The Gum Swamp Site, Dodge County, Georgia

EARLY ARCHAIC PERIOD:

The occupation of the site during the Early Archaic period saw a distinct increase that lead to the use of many different point types and a move from side-notched to corner-notched and, finally stemmed points. This transition took about 2000 years, not all of which seems to have been spent at the site by its occupants. The site probably served as a seasonal camp site by the upland cultures that moved with the game animals and the environment. This same movement was typical of most cultures of this period. The tools recovered from the site were distinctly campsite tools used for hide preparation and a considerable amount of bone work, given the recovery of three Edgefield Scrapers. There is no record of flake counts or types such as primary percussion and secondary pressure flakes, but the absence of broken, unfinished points and an excessive number of hammerstones suggests that blade manufacturing was done elsewhere. Given that this was a pre-ceramic period, the absence of pottery does not detract from the evidence that this was a campsite during this period.

The side-notched blades pictured above are, from left to right, a recurved Bolen, a Wide-notch Bolen, 2 Bolen type 2 points, a Big Sandy point, and a Taylor point. Ripley P. Bullen named the Bolen point for Bolen Bluff, located along the southern edge of Payne’s Prairie in Gainesville, Florida. Bullen identified several sub-types in this family of points.  His type 2 is the most dominant type in Georgia.  The corner notched type 4 is probably a very early Kirk Corner Notched, both in Florida and Georgia.  The type three is a stemmed form and is rare in Georgia.  The Wide Notch and Recurve are all but unknown.  The type one should probably be a Taylor in both states. 

Radiocarbon dates from the Page/Ladson site along Florida’s Aucilla River place the Bolen point at between 10,200 and 10,000 years BP.

The Bolen is a medium-sized, side-notched point between 1.25 and 4 inches in length. They are left-side beveled with a rhomboid cross-section. Unlike the Taylor, the Bolen’s basal edge can be excurvate, flat, or incurvate with pointed or rounded corners. Bolen notches are usually rounded. Basal edge construction is the primary difference between the Bolen, Taylor, and Big Sandy points. The Bolen point may have functioned as a double-sided Edgefield for working lighter material such as wood. The hafting area is smaller and thinner than the Edgefield and the angle of the blade edge is of a lesser degree. Bolen points are scattered along and north of Georgia’s fall line.  Recoveries below the Fall Line and throughout the Coastal Plane have not been mapped, but will probably include more of the typical variety known in northern Florida including the Recurved type in the Lake Seminole area, the type 5 in south central Georgia, and the Wide Notch type near the St. John’s and St. Mary’s regions.

The Big Sandy point was named for the Big Sandy I phase of the Archaic period described by T.M.N. Lewis. Big Sandy points were dated at the Stanfield-Worley Bluff Shelter at 10,000 BP.[1] The Auriculate form has been dated at 9689 years BP, and the Broad Base and Contracted Base forms share a similar date.

Unlike the Taylor point, the Big Sandy has a variety of basal constructions that have been given a variety of names to describe their differences.  Generally, the Big Sandy is a medium to large point measuring 1.5 to 3 inches in length. The cross section is bi-convex. Big Sandy points are not typically beveled, but may be serrated. Basal construction variations give way to such names as Big Sandy, Big Sandy Auriculate, Big Sandy Broad Base,  and Big Sandy Contracted Base. Big Sandy points are rare in Georgia and occur most frequently along Georgia’s fall line.

Corner-notched points and other point forms began to appear in the archaeological record at the same time or shortly after the side-notched forms appeared. From left to right, the points recovered from the Gum Swamp sites include a Palmer, Kirk Corner-Notched points, a Chipola, a Hardaway Side-Notched, and an Arredondo Point.

Joffre Coe named the Palmer point for examples from the Archaic levels of the Hardaway site in North Carolina. Coe stated that the Palmer points followed the Hardaway occupation in the Early Archaic levels of the site. The Palmer is believed to be the forerunner of the Kirk Corner Notched point that also followed the Palmer at the Hardaway site. The Palmer point is believed to date between 9,500 and 8,000 years BP.

The Palmer is a small corner-notched point, usually measuring less than 2 inches in length. There is no clear definition of size between the Palmer point and the Kirk Corner Notched point, though small, basally smoothed corner notched points are regarded as Palmers. The blade edges of Palmer points are usually beveled on the left side or serrated. The basal edge is flat to incurvate and is always smoothed. John Whatley notes that these points are usually made of local chert or quartz, yet maintain the same diagnostic characteristics. Palmers also share many diagnostic similarities to the Decatur point of Alabama, which is also an Early Archaic type.

Palmer points are widely distributed across Georgia. John Whatley correctly suggested that small corner notched points, labeled as Bolen corner notched points would be better labeled as Palmers. Larger “Bolen Corner Notched points (Bullen’s type 4) would be better labeled as Kirk Corner notched points in Georgia.

Joffre L. Coe named the Kirk Corner Notched point for examples recovered from the Early Archaic levels of the Hardaway site in North Carolina.

Coe believed that the type evolved from the Palmer points and dated to about 8,000 years BP. Their recovery at Florida’s Page/Ladson site on the Aucilla River with other Early Archaic materials dated the type between 8,500 and 7,500 years BP.

The Kirk Corner Notched is a medium-sized point measuring 1.5 to 4 inches in length. The blade is lenticular in cross-section. The blade edges may be serrated or beveled. Beveling, seen along the northern end of Georgia’s fall line, makes the type difficult to distinguish from the corner notched Bolen. Older examples have retained basal smoothing. The basal edge is concave or straight and basal corners are rounded or pointed. Barbs are angled downward at about 45 degrees, but do not extend to the basal edge like the Lost Lake. John Whatley noted that the Kirk Corner Notched point “marks the full transition” from side-notched to corner-notched technology. The Kirk Corner Notch is widely distributed across Georgia above and below the fall line.

John Powell named the Chipola point for the Chipola River where many examples have been recovered. Temporal placement is based on diagnostic features alone. The similarities to other Dalton types and the presence of basal smoothing would suggest an approximate date of 9,500 to 8,500 years BP.

At first glance there seems to be little difference between the Greenbriar Dalton, Hardaway Side-Notched, and Chipola points. The differences come primarily in the way the aricule meets the blade shoulder. The Chipola is more angular and separated from the blade than the Greenbriar Dalton and is more pronounced than the Hardaway Side-Notched point. The blade is triangular with excurvate edges. The shoulders may be angular or rounded. The primary area of distribution is along and southwest of Georgia’s fall line.

Joffre Coe named the Hardaway Side-Notched point for examples from the Hardaway site in Stanly County, North Carolina. The “Nipple Point” described by John Whatley is generally classified as an Osceola Greenbriar. Hardaway points were recovered in zone D of the Stanfield-Worley Bluff Shelter site in Alabama where a date of 9,640 radiocarbon years was secured. This suggests a transitional Paleoindian to Dalton period context and the early stages of side-notching.

The Hardaway Side-Notched is a small to medium-sized point measuring 1 to 2 inches in length. The blade is triangular with straight to slightly convex edges that meet at a broad to acute distal end. Rejuvenation may be bifacial or beveled. The hafting area is deeply side-notched and expands to squared, rounded, or pointed basal ears. The basal edge is the widest part of the point. The basal edge may be recurved and fairly shallow or deeply concave and is heavily smoothed. Basal thinning is common. Hardaway points are found across Georgia with the heaviest concentration along the fall line.

The Arredondo point was first reported by Ripley P. Bullen in 1958 (Bullen 1975:39). The type was named for the town of Arredondo in Alachua County, Florida. The Arredondo is a medium to large sized knife form measuring from 1.75 to 4.75 inches in length.  The blade is developed by generally crude percussion flaking resulting in a fairly thick by-convex cross-section. Excurvate blade edges are pressure flaked to meet at a broad distal and. The type is usually found in an exhausted state with weak, but not rounded shoulders. Stem edges may be straight, tapering or expanding to a concave basal edge. The basal edge is frequently so deeply concave as to appear bifurcated. Basal corners are typically rounded, but can occasionally appear pointed.

Examples of Arredondo Points were recovered from the Trilisa pond site in Marion County, Florida. Examples were recovered at the top of the hardpan layer below 3 feet of sand. The context also contained tools similar to those from the early pre-ceramic levels of the Bolen Bluff site in Alachua County, Florida. That context also produced Arredondo and Bolen points. The Blue Springs site near Marianna in Jackson County, Florida yielded 18 Arredondo points as well as Hamilton points and a Sumter-like point.

