2000 Florida Anthropological Society Inc.
The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
all rights to the
and shall be
and images of
The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.
The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.
Contact the Florida Anthropological Society for additional
information and permissions.
THE FLORIDA ANTHROPOLOGIST
VOL. X JULY, 1957 NOS. 1-2
THE MADDEN SITE _-__----- ------ -----------..-_____ D. D. LAXON 1
THREE SMALL DADE COUNTY SITES ---_---_ ----_ D. D. LAXON 17
THE BARNHILL MOUND, PALM BEACH COUNTY, FLORIDA,
RIPLEY P. BULLEN 23
THE NORMAN MOUND, McINTOSH COUNTY, GEORGIA,
LEWIS H. LARSON, JR. 37
THE MADDEN SITE
Situated in the typical Everglades ridge and slough top-
ography, one mile south of the Golden Glades levee, in the SW
quarter of Section 15, Township 52 South, Range 40 East, is
located a heavily wooded area, roughly seven hundred feet
North to South and five hundred feet East to West.
The hammock is in the shape of a crude crescent with its
opening to the southeast and east. It is five feet above the
surrounding terrain which is, in turn, five feet above sea
level (Fig. 1). This area is athwart the opposing erosional
forces of a drainage pattern from the northwest and prevail-
ing winds from the southeast. Continual erosion has therefore
altered the original shape somewhat.
The soil is composed of quartz sand, colored black and
various shades of grey by organic material and impurities.
Tests seemed to indicate that muck did not underly the site.
Sands of the Pamlico or Silver Bluff terrace extended to the
basal formation of limestone. On the hammock proper the or-
ganic soils run from one to two feet thick with the greater
depth in the north and western part.
In the southwest quadrant of the hammock there is a natu-
ral sand hillock, the top of which is nineteen feet above sea
level (Fig. 1). This sand mound is shaped somewhat like a
truncated pyramid with the top fifty feet wide and one hundred
and fifty feet long. Sides slope gently at an angle of from
eight to ten degrees for a distance of eighty-five to one hun-
dred feet. Vegetation on the crest of this hillock is sparse
and consists mostly of ragweed, scrub palmetto and wild citrus.
A trail, ten feet wide, kept clear by swamp buggys and
jeeps, crosses the mound from west to east, skirts the ham-
mock's eastern and northern edges and meanders in a general
northeasterly direction to a point just south of the inter-
section of Golden Glades Road and Peter's Pike.
This tract was among the vast acreage purchased by R.
Bolles in 1909, during the formation of the Everglades Fruit
and Land Company. This organization still owned the land in
1911. In December, 1928 the Board of Commissioners of the
Everglades Drainage District sold the hammock to Mr. A. C.
Madden, hence the name, Madden's Hammock.
Since the site is conspicious and readily observed from a
distance, it is known to have come to the attention of archeo-
logist, both professional and amateur. Collections were made
by Dr. John M. Goggin of the University of Florida and Karl
Squires, a former county engineer.
The site is officially recorded as Dd45 in the University
of Florida records. It would, if for no other reason than
completeness, be unwise to omit so imposing a site from any
general survey of the Hialeah area.
The flora is, of course, profuse and of considerable va-
riety. A list is included in this site report. It is inter-
esting to note the presence of coontie on the east side and,
considering the multitude of airplants, the lack of orchids.
This scarcity of orchids probably attests to the dryness of
This location was used by the Seminole for the Green Corn
Dance until roads were built into the vicinity. While theo-
retically the ceremony was held in the same place for four
years, in order to maintain secrecy, they often changed the
site at intervals of one, two, and three years; and would have
resented easy access to the grounds.
With relation to the Seminole it is also interesting to
note physic nut trees growing at the base of the eastern slope.
This tree could be an excellent substitute for the yaupon, us-
ed in concocting the ceremonial, "black drink." There is
scattered evidence of Seminole occupancy to fairly recent
times. Numerous beads have been screened from atop the sand
SEC 5 \
._v i N
+ TURTLE BONE CACHE A
o rrar pir
Figure 1. MIap of AMden Site Showing
+ TURTLE BoNE CACHE -
f SCRUB PfLMCT e ro
Figure 1. Map of Miadden Site Showing
Location of Pits
mound in the last few years. Two large turtle bone caches,
representing gopher, terrapin, and soft shell were uncovered.
One was on the south side of the west slope, the other across
the hillock on the east side of the opposite slope approxi-
mately half way down. These probably were Seminole garbage
heaps, considering their tendency to throw all the shells in
the same place. The preponderance of turtle remains, reflect-
ed in the Indian midden material, is probably due to the high-
er water table during Indian occupancy.
