); Caldwell (1952, 1971); Caldwell and McCann (1941); Waring (1968c); DePratter (1979:128-131, 1991:).
WILMINGTON SIMPLE STAMPED
Sorting criteria: Simple stamped exterior finish. The paste is characterized by crushed sherds or grog from 3 to 5 mm in maximum dimension, although larger inclusions up to 10.0 mm are sometimes noted.
Distribution: Poorly documented. The type is an extreme minority at the mouth of the Savannah River, and is rarely noted in Wilmington assemblages on the southwestern South Carolina coast.
Chronological position: Late Woodland period (ca. A.D. 500-1000).
Background: Not previously defined. A few sherds of what was called Wilmington Simple Stamped by both Caldwell and McCann and DePratter have been reported at the Deptford burial mound near Savannah (DePratter 1991:119-120). Additionally, DePratter (1991:175) illustrates a sherd of Walthour Simple Stamped. A single sherd of what appears to be a simple stamped, clay-grog tempered ware was also recovered at Mattassee Lake that appears to belong to the Wilmington or possibly Refuge series. The stamp impressions are very faint, reflecting a fair amount of smoothing after stamping; it is possible that the finish may be due to a thong or even cord wrapped paddle. Small (1.0 mm) lumps of clay/grog are present but infrequent in the paste, which is otherwise temperless. Clay/grog tempered simple stamped and cord-marked ceramics have been occasionally reported from elsewhere along the Santee (e.g., Anderson 1975b).
Primary references: Waring 1968:220; DePratter 1991:119-120, 175.