The Triangular End scraper was named at the Stanfield-Worley Bluff Shelter site by David DeJarnette, Edward Kurjack, and James Cambron (1962). Sometimes referred to as a Thumb Scraper, they are made from a unifacial flake that is intentionally flaked to a pointed triangular shape. A bulb of percussion may appear at the pointed end with a bit at the broad end of the scraper. The scraping edge is steeply chipped to a sharp, but reinforced edge. Jerald Ledbetter (1995) classified these and other scraper forms as formal unifacial scrapers that he recovered on the Mill Branch site (9WR4). Ledbetter described these “hafted Teardrop-shaped end scrapers” as being made from chert and quartz.

Ledbetter likened them to tool forms commonly attributed to the Paleoindian and Early Archaic periods, some of which may have continued in use into the Middle Archaic period. The context of Ledbetter’s recoveries appeared to have been Early Archaic. The illustration at the center has a spur. This spur is the typical identifying factor of a Paleo End Scraper. There were 115 examples of triangular end scrapers recovered during the Central Georgia Surface Survey, 20 of them from the Deep Creek site in Glasscock County, Georgia. The site contained two Paleoindian point types, seven Dalton points, and forty-two Early Archaic points. Gordon Willey listed Triangular scrapers among those recovered from the Safety Harbor period Safety Harbor Mound site in Pinellas County, Florida. This mound remained in use until near the end of the recorded history of the Safety Harbor culture that ended in about 1725. It is entirely possible that the large numbers of scrapers recovered from the mound may be the result of mound fill taken from earlier sites.

The Core or “turtleback” scraper was also identified at the Stanfield-Worley site. H. Trawick Ward (1993) defined a core as a chunk of raw material from which two or more flakes have been purposefully detached. Cores may be the result of direct or bi-polar percussion. The examples of these cores that could be classified as scrapers from the Central Georgia Surface Survey were all the result of direct percussion. The core scraper is made from an exhausted core that has then had one or more sides modified through pressure flaking to create a scraping face. Core edges that remain unmodified retain their typical percussion flake scares. The Central Georgia Surface Survey that included sites across 18 counties only produced 55 of these tools. The Pool Road Hill Top site that produced 17 of these examples contained predominately Early Archaic point types and pottery was virtually absent. The high number of these scrapers in a heavily Early Archaic context suggests that they, like other unifacial tools, were used extensively during the Late Paleoindian to Early Archaic periods. Gordon Willey[i]recovered a number of what he called “turtle-back” scrapers. While Willey did not define what he recovered, they were consistently recovered from sites that also contained Paleoindian and Early Archaic projectile points including the Safety Harbor site and the Parish Mound 3 site. Research done by James Cambron and David Hulse[ii] on nine sites that included Paleoindian, Transitional Paleoindian, and Early Archaic materials and context along with examples of this scraper type support a Paleoindian and Early Archaic context.

James L. Michie[iii] named Edgefield Scraper for Edgefield County, South Carolina, where twenty-one examples were recovered. Lyman O. Warren[iv]referred to this form as Piper-Fuller knives after recovering them at the Piper-Fuller Airfield in St. Petersburg, Florida. The type is known as the Albany Scraper in Louisiana.[v]The Edgefield Scraper is a medium to large-sized blade ranging between 1.5 and 4.0 inches in length in its un-resharpened state. After rejuvenation, it averages from 1.25 to 2.5 inches long. The blade is unifacially flaked and plano-convex in cross-section before rejuvenation. It has convex edges and a broad distal end. The worked face of the blade is developed through random percussion flaking. Resharpening is done through diagonal beveling on the left side of the blade resulting in an asymmetrically triangular shape. The angle of the bevel across the thick blade averages about 60 degrees. The steep bevel, coupled with the occurrence of small step fractures along the blade edge resulting from heavy pressure during use, indicates that these blades were probably used for working bone.[vi] Step fractures are not evident on the un-resharpened form. The width of the hafting area, required to support this kind of pressure, also tends to support the bone working tool hypothesis.   The hafting area is usually side-notched (although some corner-notched forms are known) and bifacially worked, indicating that secure, uniform hafting was necessary. A smoothed, flat to convex basal edge is normal.

Barbara Purdy’s placement of the Edgefield Scraper within the late Paleoindian tool kit is based on the side-notched and smoothed basal characteristics of the blade and hafting area. This temporal placement is supported by its recovery with Hendricks scrapers at the surface of the hardpan layer at the Piper-Fuller Airfield site.[vii] It has been recovered in Texas with San Patrice points, in Georgia with Taylor points, and with Bolen points in Florida. These findings give the Edgefield Scraper a late Paleoindian to Dalton period association dating between 10,000 and 8,500 years BP.        Its wide distribution from South Carolina to Texas and south to St. Petersburg, Florida indicates that the Edgefield Scraper was a very effective tool form.

MIDDLE ARCHAIC PERIOD:

The recovery of 32 examples of Marion points suggests that this was the primary knife form. The recovery of other point forms is very limited, two of which were South Prong Creek points, suggesting limited woodworking activities at the site. The two Morrow Mountain I points predate the four later Morrow Mountain II points, but the total suggests very light use of the site, primarily as a work site during the Middle Archaic period.

The South Prong Creek type was named from a surface site on South Prong Creek in Richmond County, Georgia. Whatley references Webb’s work at 9DW77 in Dawson County, Georgia in dating the type between 4990 and 3610 BP. The South Prong Creek point is a large blade measuring 2.75 to 3.25 inches in length. The blade is triangular with excurvate edges that are serrated approximately 2/3 of the way from the shoulder to an acute distal end. Shoulders may be straight to tapering. The stem is short and has a flat to excurvate basal edge that may be lightly smoothed. Whatley suggested a statewide distribution. Documented recovers indicate the heaviest concentration may lay along Georgia’s Fall Line.

The Marion point is one of four Middle Archaic stemmed points named by Ripley Bullen and Edward Dolan at the Johnson Lake site in Marion County, Florida. Bullen estimated the Marion point to date between 7000 and 3000 BP. Several Marion points were recovered from the Tick Island site in Volusia County, Florida. The oldest part of the cemetery dated between 4163 and 4375 BP. No Georgia dates have been established. The Marion is a medium to large point measuring between 2 and 4 inches in length. The excurvate blade edges meet at an acute distal end. The blade is fairly heavy with a lenticular cross-section. The shoulders are broad and taper to meet the stem at an angel greater than 90 degrees. The stem is tapering with a flat to rounded basal edge. Marion points can be found from Florida to Georgia’s Fall Line.

Joffre Coe named the Morrow Mountain for examples from the Doerschuk site near Morrow Mountain in the Piedmont area of North Carolina. Jerald Ledbetter suggests that , based on the work of Joseph Caldwell, simply grouping these points as Morrow Mountain points may be an over simplification.  Caldwell’s early discussions of the Old Quartz culture that stretched from today’s Early Archaic to Late Archaic periods were based on tool types.  Coe, on the other hand, used projectile types to define cultural chronology.   Cambron and Hulse documented three different types, Morrow Mountain, Rounded Based, and Flat Based. These divisions reflected much of what Caldwell saw across north central Georgia and South Carolina.  Based on Coe’s discussions, John Whatley designated Morrow Mountain I and II types. Caldwell saw the use of small quartz points, blades, and scrapers covering an expanse from Paleoindian to Late Archaic times.  Coe believed the Morrow Mountain point to be Middle Archaic as do Cambron and Hulse and the generally accepted date has been 7500 to 7000 BP based on work done in Tennessee.  Sites in Baldwin and Richmond counties and the Morrow Mountain burial at the Mims Point site in Edgefield County, South Carolina date the use of the type as late as 5670+/_60 BP., but 6660 years BP in Georgia. Morrow Mountain II points date to as late as 5500 years BP.  Research in Georgia indicates a temporal difference between Morrow Mountain I and II types.  The smaller Morrow Mountain I dates earlier as suggested above while the larger Morrow Mountain II type dates between 6500 and 5500 BP. Whatley also pointed out that examples made of Coastal Plain chert range slightly larger in size. Morrow Mountain I points are small, ranging in length between .75 to 1.5 inches. Morrow Mountain II points range in length between 2 and 3 inches. Some examples from the Savannah River site had what Whatley described as “hanging shoulders,” perhaps allowing for the Flat Based classification by Cambron and Hulse (Putnam County example). Their distribution centers north of the fall line in Georgia.