The north and northwest sides of the hammock drop sharply
into a low, swampy trough that encircles most of the site. In
this area is the bulk of the midden. Several observations
concerning the natural sand hillock are noteworthy.
While the shape of the natural sand ridge is irregular,
as compared to a temple mound, it is felt, at the risk of be-
coming a polemic, that the scarcity of aboriginal material,
considering the nearness of the midden, gives rise to the
speculation that the mound was used as the base for a caci-
que's house or a temple mound during part of the aborigine's
occupation. The top of the mound, under the above conditions,
would be consecrated ground.
Since occupation was contemporary with the period of
Temple Mound influence from the north, and unless the influ-
ence had been felt to some degree, aboriginal material would
have extended to the mound proper. Close scrutiny of its
crest turned up only Seminole and modern artifcats. Yet mid-
den material can be picked up on the surface at the base of
the northwest slope of the sand ridge, less than one hundred
feet from the top. It is an unlikely possibility that the
Seminole had removed the prehistoric material in the process
of occupying the ground themselves.
The top of the sand ridge and most of the midden area was
swept carefully with a metal locator. For the most part, ket-
tle iron, nails, and several unidentified iron stakes made up
the bulk of metal material found. The amount of refuse, in-
cluding pottery, numerous sea turtle, alligator, deer bones,
clam, and oyster remains attested to a typical Everglades sub-
sistence gathered from a considerable distance.
Digging was started in July, 1955 and continued until May
1956, when the last of the twelve test pits were finished.
Excavation of the first three test pits took place on the west
side of the trail, very near the point where it leaves the
hammock's edge for the sawgrass. Test pits are located in
Pit 1 produced the usual familiar pottery with Glades
Plain predominating. An excellent example of a Strombus celt
was found at a depth of fourteen inches.
Pit 2, along with Glades Plain, showed rounded and flat
rim sherds with Glades paste. Three of the sherds were of
soft paste, not St. Johns, and were included under Glades
Plain but were far from typical. Fragments of bone awls and a
Macrocallista knife were also found.
Pit 3 showed a small Strombus celt, more of the ever pre-
sent Glades Plain, and a quantity of sea turtle bones.
In each of the pits there were conch, clam and oyster
shells, and many varied bone fragments. As these tests were
thought disappointing and tests showed the bulk of the midden
further west, the scene of operation was moved slightly west
of north, more on the hammock's slope.
Pits 4 and 5 were therefore started about twenty feet
west of the first three tests. The upper level of Pit 4 pro-
duced more Glades Plain, Belle Glade ware, and a single Glades
Tooled rim sherd. A Strombus celt was also found at this lev-
el. A Strombus celt, short bone points, and fragments of four
shell tools were in the second six-inch level.
It was decided to excavate further in this vicinity, Pit
5 was therefore made a five foot extension of Pit 4 southward.
In both the 0-6 inch level and the 6-12 level of Pit 5, Fort
Drum Incised was found, coupled with a Strombus celt and short
bone points in the 6-12 inch level, giving this pit a Glades
II flavor. In the 6-12 inch level was also found a broken
stone celt, possibly imported from the north. Six of the
t 1 T I l N.
Figure 2. Glades Tooled Rim Sherds From
The Madden Site
-sbr a-,;R ^
Figure 3. Glades Tooled Rim Sherds From
The Madden Site
! I T ,f
Glades Plain rims in the top level appeared to be extremely
thick and from their contour, from large, shallow vessels. A
single pewter bead, a disc shell bead and several "flaked
surface" sherds, possibly from north Florida, were uncovered.
Pit 6 was excavated paralled to, and fifty feet west, of
Pit 4, Pit 7 being another five foot extension of Pit 6 south-
ward. Among the first material removed from these pits were
several variants of Glades Tooled. In the lower levels bone
points, a Strombus celt, and Key Largo Incised sherds were
found. This suggests normal stratigraphy with Glades II re-
mains lying under those of the Glades III period.
An exciting find in the four to eight inch level of Pit 6
was seventeen sherds of a Glades Tooled variant best described
as "crimped". Obviously, all sherds were from the same vessel,
twice evidence of repair during manufacture appeared. There
are also, in both cases, repair holes on each side (Fig. 3,
third row). Apparently, the vessel cracked during the drying,
patches were added at the rim in an attempt at repair before
firing. After firing, it was found the patching did not alle-
viate the trouble and drill holes were made for lashing. It
could be seen that the patch was added before firing, the area
was fired with the rest. That the holes were made after fir-
ing is clear from the appearance of the edges of the holes.