Side scrapers were described at the Stanfield-Worley site as an otherwise unshaped flake that was worked to a scraping edge along one or more of its long edges. H. Trawick Ward, who recovered two of these scrapers from the Guthrie site in North Carolina, noted that these scrapers were used for hide scraping and cutting. In speaking of scrapers as a class of tools, Jerald Ledbetter further described scrapers as having deliberate unifacial retouch along an edge to produce a steep, convex working edge profile. He added that the shape of flake scrapers suggests that they could have been used as hand-held or possibly hafted tools that were meant for light-duty scraping. The location of scrapers in association to the Late Archaic pit house at the Mill Branch site suggested that they were used most often in a domestic setting. Other archaeologists have agreed with Ledbetter that these tools occur most frequently in household midden deposits.

The earliest recovery of side-scrapers was from the pre-Clovis layers of the in area B of the Cactus Hill site in Sussex County, Virginia where one example was recovered that measured 5.4 cm. long. The pressure flaking along one edge suggested use as a cutting or scraping tool. The sites surveyed in central Georgia contained 109 examples of the side-scraper. The highest number of these blades came from the Deep Creek site in Glasscock County where 12 blades were recovered. The site contained 42 Early Archaic point types, 135 Middle Archaic points, 162 Late Archaic points, 20 Woodland points, and 28 Mississippian to Historic period points. This high number of points from each archaeological period may suggest a consistent use of these blades from Early Archaic to Historic times. Examples recovered by Gordon Willey from the Safety Harbor Mound site that was heavily used as late as 1700 A.D.

Bennie C. Keel (1976) described these as “small ovate knives,” stating that they had straight to slightly excurvate blade edges and rounded basal edges that were somewhat pointed. The distal ends of Keel’s examples were pointed at an average of about 60 degrees. The blades were percussion made, averaged between 32 and 42 mm. in length, 15 to 24 mm. wide, and 8 to 13 mm. thick. All of Keel’s examples, of which there were nine at the Garden Creek Mound 2 site, were made of quartz.

Keel’s Garden Creek recoveries dated with Mound 2 and the Swannanoa and Pigeon series pottery it contained between 200 and 400 A.D. Keel noted that Faulkner and Graham[viii] referred to similar blades as “bifacial knives” in their work at the Nickajack Reservoir in Tennessee. During the Central Georgia Surface survey, 39 examples of Ovate Knives were recovered. The largest group of knives recovered at any one site was at the Rutland Lower Field site where five examples were recovered. That site had a substantial Late Archaic component.

The Triangular End Scraper was named at the Stanfield-Worley Bluff Shelter site by David DeJarnette, Edward Kurjack, and James Cambron (1962). Sometimes referred to as a Thumb Scraper, they are made from a unifacial flake that is intentionally flaked to a pointed triangular shape. A bulb of percussion may appear at the pointed end with a bit at the broad end of the scraper. The scraping edge is steeply chipped to a sharp, but reinforced edge. Jerald Ledbetter (1995) classified these and other scraper forms as formal unifacial scrapers that he recovered on the Mill Branch site (9WR4). Ledbetter described these “hafted Teardrop-shaped end scrapers” as being made from chert and quartz.

Ledbetter likened them to tool forms commonly attributed to the Paleoindian and Early Archaic periods, some of which may have continued in use into the Middle Archaic period. The context of Ledbetter’s recoveries appeared to have been Early Archaic. The illustration at the center has a spur. This spur is the typical identifying factor of a Paleo End Scraper. There were 115 examples of triangular end scrapers recovered during the Central Georgia Surface Survey, 20 of them from the Deep Creek site in Glasscock County, Georgia. The site contained two Paleoindian point types, seven Dalton points, and forty-two Early Archaic points. Gordon Willey listed Triangular scrapers among those recovered from the Safety Harbor period Safety Harbor Mound site in Pinellas County, Florida. This mound remained in use until near the end of the recorded history of the Safety Harbor culture that ended in about 1725. It is entirely possible that the large numbers of scrapers recovered from the mound may be the result of mound fill taken from earlier sites.

LATE ARCHAIC PERIOD:

The amazing part of this site is that, while the site produced 566 projectile points related to the Late Archaic period between 5000 and 3,000 years B.P., it produced no tool types specific to the period, only 46 Stallings pottery sherds, 13 steatite boiling stone fragments and only 7 steatite bowl fragment. Even though the site was only a surface collection, all other sites that produced points and tools of this period also produced a more substantial amount of pottery and steatite fragments. Excavation would have, no doubt, have produced greater numbers of sherds, but given what was present at the site in the surface collection, the large number of Elora and Savannah River points would suggest a period of fairly extensive habitation by Late Archaic people. It is possible that a base camp may have existed close by, but was not included in the parameters of the plowed field.

James Roshto named the Bascom blade for the Bascom site in southern Georgia. Blades with rounded basal edges were referred to as Rocker Base blades Ripley Bullen identified these blades as Morrow Mountain related. Cambron and Hulse named a very similar blade the Maples point from sites along the Elk River near the Maples Bridge. All of the names listed above are given for the same large unfinished blade form that dates to the Late Archaic period between 4100 and 3900 BP.

The descriptions of all of these blades describe them with large, shallow flaking and little or no pressure flaking along blade edges. The basal edges or “stems” are either flat or rounded. These seem to be preforms for such large Late Archaic points as the Savannah River, Pickwick and/or Ledbetter points. The distribution for these large blades seems to mirror these same point types. These blades are widely distributed across Georgia. Most are made of coastal plains chert with only a few quartz examples like the one from Richmond County.

The Savannah River point was named for examples from the Savannah River Focus of the Late Archaic period in the Piedmont region of North Carolina. The larger Savannah River points date with the occurrence of fiber-tempered pottery between 4150 and 3800 BP. The smaller version of the Savannah River point was noted at the Chickasawwhatchee Creek site in Dougherty County, Georgia dated slightly later. There were a fairly significant number of Small Savannah River points at the site that suggested a continued presence between 3800 and 3400 B.C. was significant.

The earlier Savannah River is a large point measuring between 35 and 125 mm in length. The smaller Savannah River averaged between 38 and 52 mm in length. Both early and late varieties have a triangular blade with excurvate blade edges and shoulders are straight to tapering. The cross-section is lenticular with an acute distal end. The stem is straight to slightly tapering with an incurvate basal edge. Both the early and later versions of the Savannah River point are widely distributed across Georgia.

The Lafayette name was given by Ripley Bullen for this type for Lafayette county in Florida. The type has been is sometimes combined with the Clay point because archaeological evidence at Shell Point in Bay County, Florida, indicates that both point types are the same except for the barb shape. A cash of points were recovered having identical basal construction, but with Lafayette and Clay variant barb styles.

Bullen believed Lafayette points dated to the pre-ceramic Late Archaic period between 5000 and 3000 RCYBP. Lafayette examples were dated to 960 BC at the Zabski site in Florida. Whatley reported that nine points of this type were recovered at the Furman Sholes site (9BL69) in Baldwin County. Six examples were in an Early Woodland Refuge context and three were from three were from deeper levels. The use of Lafayette points seem to extend far beyond the Clay type.

Lafayette points are medium to large measuring 1.75 to 4 inches or more in length. The cross-section above the barbs is lenticular, but becomes flattened at the barbs. Rejuvenation normally occurs above the barbs. Stem configuration can be straight to expanding with straight, excurvate, or incurvate basal edges. Basal thinning is often evident up the stem. Along the southern end of the Fall Line and at least as far north as Baldwin County.

The Pickwick point was named for examples recovered from sites along the Pickwick Basin of the Tennessee River Valley. The Ledbetter point (not illustrated) was named for examples recovered from the Ledbetter site in Benton County, Tennessee.

Recoveries of Pickwick points in Alabama site indicate a span of use between late Middle Archaic and early Woodland sites. Recoveries of Ledbetter points indicate a Late Archaic to Woodland date between 4990 and 4020 BP.

Both types are medium sized points measuring between 2.75 to 3 inches in length. Both types have a large triangular blade with excurvate to recurve blade edges. The stems are straight to tapering with a flat to slightly excurvate basal edge. Both have shoulders that are straight to tapering in relation to the stem. The key difference is the blade of the Ledbetter that is typically offset slightly from the center of the stem. Even this is due to resharpening. Examples in Georgia are most often recovered along the fall line.