It is possible that these "crimped" rims are a transit-
ional form between Glades II to Glades III, and should not be
included with Glades Tooled.
Test Pit 8 was excavated opposite Pit 9 but a few feet
west and with its lower edge resting almost in the trough at
the bottom of the slope.
In the upper levels of Pit 8 were found several unique
sherds. There were two with a brushed surface, suggestive of
Seminole, although they were not classified as such. Another
unclassified sherd, with a five-line incising, was reminiscent
of Fort 'Talton or Weeden Island. A single rim sherd with ap-
pliqued bosses a half-inch below the lip was also found.
While different from the type usually found, it is probably of
the same time period, Glades II. The nodes had been made by
pushing a tool in the still damp clay from the interior of the
vessel, making small knobs on the surface, and leaving small
holes in the interior.
Noded pottery is not new in the southeast. Jennings, in
his "Prehistory of The Lower Mississippi Valley", states the
type has been found in Louisiana, Georgia, Alabama and Missis-
sippi. On the local scene it was found by Goggin and Sommer
at Upper Matecumbe Key, eighty miles to the south, and by
Brooks at Grossman's Hammock, twenty miles to the southwest.
A small sherd with a zig-zag motif on top of the flat rim
(Fig. 4), parallel lines below the rim with vertical lines
connecting, was suggestive of Matecumbe Incised. Many pieces
of nineteenth century stoneware and iron kettle fragments were
also found. The uncovering of Spanish majolica, St. Louis
Blue-on-White, was definite evidence of occupation during
Spanish times. A portion of greenstone celt which had been
reworked into a chisel or gouge-like form was found in the 0-6
Test Pit 9 was excavated a few feet north and east of Pit
8. Pit 10 was placed at right angles to Pit 8 on its southern
edge. In the Pit 9 test Strombus celts were found at both the
0-6 and 6-12 inch zones. One unique incised sherd and a
chamferedd" surfaced sherd was also found. Mexican majolica,
Pueblo Polychrome, of the Spanish Mission period, circa A.D.
1700 was found in the 0-6 inch level.
In one small area of Pit 10 at the 0-6 inch level were
found eighty body sherds and five rim sherds from the same
Glades Tooled pot. The vessel was nicely made and exhibited
some form of burnishing or polishing. At the same level was
found an example of Little Manatee Shell Stamped, apparently
an importation from the west, where it would be probably late
Weeden Island in time period. This pit, while the lower level
seemed older than the upper, produced very few Glades II items.
Pit 11, a "feeler", was excavated sixty-five feet west of
pits 8 and 10. Among the usual midden refuse of shell and
bone and "run of the mill" sherds, were two large St. John
Check Stamped, Spanish majolica, and several unique incised
rims. One of these had a notched lip resembling a Safety Har-
bor (Fig. 4, last sherd) trait. Fourteen Glades Tooled rims
in the 0-6 inch level and four in the 6-12 inch level includ-
ed new variants of this type.
Pit 12 concluded the work at this site and was excavated
as an extension of Pit 9. In the 0-6 inch level of this pit
were found two fascinating specimens of unique incised with
notched lips. There were also more unique sherds with bosses
on the surface. These differed from those previously found as
they were formed by an applique. A Strombus celt and Spanish
majolica, St. Louis Polychrome, was also uncovered. The last
level contained two hundred and thirty-two sherds; forty-two
being rim sherds, for the most part, Glades Tooled. Belle
Glades Plain were also represented.
Because of the wide variety of Glades Tooled rims found
at the Madden site, important variations have been illustrated
in Figures 2 and 3. Allowing for minor variations, a total of
sixty-one different Glades Tooled vessels were represented by
sherds from this site. Four unique sherds have been included
as Figure 4.
Figure 2 covers Glades Tooled rim sherds with lips ap-
proximately the same width as bodies of sherds. These lips
measured from three-sixteenths to three-eights of an inch in
width. Of thirty-one variations of such sherds, ten are il-
lustrated. Unillustrated sherds consisted chiefly in minor
variations in spacing of tool marks.
Figure 3 illustrates two groups of rims both with lips
wider than vessel walls. Above the scale are six sherds re-
presenting a total of twenty-three variations. These rims are
from extremely shallow vessels at least vessel walls curve
substantially just below the lips. Lips range from three-
eights to three-quarters of an inch in width.