The Otarrie point was named for examples from the Appellation region of North Carolina.  John Whatley applied the Otarre name to this type sighting Keel’s work in the Appellations and Ledbetter’s work Mill Branch sites in Warren County, Georgia. Whatley suggested a Late Archaic to Early Woodland date of 4600 to 2600 BP.

The Otarre is a medium to large point measuring from 2.25 to 3.5 inches in length. The blade edges are straight to slightly excurvate with slight shoulders and a lenticular cross section. The stem is straight with a flat basal edge.  Whatley described the type as a scaled down “junior version” of Middle Archaic Elora points. Otarre points are widely distributed across Georgia from the Fall Line northward.

The Pickwick point was named for examples recovered from sites along the Pickwick Basin of the Tennessee River Valley. Recoveries of Pickwick points in Alabama site indicate a span of use between late Middle Archaic and early Woodland sites.

Pickwick points are medium sized blades measuring between 2.75 to 4 inches in length. The blade is large and triangular with excurvate to recurved blade edges. The stem is straight to tapering with an excurvate basal edge. Shoulders are tapering in relation to the stem. Examples in Georgia are most often recovered along the fall line.

The Elora point was named for examples from several sites along the Elora area of Lincoln County, Tennessee. John Whatley regarded the Elora as a prime indicator of the Late Archaic period. They are regarded as related to the Paris Island type and date between 4800 and 4300 years BP.

Elora points are medium to large sized measuring between 1.75 and 3.25 inches in length. The blade is triangular with rounded and tapered corners. Blade edges are straight and may be serrated through retouch or purposeful serration with flakes taken from both faces. The stem is thick and contracting with a straight to slightly excurvate basal edge. The basal edge may appear snapped or thickly finished. Elora points can be difficult to distinguish from the South Prong Creek type. Serration on the Elora point typically extend the full length of the blade edge while South Prong Creek serration extend only about 2/3 the length of the blade edge. Elora points are widely distributed from the Fall Line northward.

The South Prong Creek type was named from a surface site on South Prong Creek in Richmond County, Georgia. Whatley references Webb’s work at 9DW77 in Dawson County, Georgia in dating the type between 4990 and 3610 BP.

The South Prong Creek point is a large blade measuring 2.75 to 3.25 inches in length. The blade is triangular with excurvate edges that are serrated approximately 2/3 of the way from the shoulder to an acute distal end. Shoulders may be straight to tapering. The stem is short and has a flat to excurvate basal edge that may be lightly smoothed. Whatley suggested a statewide distribution. Documented recovers indicate the heaviest concentration may lay along Georgia’s Fall Line.

Allendale points, also sometimes called MALA (Middle Archaic, Late Archaic) points, were reported by K.E. Sassaman in his survey of the Savannah River site in Aiken and Barnwell counties in South Carolina. John Whatley used the name Piedmont Allendale to represent a series of similar non-provenience points from the Piedmont region of Georgia.

Allendale points from the Big Pine Tree site in Allendale County, South Carolina dated to the Late Archaic period. Whatley also tentatively dated the Piedmont Allendale of similar morphology to the Late Archaic period. The Allendale is a medium-sized point measuring between 33 and 51 mm in length. Examples generally fall into stemmed or notched categories. The blade of the stemmed form is thick and lanceolate with slight shoulders and a short square stem and flat to rounded basal edge. The blade of the notched type is often more broad and may have barbed shoulders and an expanding stem. Whatley noted that one corner of the stem is often removed through damage during use. Allendale point distribution focuses along the central Savannah River coastal plains area.

Both the Clay and Lafayette names were given by Ripley Bullen for these types for Clay and Lafayette counties in Florida. The types have been combined here because archaeological evidence at Shell Point in Bay County, Florida, indicates that the point types are the same except for the barb shape. A cash of points were recovered having identical basal construction, but with variant barb styles.

Bullen believed Clay points dated to the late pre-ceramic portion of the Late Archaic period between 5000 and 3000 BP. Clay and Lafayette points are medium to large measuring 1.75 to 4 inches or more in length. The cross-section above the barbs is lenticular, but becomes flattened at the barbs. Rejuvenation normally occurs above the barbs. The barbs seem to have been a specialized tool and are frequently found broken off, perhaps through prying to open oyster or clam shells. Georgia examples sometimes have a few serrations along the end of the Barb. Stem configuration can be straight to expanding with straight, excurvate, or incurvate basal edges. Basal thinning is often evident up the stem. Clay points are distributed along the southern end of the Fall Line.

The examples shown above demonstrate the variety of small and large Savannah River points and other types found at the Gum Swamp site.

While not exactly pottery, steatite vessels were the first forms used for cooking beyond the use of gourd containers, wooden bowls, shell cups, baskets or animal skins and bladders used to contain food or drink.  Many objects have been made from steatite throughout history, but bowls are one of the most prominent objects in the Southeast. Archaeologists have noted the close resemblance of early pottery and carved soapstone containers, which may have been manufactured in Virginia by 4,500 years ago. Not only was soapstone fashioned into thick, heavy containers, a few appearing like mortars, but also into rather elegant, thin bowls. Oval and round shapes appear on small-to-large, shallow-to-moderately-deep bowls. Black smudges on soapstone vessels indicate that they were used for cooking. Soapstone vessel manufacture was limited to a few places in Virginia and soapstone ridge near Atlanta, Georgia where the stone occurs naturally. Archaeologists have identified quarries in other areas as well, where boiling stones were manufactured along the Savannah and Ogeechee rivers in Georgia. Manufacturing vessels, establishing trade networks, and transporting and repairing soapstone vessels took a great deal of effort. The original demand for soapstone vessels may have created a need for inexpensive durable vessels that could be produced anywhere from local clays and used by everyone. Although the quarries contained an unending supply of soapstone, the production and exchange networks may not have been able to meet the demand. This may in part explain why ceramic production was accepted first in areas far removed from soapstone quarries.

Our friend and primitive skills expert, Scott Jones, has laid out a full set of instructions for bringing 5 liters of 60 degree water to a full boil in 10 minutes with just a few of these steatite slabs, but suffice it to say, it works. The slabs are about ½ inch thick and range in diameter from 8 to 10 inches. The hole averages about 3/8 to ½ inch. Before the invention of pottery, cooking was not done over an open fire, but it was done beside the fire in a small pit. Tabular steatite was the stone of choice because it could endure the quick expansion and contraction of extreme temperatures. The hole is provided to remove the stone slabs from the fire and from the hot liquids being cooked. Each slab was smoothed with sandstone like the one pictured below from a site along the Ogeechee Rover. One side has a broad, smooth curved surface for finishing the flat portion of the slab while the wide grooves were the result of smoothing the rounded edges of the slabs.

The stone slab industry along the Savannah River at Stallings Island and at several locations along the Ogeechee River in Georgia began some time prior to the invention of fiber-tempered pottery, about 5,000 years B.P. These sites are littered with drills for making slabs, broken slabs themselves, sherds of fiber-tempered pottery, and not much else. The industry ended shortly after people realized that clay pots tempered with sand or other non-organic materials could be placed directly into a fire.

 Stallings pottery was defined by James B. Griffin in 1943.   Antonio Waring used the name Stallings Island with the addition of Plain, Punctated or Incised to describe the surface decoration. This type has been known for a long time from the Stallings Island site near Augusta in the Savannah River. Stallings pottery is tempered with plant fibers such as grass or Spanish moss. Decorative punctations of two types appear in this type. Punctation lines are arranged in rows encircling the vessel parallel to the lip, in V-shapes along the sides, randomly places across the surface, or in meandering designs. Linear punctations were made by a drag and jab method. The second method of punctation appears as rows of individual punctations. Punctations appear as hemispherical depressions, circular depressions, semicircular, fingernail or hollow circular punctations. Known vessel forms include bowl shapes with straight rims. Rims normally slant outward, however inward slanting rims are known at the Stallings Island site. Lips are most often rounded, but may be thickened with a flattened rim to appear as a “T” shape. The “T” rims are regarded as older than other rim forms. Stallings pottery belongs to the Late Archaic period and is believed to date to about 4,500 years B.P. Related point types include Savannah River, Adena, Allendale, Thelma, Culbreath, Clay, and Santa Fe points. This type is known along the Atlantic Coast from North Carolina to St. Simons Island and up the Savannah River well into the Piedmont area and across central and southern Georgia

WOODLAND PERIOD:

The Woodland period, which lasted roughly between 1500 B.C. until about A.D. 850 in Georgia, yielded 661 artifacts with it. There was a significant variety of points and tools present at the site that included 625 sherds of pottery, almost all of which represented the Middle and Late Woodland periods. There were only 3 sherds of Dunlap pottery from the Early Woodland period.

Point types that could have been classified as projectiles included only 3 Duval points, 1 Hernando and 1 Hamilton Triangular point, 1 Florida Copena, 1 Tampa, and 1 Leon-like point. Knife blades included 1 Copena Triangular, 2 Corner Tang Knives, and one Adena-like point. These points and blades are so few in number that no reasonable conclusions can be drawn from this period at the site other than there seems to have been a varied number of cultural influences present during this period, most of which were probably due to the Hopewellian trade influence across the Southeast as seen with the Copena and Adena points.

Jerald Ledbetter applied the generic term Woodland Spikes to a variety of spike forms in Georgia. Individual spike form names have been given here to identify their differences. The Duval Spike was named by Ripley P. Bullen was named for Duval County, Florida. Ledbetter noted that Woodland Spike forms were present during the Early Woodland period with Dunlap Fabric Impressed ceramics dating to between 2600 and 2000 BP. Spikes were scarce in Georgia during the Middle Woodland period with only four examples at the Leake site in Bartow County. Spikes were again used in southern Georgia during the Late Woodland period between 1500 and 1200 BP.

Spike forms are typically small to medium-sized points with a few large exceptions exceeding 3 inches in length. The cross section is diamond-shaped with straight blade edges. Stem configurations have been the separating factor between types as can be seen above. Duval Spikes have straight to bulbous stems.

The Copena Triangular form was first illustrated by William S. Webb and David DeJarnette, naming them Copena after the Hopewellian related Copena culture that spread across portions of Alabama, Mississippi, and Tennessee. Cambron and Hulse later added the Triangular name to distinguish it from the recurved Copena point of Alabama.

Copena Triangular points date to the Early and Middle Woodland periods between 2600 and 1800 BP. John Whatley noted the research of Jerald Ledbetter at the Pumpkin Pie site in Polk County where a Copena Triangular point was recovered from a feature containing Coosa notched points and fabric marked pottery that dated to 2520 BP.

The Copena Triangular is a medium to large sized blade measuring between 1.25 and 3 inches in length. The blade edges are straight to slightly excurvate with an acute distal end. The cross-section is lenticular and the basal edge is straight (top row) to slightly incurvate (bottom center and right).  The incurvate-base form is often referred to as a “Copena Auriculate” point. The excurvate form (left bottom) is often referred to as a “Copena Round Base. Rejuvenation typically occurs along the distal end resulting in a shortened length. As this continues to exhaustion, the blade can appear almost triangular.  Examples made from quartz from the Piedmont area (bottom right) can be much more crude and thicker than those made of chert.  Examples from Florida (Swift Creek site 8LE148 top right) were straight-sided and resharpened from the distal end until the blade became triangular. Copena Triangular blades seem to be closely related to Early and Middle Woodland sites in northwestern Georgia.

Ripley Bullen named the Hernando point for Hernando County, Florida. Hernando points have an Early Woodland association between2500 and 2000 BP in southern Georgia. Their association with Tallahassee points in Middle Woodland times may have extended their use in Florida to as late as 1200 BP.

The Hernando is a small to medium sized point measuring from 1.5 to 2.5 inches in length. They are Georgia’s only basil-notched point. The blade is triangular from its acute distal end to its basal edge. The cross-section is flattened with basal notches that may, on rare occasion appear to form a central stem (figures 4,7, & 8). Blade edges are straight to incurvate. Hernando points are found along the southern counties of Georgia to the Fall line.

TMN Lewis and Madeline Kneberg named the Hamilton point for Hamilton County, Tennessee where they were first reported. John Whatley applied the name Late Woodland Triangular to this type to demonstrate the appearance of these small triangular blades in the archaeological record.

Sassaman estimated the appearance of these points at AD100, “but certainly by AD500.” The lithic technology of the McKeithen Weeden Island sites was dominated by these triangular points. These sites began to appear directly from the Deptford sites by 200 A.D. across northern Florida, spreading across southwestern Georgia.  Mound construction began at the McKeithen site near Lake City, Florida about 350 A.D.  Point types identified as Pinellas (here named the Woodland Triangular), Tampa, and Ichetucknee were scattered throughout the site.  Mound B was constructed as a low platform mound, upon which had stood the house of a prominent person.  A burial excavated beneath the floor of the house contained the presumed occupant.  The individual had been shot in the left buttock with what Jerald Milanich described as an “arrow” point.  The 1.25 inch Ichetucknee point was still embedded in the bone.  This seems to be the beginning of use for the bow and is the earliest fatality by an arrow.

The Hamilton point is small, measuring only between .75 and 1.5 inches (19 to 34mm) in length with occasional examples reaching as much as 2 inches (50mm) in length. The blade ranges from equilateral to isosceles triangular in shape. Blade edges are usually straight, but can be concave. The basal edge is flat to incurvate, or rarely slightly excurvate. Late Woodland Triangular points are expected to occur in nearly all counties in Georgia.

TMN Lewis and Madeline Kneberg named the Greenville point for examples from the Camp Creek site near Greenville, Tennessee. W.H. Baker estimates the Greenville point to date between 2150 and 1700BP.[1] Like Camp Creek and Eared Yadkin points, these points date from the latter part of the Early Woodland to the Middle Woodland periods.

The Greenville point is medium sized measuring between 1.25 and 2 inches (29 to 50mm) in length. The blade edges are parallel in the hafting area and may become excurvate through the upper portion of the blade with an acute distal end. The basal edge is flat to slightly incurvate and the flaking is random. Greenville points range in distribution from northwestern to central Georgia.

John Whatley identified these as Florida Copena points using the name applied by Ripley Bullen for similar points found in Florida. The “Florida” name has been applied here to examples recovered below Georgia’s Fall Line and to those with similar characteristics. Examples of small points with the same cultural affiliation from areas above the Fall Line seem to have more to do with Copena and less to do with Florida. Whatley reported that these points are recovered in Deptford and early Swift Creek sites in central and southern Georgia, dating them between 2500 and 1800 BP.

The Florida Copena is a small blade form measuring between .75 and 1.5 inches in length. The blades are narrow and blade edges, especially among examples from northwestern Georgia, maintain the typical recurved shape of the larger Copena blades of Alabama. Examples from the Fall Line southward are cruder with less of a recurved shape along the blade edge. The basal edge is flat to slightly excurvate or incurvate. The distal end is acute and the cross-section is lenticular. The Florida Copena points range from southern Georgia and Florida to Georgia’s Fall Line.

James Cambron and David Hulse name the Swan Lake type for examples from the Swan Lake area of Limestone County, Alabama. These points are also known as Godly points in central Texas. Swan Lake points may be related to Florida’s much larger Jackson point of the same period that were named by Ripley Bullen. John Whatley placed Swan Lake points in the Middle Woodland period with a transitional Cartersville and Swift Creek ceramic association between 1500 and 2000 BP. His conclusions were based on recoveries from sites in Cobb and Bartow counties.

The Swan Lake is a small point measuring from .75 to 1.5 inches in length. The cross-section is lenticular and thick with respect to its overall size. The blade is excurvate with an acute distal end. The hafting area consists of wide, shallow side notches that extend to an expanding, flat to excurvate basal edge. Swan Lake points are primarily distributed throughout the extreme northwestern portion of Georgia above the Piedmont region, but can also be found in limited numbers as far south as Georgia’s Fall Line.

The Corner Tang Knife is an unusual blade that was first brought to my attention by Jerry Scott of Suwannee County, Florida and was compared to a similar example from the Dent Mound in Duval County. Later research revealed that Ripley P. Bullen had collected several examples, but did not report them or name them before his death. Wilfred Neill illustrated the Corner Tang Knife in his 1963 article on the Taylor point and a Taylor county site, referring to them as a “single-shouldered point or knife.” Cambron and Hulse identified a very similar blade as a provisional Type 6 without the notch.

The Corner Tang Knife is a medium-sized blade, measuring 2 to 3 inches in length. The blade is developed through collateral flaking with very little or no fine pressure flaking along the blade edges. The blade edge opposite the notch is beveled on one face only, forming a median ridge. The excurvate blade edges meet at an acute distal end. The hafting area is formed by a single side or corner notch that may appear on either the right or left side of the blade. The notching styles do not seem to dominate any particular distribution pattern, nor do they conform to any particular point or blade notching of the period. The opposite basal corner is angled up slightly from the basal edge. The basal edge may be convex, concave or straight and is lightly smoothed for hafting. Thinning from the basal edge is typical.