Below the scale in Figure 3 are illustrated four rims
which have "thinned" rims. These rims have been squeezed or
pinched and drawn "upwards". Usually, most of the "thinning"
takes place on the inside of the rim. This is quite notice-
able in the middle sherd of the lowest line of Figure 3. Lips
are about a half of an inch in width and nearly vertical in-
stead of being nearly at right angles with vessel walls. Of
seven variations of these lips, four are illustrated.
Tests seemed to indicate not a small midden, but literal-
ly a Glades III metropolis, with signs of occupancy during the
entire period, but retrogressing, considering the paucity of
material, almost reluctantly into late stages of Glades II.
Among the excavated refuse, was found pottery left by
transitory residents, that reflecting competitive artistry of
a large number of local potters, and the residue of trade cer-
amics. There was also uncovered evidence of Seminole occupat-
ion, Spanish contact, and of influences from north, south, and
In the immediate vicinity of this large site are several
small family middens, all showing definite occupation in
Glades II times. It is possible this large sand dune didn't
support vegetation, other than scrub, during these times.
Thus the site may not have been considered habitable by the
aborigine, except as temporarily a stop-over point, until soil
had built up sufficiently to support the larger trees.
Appreciation is expressed to the following: Mr. H. B.
Smith, owner of the land, for permission to dig; Bob Masters,
Noel Herrmann, and W.R.Angleton for digging and screening; Dr.
Wilfred T. Neil for information concerning the Seminole;Dr.
Virgil Sleight of the Geology Department, University of Miami,
for identification of sands; Mr. F. D. R. Parks, of the Dade
County Engineer Office for permission to trespass the Golden
Glades Levee; Dr. Malcolm Birdsey, of the Botany Department,
University of Miami, for identification of flora; and Mr.
Ripley P. Bullen, Curator of Social Sciences, Florida State
Museum, Gainesville, for help with typology and photographs.
Figure 4. Unique Sherds From The Madden Site
LIST OF PLANTS EITHER COLLECTED OR OBSERVED AT
HIALEAH # 6
Dr. Malcolm Birdsey
Quercus virginiana Live Oak
Elaphrium simarubra Gumbo Limbo
Simarouba Glauca Paradise Tree
Zanthoxylum fagara Wild Lime
Salix amphibia Southern Willow
Myrica cerifera Wax Myrtle
Ficus aurea Strangler Fig
Celtis mississippiensis Hackberry
Eugenia axillaris White Stopper
Eugenia uniflora Surinam Cherry
Chiococca alba (?) Snowberry
Cassia (2 species)
Nephrolepos exaltata Boston Fern
Kalanchoe crenata Succulent Plant
Icacorea paniculata Marlberry
Rapanea guayanensis Myrsine
Tillandsia usneoides Spanish Moss
Tillandsia (several species)