Evidence for the cultural association of the Corner Tang Knife in Florida has been consistently Woodland. One example was excavated from the Dent Mound site in Duval County. The mound dated between A.D. 260 and 590. A second example, described as a reworked Columbia or Bradford, was recovered from a Cades Pond site. One example was recovered from a Pasco County site that also contained Pinellas points. Two examples are from the mixed context of the Suwannee River between Hamilton and Columbia Counties, which has produced many ancient points. Neither example, however, displays the smoothed basal edge typical of Bolen age tools. They may, in fact indicate the use of these blades by the McKeithen Weeden Island people. One example is from a predominately Bolen site near Lake Bird in Taylor County, Florida. Pinellas points, however, have also occasionally been recovered from the site. One example was recovered from the Mayport Shell Midden in Duval County in a disturbed context that contained Weeden Island ceramics.

In Neill’s report of a Taylor County site, although plowing mixed the multiple occupations, the chert material of the “single-shouldered” blade was the same as one of the Taylor points. Jeffries Wyman illustrated a similar example from a shell mound near Lake Beresford along the St. Johns River. Of the notchless examples cited by Cambron and Hulse, two were recovered from Stratum I (Woodland and Mississippian) of the Flint Creek Rock Shelter and three examples were from the upper half of Zone A (Archaic and later) of the Stanfield -Worley Bluff Shelter.

There is some strong feeling that these blades may be related to the Bolen tool assemblage and may have been repeated during the Woodland period. This hypothesis is based on general diagnostic traits. The example in figure 6.13f of unknown provenance, has a distinctly Bolen type five notch. Examples 6.13b and 6.13d may support an older association. This would not be the first occurrence of older points or blades being recovered and reused. There has been, to date, no single component contextual recovery of these blades within a Bolen association.

To date, the known distribution of this type extends from the Central Gulf Coast region through the North-central and St. Johns regions of Florida to as far north as Georgia, Alabama and Tennessee. This distribution appears to parallel the cultural regions of the peninsular Weeden Island cultures.

The Tampa point was first identified by Ripley P. Bullen (1950) for examples from the Tampa Bay and North central Florida areas. The Tampa point is a small tear-drop or leaf-shaped lanceolate point measuring 1 to 1.75 inches in length and .5 to .75 inches in width. The distal and is acute to acuminate. The basal edge is rounded and the cross-section is typically thick. Flaking is random and workmanship is usually poor. The examples from the Parish Mound site are exceptional.

The examples from the McKeithen site may be the earliest forms of the Tampa point. The examples date to the heaviest years of site occupation between A.D. 350 and 475. Recoveries of Tampa points at the Fox Pond site (Milanich 1971) in Alachua County, Florida indicate their association with the Alachua and Hickory pond culture. Gordon Willey’s report (1949) of four Tampa points from the safety Harbor period Parish mound 1 illustrates their association with Tampa Bay area cultures. John Powell (1990) reported one Tampa point made of leaded bottle glass from eastern Florida. A recovery from a colonial association indicates their span of use from between A.D. 300 to 1700.

The predominant area of distribution in Florida tends to be the Tampa Bay region in association with Safety Harbor sites. Their appearance in the Alachua sites of north-central Florida seemed to be limited. Tampa-like points have also been recovered in northeastern Florida at the Dent Mound, level I, as well as the glass example noted above.

Ripley P. Bullen named the Leon point type for Leon County, Florida. Bullen’s research was centered on examples from the Florida State Museum’s Simpson Collection. The Leon is a small, corner-notched point measuring 1.0 to 1.5 inches in length. The blade is triangular with slightly excurvate sides and an acute distal end. The tang is expanding and the basal edge is flat to excurvate. Flaking is random and workmanship is poor to fair.

Examples of the Leon point were recovered from the 8-12 inch and 16-20 inch levels of the Weeden Island component at the Crystal River site.   These examples were comparable to points recovered from the Butler site, a Weeden Island site on the Chattahoochee River.   Their occurrence among surface collections at 8LE156, a special-use upland site belonging to the Weeden Island culture[ix] confirms Bullen’s earlier conclusion concerning their temporal placement, dating the type between A.D 600 and 1200. The wide­spread influence of the Weeden Island culture carried the Leon point from Northwestern Florida to as far south as the Peace River. The Thelma point of Georgia may be a northerly extension of their distribution, although a ceramic association for Thelma points has yet to be established.

The Adena-like point illustrated above is a finely worked example that is similar to the Adena points of Ohio. Adena points are fairly common in northern Georgia. This example has finely finished serrations along the blade edges with a thick, biconvex cross section. The stem is slightly longer than most classic Adena points, but falls within the size ranges described for Adena points.

The Adena name is derived from the point’s association with the Adena culture of Ohio (Cambron and Hulse 1990. Early research from Adena sites in Ohio and Kentucky suggest that these blades were used during both Late Archaic and Woodland periods. Five examples of the Adena point were recovered from Florida’s Watson’s Field site (8JA93) with Santa Rosa Swift Creek ceramics and Bakers Creek points (Schroder 2006:139-140).

The Adena is a medium to large-sized point measuring between 2.5 and 4.25 inches (65 mm. – 115 mm.) in length. The blade is triangular to ovate with an acute distal end and lenticular cross-section. The shoulders are weak and rounded. The stem is only slightly narrower than the blade. The minimum width of the stem in both the Narrow Stem and Wells variants has greater variance with shoulder width. The stem is straight with a rounded basal edge. Examples with a more flat basal edge on the stem are often called Adena Robins points. Adena points may appear in Swift Creek and other Hopewell related Early and Middle Woodland sites in Northwest Georgia.

Dunlap pottery was named by Jesse Jennings and Charles Fairbanks in the SEAC Bulletin in 1940. The type was named for the Dunlap Mound at the Ocmulgee National Monument. It is a sand-tempered ware with a surface decoration that was done with heavy, twisted cord fabric impressed on thin-walled pottery. Like other fabric-impressed types, many have suggested that the pottery was formed in baskets that left the impressions in the paste. Known vessel forms are deep, cylindrical beakers and simple bowls. The lips are most often flat, but are occasionally rounded. The rims usually have a slight flare, but straight bowl rims are known. Dunlap pottery belongs to the Early Woodland period. It occurs in northern Georgia, eastern and northeastern Alabama, southeastern Tennessee and, perhaps, extreme western South Carolina.

The pottery types that did appear in the site belonged to the Middle and Late Woodland periods, suggesting that the campsites of these periods were scattered wide enough to take in part of the site.

The Deptford type was named by Joseph Caldwell and Antonio Waring in the 1930’s. The name. The type was named for the Deptford site in Chatham County, Georgia. This is the same as Cartersville Check Stamped north of Georgia’s Fall Line.  Sand is more often used for temper in Florida and southern Georgia while grit was used elsewhere.  The entire surface of the vessel is covered with check stamping. Stamps are generally square or slightly rectangular and clearly stamped. Some stamping was purposely obliterated after stamping.  Deptford pottery is usually deep, straight-sided jars with rounded or flattened rims.  Vessels often have short, stamped tetropodal legs. Deptford pottery is part of the Middle Woodland period. Deptford pottery is found over a wide range from the South Carolina Coast across Georgia and parts of eastern Tennessee and eastern Alabama and northern Florida as far south as the Tampa Bay area. North of Georgia’s Fall Line it is known as Cartersville Check Stamped, but the pottery is the same as Deptford pottery.

Gordon Willey named St. Andrews Complicated Stamped pottery from sites along the northwestern Florida coast between Choctawhatchee Bay and St. Andrews Bay for which it is named. Fine sand and mica were used as temper in this pottery. Surface is well smoothed, even to a low polish on interiors.  The color is Orange-buff throughout or a gray-black core with one or both surfaces fired. St. Andrews pottery dates to the Middle Woodland Santa Rosa-Swift Creek period. Vessels are straight-sided with straight or slightly out-flaring rims.  The lips are scalloped or notched and the bases are rounded. The vessels are decorated with rectilinear complicated stamping over the entire body of the vessel. The designs include line block, hatched rectangles, diagonally bisected rectangles with each triangular half filled with hachure, and concentric rectangles or triangles. The vessels are pot forms with a flaring rim. The rims are widely scalloped. The lips are slanted inward and flat-round or flat.

Gordon Willey first formulated the “early” and “late” divisions of Swift Creek Complicated Stamped pottery based on his observations at the Carrabelle site in 1949. This division, first applied to Florida examples, has now been applied to Georgia examples as well. The determination between early and late types was based on rim forms, stamping quality and vessel form. Arthur Kelly had done research on the Swift Creek site near Macon, Georgia in 1938. Fine sand mica was used as temper in this pottery with occasional coarser particles. The core is usually gray to black in color with a buff exterior after firing. The decoration on this type is primarily curvilinear complicated stamping in several distinct curvilinear and rectilinear motifs. The design variations are almost limitless. Most designs are highly stylized natural shapes, many of which have been described by Frankie Snow as turtles, masks, flowers, rattle snake rattles, snow shoes and perhaps celestial formations. Design elements include concentric circles, loops, triangles and lines as well as nested ovals and lobes with eyes and irregular shapes. The known form for the early variety of this type is a deep jar or pot with an out-flared orifice. The base is rounded and may have podal supports. The rims are straight and vertical, out-slanted or out-curved. The lips are small, close-spaced notches or round-bottomed broader notches. Swift Creek pottery has been assigned to the Middle Woodland, Santa Rosa-Swift Creek period. Swift Creek pottery is spread across northwestern Florida and all of Georgia, eastern Alabama, eastern Tennessee and southern South Carolina. The illustrated map is after Louis D. Tesar 1980 Map 5, p. 90.[x][i]

Ocmulgee Cord Marked 1 pottery was first identified by Frankie Snow in 1977 following his survey of the Big Bend region of the Ocmulgee River. This pottery was tempered with sand. Many sherds show little to no tempering while others have red zones, indicating a clay temper. The surfaces of these vessels were decorated with a cord-marked paddle. Type 1 markings are clear and run perpendicular to the rim. There is little overlapping of stamping in this type. Cord markings are deeper and clearer than types 2 or 3. The vessels seem to be simple globular jars with straight sides and folded rims. When rims are folded, they are also stamped. This type belongs to the Late Woodland period. Associated points include Woodland Triangular, spike, and Yadkin points. Each of the three types have a distinct area of the Ocmulgee River basin along which it occurs. Type 1 occurs along the most northern section of the river from Macon, Georgia to the area of Hawkinsville, Georgia. There is a blending of areas between each type. Type 2 appears between Hawkinsville and Dodge Boon Landing. Type 3 appears between Dodge Boon Landing and Eason Bluff.

Ocmulgee Cord Marked 2 pottery was first identified by Frankie Snow in 1977 following his survey of the Big Bend region of the Ocmulgee River. This pottery was tempered with sand. Many sherds show little to no tempering while others have red zones, indicating a clay temper. On the surface of this type, cord markings are perpendicular with some cross stamping. Rims are not folded, but an incised line typically runs parallel to the rim to give the impression of a fold. Broad, straight incised lines may also appear in random directions along the body of the vessel. These vessels seem to be simple globular jars with straight sides and folded rims. When rims are folded, they are also stamped. This type belongs to the Late Woodland period. Associated points include Woodland Triangular, spike, and Yadkin points. Each of the three types have a distinct area of the Ocmulgee River basin along which it occurs. Type 1 occurs along the most northern section of the river from Macon, Georgia to the area of Hawkinsville, Georgia. There is a blending of areas between each type. Type 2 appears between Hawkinsville and Dodge Boon Landing. Type 3 appears between Dodge Boon Landing and Eason Bluff.

Ocmulgee Cord Marked 3 was first identified by Frankie Snow in 1977 following his survey of the Big Bend region of the Ocmulgee River. This pottery was tempered with sand. Many sherds show little to no tempering while others have red zones, indicating a clay temper. On type 3 pottery, Cord markings are straight to angular and cross stamped. Rims are simple with cord markings from the rim. No incised lines appear. Stamping is not as deep and clear as with type I. These seem to be simple globular jars with straight sides and folded rims. When rims are folded, they are also stamped. This type belongs to the Late Woodland period. Associated points include Woodland Triangular, spike, and Yadkin points. Each of the three types have a distinct area of the Ocmulgee River basin along which it occurs. Type 1 occurs along the most northern section of the river from Macon, Georgia to the area of Hawkinsville, Georgia. There is a blending of areas between each type. Type 2 appears between Hawkinsville and Dodge Boon Landing. Type 3 appears between Dodge Boon Landing and Eason Bluff.

The tube pipe pictured above was recovered in the Gum Swamp site. As a surface find, it has no cultural context, but certain things can still be learned from other similar research. The tube pipe occurs primarily during the Late Archaic and Early Woodland to early Middle Woodland periods. The absence of Early Woodland pottery seems to rule out the possibility of an Early Woodland context for the pipe. Emma Fundaburk (Sun Circles and Human Hands) sited one Archaic burial of a thirteen year old girl with a Tub pipe under her arm that was excavated in Limestone County, Alabama. The pipe was made from a soft, red stone.  One clay tubular pipe was plowed up in Burke County, Georgia and brought to an artifact ID day in Augusta. The pipe had all three Refuge decorative patterns of simple stamping, incising and punctation that was tempered with sand and grit, typical of Refuge pottery in Georgia. The decoration on the pipe suggests that it was made by a potter that was probably a woman. It is perhaps the earliest form of ceramic pipe in existence. Bennie Keel also recovered two ceramic tubular pipes from the Swannanoa phase of the Warren Wilson site in Buncombe County, North Carolina. The tubular pipe, made of limestone, was recovered from site JAv176A, a large village site along the Tennessee River in Jackson County Alabama. The pipe was not associated with a burial, but was found with an elbow pipe made of fine-grained sandstone. The site also contained Mulberry Creek and Flint Creek pottery dating to the Early and Middle Woodland period.

Lyman O. Wyman studied the use of tobacco and pipes along the shell “fields” of the St. Johns River in Florida during the mid 1800’s, but had recovered no pipes during his research.  C.B. Moore continued Wyman’s search over a four year period and had recovered just five pipes, four broken and one whole tube pipe. Those examples came from the Mulberry Midden, a much younger midden (possibly Early Woodland) than the older Orange period St. Johns middens. Moore concluded that the use of tobacco was unknown to the builders of the earlier middens while the builders of the later Mulberry Midden used tobacco with a tube form of pipe.

MISSISSIPPIAN PERIOD:

The Mississippian period dated between A.D. 900 and 1650. Only 46 artifacts were recovered from this period and only 8 were projectile points, 1 Guntersville, 5 Madison points and two Ichetucknee points. There was only a hint of the cultures that moved into the area of Gum Swamp during this period.

The Savannah culture left only 16 sherds of pottery that were recovered from the site, most of which was the Complicated Stamped variety. The Lamar culture left little more with only 22 sherds recovered, eleven were Bold Incised and 11 were Complicated Stamped.

TMN Lewis and Madeline Kneberg originally named this type Dallas points for the Dallas focus of Tennessee. John Whatley noted that James Cambron renamed the type Guntersville for examples from the Guntersville Basin of the Tennessee River. Guntersville points appear to be associated with Late Mississippian and Historic period cultures in Tennessee. Dallas points are known only from burials associated with Late Lamar sites in northwestern Georgia. This association would date the type between AD1250 and 1650. Frankie Snow reported a group of six Guntersville points in a Historic period site (9Cf17) dating to AD1680. Guntersville points are small to medium in size, ranging between 1 and 1.5 inches (27 to 40mm) in length. The blade is lanceolate in shape with an acute distal end. The blade is formed through random flaking. Dallas points are especially well made. The basal edge is most often flat, but can be slightly concave. Guntersville points are more concentrated in northwestern Georgia with a scattering below Georgia’s Fall Line.

Edward G. Scully initially named this type Mississippi Triangular points, later changing the name to Madison points because of their distribution in Wisconsin and Missouri.  John Whatley used Scully’s original name in his work with The Society for Georgia Archaeology’s Early Georgia publication in its April 2002 addition.  The return to the Mississippian Triangular name seems to encourage lumping together the Hamilton point reported by James Cambron and David Hulse in their Handbook of Alabama Archaeology that was published by the Alabama Archaeological Society in 1964. Scully believed these points to date to the Middle and Late Mississippian period. Whatley dates them between 800 and 1700 AD with their main focus between 850 and 550 BP. Mark Williams states that no small triangular points occur after AD 1350.  Cambron and Hulse date the Hamilton slightly earlier from the Late Archaic period to the Late Woodland period, using examples recovered at several Alabama sites. Madison points are small, measuring from only .5 to 1 inch (14mm to 29mm) in length. The blade ranges from isosceles to equilateral triangles with straight to slightly incurvate blade edges. The basal edge ranges from flat to slightly incurvate or excurvate. Blade edges can be finely serrated.  This same description generally fits all of the triangular points that fall under these three names. Madison points are believed to be present in virtually every county in Georgia.  The map below illustrates the distribution of known Mississippian sites.  Maps like this one contain a survey bias as not all areas have been surveyed.  Clinch County, for example, has many known Woodland sites, yet there are none indicated because they have yet to be surveyed.

The Ichetucknee point was named by John Goggin (1953) for examples recovered along the Ichetucknee River. Ripley P. Bullen (1975) later described the point, comparing it to the Guntersville and Nodena points described by Cameron and the Hulse (1964). The Ichetucknee is a small sized lanceolate shaped point measuring between 1 and 1.75 inches in length. The point is developed through random pressure flaking with a bi-convex cross-section. The blade edges are excurvate with an acute distal end. The blade edges curve in slightly through the hafting area and meet the flat, slightly concave, or rounded basal edge. Workmanship is usually very good.

Ripley Bullen correctly assign these points to the Safety Harbor and Alachua cultures. The earliest occurrence of the Ichetucknee point was at the McKeithen village site (Milanich 1994). The lithic assemblage was most extensive at this site during its middle years of occupation about A.D. 350. Milanich also ascribes the Ichetucknee point to the lithic assemblage of the Suwannee Valley people who ascended from the McKeithen Weeden Island culture about A.D. 750 and settled along the Ichetucknee River. Perhaps it was one of their points that John Goggin recovered and named. Examples of the Ichetucknee were recovered at Hickory Pond and Alachua sites dating between A.D. 600 and 1539 (Milanich 1971). Gordon R. Willey (1949) illustrated Ichetucknee point examples from the Safety Harbor period Parish mound 1 and from the Safety Harbor Village and mound site in Pinellas County. Both of these sites contain colonial period Spanish artifacts and belong to the Tatham or Bayview phases of the Safety Harbor culture dating between A.D. 1500 and 1725. The latest recovery of Ichetucknee points seems to be at the Apalachee Council house at the Fort San Luci site near Tallahassee, Florida. The site dated between 1656 and 1704 (Shapiro and McEwan 1992). One example, made from green bottle glass, was recovered from the Chiefs hut at the mission that would also date between 1649 and 1704. Projectile points from this site were almost exclusively of the Ichetucknee type, many with the excurvate basal edge. One of the few Pinellas points recovered from the site was also made of green bottle glass. One might expect to find Ichetucknee points in the colonial period Utina sites, which ascended from the Suwannee Valley Indian Pond phase, but none were reported by Johnson and Nelson (1990) as their study of this culture centered around ceramics seriation.

The distribution of Ichetucknee points seems to be throughout the North and North-central Florida regions as well as the Tampa Bay area. Their occurrence at the Apalachee council house in Leon County, Florida may indicate a trade link between the Leon Jefferson and Indian Pond cultures occurring about the time of European contact as suggested by Johnson and Nelson.Trawick Ward (1993) defined perforators as “flakes or bifaces that have been finely retouched to produce a pointed bit.” David DeJarnette, Edward Kurjack, and James Cambron, who referred to them as “borers,” defined them as “a blade worked with a drill-like projection, but more likely a hand tool than a typical drill.” Ward interpreted them as hide working punches. This group of drills have a base that is expanded, but without any consistent shape because they required no hafting. They are generally broad, flat and thin, but without regularity of any other form. Bit wear patters that are evident on the tips can be attributed to the circular motion of drilling.

Perforators were among those tools identified with the Clovis age assemblage by Joseph and Lynn McAvoy (1997) at the Cactus Hill site in Sussex County, Virginia. Joffre Coe (1964) recovered two examples of perforators at the Hardaway site in North Carolina. Both “drills” were polished from hand-held use and were associated with the Hardaway complex. A total of 55 examples were recovered in the Central Georgia Surface Survey, most of which were from sites that were dominated with Middle and Late Archaic projectile point forms. Among these sites, the Deep Creek site in Glasscock County, Georgia contained 15 examples of this type with 215 examples of Late Archaic projectile points. The Heards Bridge site in Jefferson County, Georgia contained 13 examples in a context that was primarily a Late Archaic site for working steatite slabs. Of the 31 drills recovered at the Tick Island site in Volusia County, Florida, a predominantly a Late Archaic site that dated to 3,100 B.C., at least 6 examples could be identified as perforators. Ward recovered four examples from the Jenerette site that dated between A.D. 1600 and 1680.

Jerald Ledbetter (1995) generally defined drills as “pointed, secondary reduction bifaces with narrow, parallel blade edges that are symmetrically flaked.” He further defined Cruciform Drills at the Mill Branch sites as appearing to be “both reworked projectile points and specialized biface styles.” The drill hafts seem to represent Savannah River, South Prong Creek, and rounded stemmed points. William Webb referred to drills that were reworked projectile points as “winged drills” because of the shoulders that remained on the drill. The average length of these drills at the Mill Branch sites was 57.4 mm. and the width averaged 21.7 mm.

Ledbetter noted that drills in secure contexts at these sites were only associated with the Late Archaic elements within the sites. At the Heards bridge site in Jefferson County, Georgia, 24 of the 88 drills that were recovered were Cruciform drills. The site was a special use site for the production of perforated soapstone slabs.

Side scrapers were described at the Stanfield-Worley site as an otherwise unshaped flake that was worked to a scraping edge along one or more of its long edges. H. Trawick Ward, who recovered two of these scrapers from the Guthrie site in North Carolina, noted that these scrapers were used for hide scraping and cutting. In speaking of scrapers as a class of tools, Jerald Ledbetter further described scrapers as having deliberate unifacial retouch along an edge to produce a steep, convex working edge profile. He added that the shape of flake scrapers suggests that they could have been used as hand-held or possibly hafted tools that were meant for light-duty scraping. The location of scrapers in association to the Late Archaic pit house at the Mill Branch site suggested that they were used most often in a domestic setting. Other archaeologists have agreed with Ledbetter that these tools occur most frequently in household midden deposits.

The earliest recovery of side-scrapers was from the pre-Clovis layers of the in area B of the Cactus Hill site in Sussex County, Virginia where one example was recovered that measured 5.4 cm. long. The pressure flaking along one edge suggested use as a cutting or scraping tool. The sites surveyed in central Georgia contained 109 examples of the side-scraper. The highest number of these blades came from the Deep Creek site in Glasscock County where 12 blades were recovered. The site contained 42 Early Archaic point types, 135 Middle Archaic points, 162 Late Archaic points, 20 Woodland points, and 28 Mississippian to Historic period points. This high number of points from each archaeological period may suggest a consistent use of these blades from Early Archaic to Historic times. Examples recovered by Gordon Willey from the Safety Harbor Mound site that was heavily used as late as 1700 A.D.

[i] Willey, Gordon R.

1949       Archeology of the Florida Gulf Coast. Bureau of American Ethnology, Smithsonian Institution

[ii] Cambron, James and David Hulse

                1960       Alabama Archaeology

[iii] Michie, James L., The Edgefield Scraper, The Chesapiean, Vol. 6, 1968:30

[iv] Warren, Lyman O., Unique Knife or Chisel, Piper-Fuller Airfield, St. Petersburg, The Florida Anthropologist, Vol. 26:3, 1973:119-120

[v] Webb, Clarence H., Two Unusual Chipped Stone Artifacts from Northwest Louisiana, Bulletin of the Texas Archaeological and Paleoindian Anthropological Society, 1946:9

[vi] Purdy, Barbara A., Florida’s Prehistoric Stone Technology: A Study of the Flintworking Technique of Early Florida’s Implement Makers, University of Florida, Gainesville, 1981:26-28

[vii] Warren, Lyman O., Unique Knife or Chisel, Piper-Fuller Airfield, St. Petersburg, The Florida Anthropologist, Vol. 26:3, 1973

[viii] Faulkner, Charles H., and J.B. Graham

1966       Highway Salvage in the Nickajack Reservoir, Department of Anthropology, University of Tennessee

[ix] Percy, George W. and M. Katherine Jones, Archaeological Survey of Upland Locales in Gadsden and Liberty Counties, Florida. The Florida Anthropologist, 29:3, 1976:Figure 4f