Panicum albomarginatum Grass
Vinca rosea Periwinkle
Guilandina crista Nicker Bean
Melia azedarach China Berry
Serenoa repens Saw Palmetto
Psychotria undata Wild Coffee
Jatropha curcas Physic Nut
Mormordica balsamina Balsam Apple
Bacharis halimfolia Groundsel-tree
Tamala (Persea) borbonia Red Bay
Sideroxyon foetidissimum Mastic
S A C C 0
n 1f ii Ib
.. u 6 7 10 11 LJ Deiih
1. Strombus celt. __n--
& x x x7 6-12
3. Worked columnella. x 0-
3. Short bone points. Yx 0-6
Portable hone. 0-6
Ground stone celt. --i
________ -x 6-12
3. Fasciolaria hammer. o0-
7. Bone awl. x 0-6
Tip of Busycon pick. x 0-
9. Frag. Macrocallista knife. x 0-6
10. Disc type shell bead. 0-6
11. Pewter bead.
12. 19th Century crockery. x x x x x 0-6
13. Worked turtle carapace. 0-i
14. Columnella tool. x
15. Fasciolaria tool. M 0-
16. Frag. Greenstone celt. x-6
17. Spanish majolica. (Unident'd.) x x 0-6
S1. Pueblo Polychrome. 0-6
19. St. Louis Polychrome. 0-
20. allhasee bue-n-wite 6-12
20. Tallahassee blue-on-white. x 0 6
t eef Pi t Numb
Bone And Shell Distribution
1 4. 7 8 Q 1 0 11 19
Turtle x x x x x x x x x
lam x x x x x x x x x x x
Otter Dentary x x
Busycon x x___
Rabbit -- x x
Conch x x x x x x x x x x
Racoon Dentary x x
Crab Claws x
Alligator x x x x x x x x
Sea Turtle x x x
Deer x x x x x x x x x x
Sting Ray Spline x _
Fish x x x x x x x x
Oyster x x x x x x x
Shark Vertabrae x x x x
Opposum Jaw x
Bird x x x
Soft Shell Turtle x
Rabbit Jaw_ x__
Teeth (Lynx) x
PLAIN. shell tempered 1 1
BELLE GLADE PLAIN 44 31 75
GLADES PLAIN, with thickened 4 1 5
and inturned rims
GLADES PLAIN 1370 1062 2432
ST. JOHNS PLAIN 20 25 45
PLAIN, rounded rims Glades 2 2
GLADES PLAIN flat rims like 59 59
Belle Glade, Glades paste
PLAIN, coquina tempered 8 10 18
PLAIN, sand tempered, flated 1 1
PLAIN, brushed surface 2 2
GLADES PLAIN, fluted rims 1 1
ST. JOHNS PLAIN. rough surface 2 2
PLAIN. chamfered surface 1 1
PLAIN, coquina tempered 1 1
GLADES TOOLED 83 32 115
GLADES TOOLED,crimped rims 17 17
ST. JOHNS CHECKED STAMPED 31 32 63
ST. JOHNS CHECKED STAMPED, Belle 2 2
ST. JOHNS CHECK STAMPED, smoothed 1 1
over gritty paste
MISC. INCISED (1) 2 2
FORT DRUM INCISED 2 1 3
KEY LARGO INCISED 1 1
MISC. PUNCTATED 1 1
FIVE LINED INCISED 1 1
UNIQUE INCISED, bosses below lip 1 1
MISC. INCISED (2) 1 1
UNIQUE INCISED, zig-zag motif on 1 1
UNIQUE INCISED 2 2
LITTLE MANATEE SHELL STAMPED 1 1
MIAMI INCISED 1 1
UNIQUE INCISED, notched lip 3 3
UNIQUE INCISED, bosses on surface 2 2
UNIQUE INCISED, zig-zag motif on 1 1
rim, parallel and right angle
lines pendant to rim
Brooks, Marvin J., Jr.
1956. "Excavations at Grossman Hammock, Dade County, Flo-
rida. "The Florida Anthropologist, Vol. IX, No. 2,
pp. 37-46. n. p.
Griffin, John W. (ed)
1949. The Florida Indian and His Neighbors. Winter Park.
Goggin, John M., and Frank H. Sommer, III
1949. "Excavations on Upper Matecumbe Key, Florida." Yale
Univ. Publ. in Anthropology, No. 41, New Haven.
Jennings, Jesse D.
1952. "Prehistory of the Lower Mississippi Valley." In
Archeology of Eastern United States, J.B. Griffin,
ed., pp. 256-71. Chicago.
Laxon, D. D.
1953. "Stratigraphy at a Hialeah Midden." The Florida
Anthropologist, Vol. VI, No.l,pp. 1-8. Gainesville.
Neill, Wilfred T.
1952. Florida's Seminole Indians. Ross Allen's Reptile
Institute. Silver Springs.
Willey, Gordon R.
1949a."Excavations in Southeast Florida." Yale University
Publications in Anthropology, No. 42. New Haven
1949b."Archeology of the Florida Gulf Coast." Smithsoni-
an Miscellaneous Collections, Vol. 113. Wash.
THREE SMALL DADE COUNTY SITES
D. D. Laxson
These three, small, widely scattered Dade County middens
were excavated over a period of several years. It was felt
none of the sites were of sufficient size to warrant individu-
al publication. They are, therefore, presented combined as a
matter of record and to reveal several interesting, possibly
transitional, pottery forms.
SITE NO. ONE
In the NE quarter of Section 30, Township 52 South, Range
41 East, on property owned by the U. S. Government as the
Marine Air Station, is located an elongated hammock best fa-
cetiously described as "ham hock" shaped. It is about eleven
hundred feet long north-south and several hundred feet wide on
the northern end. A road had been cut east-west and bisected
by a firebreak running across the top of a drainage ditch
spoil bank. The midden was located several hundred feet south
of the road and twenty-five feet east of the drainage ditch.
Three test pits were excavated in this area. In recent months
during the course of work on the airfield this hammock was
destroyed and no longer exists.
It would indeed be repititious to describe local flora
and fauna. It is identical with other hammocks in the area.
There are several interesting features about the midden
area. Faceted blue beads in the surface material hinted at
recent Seminole occupation. There was a large percentage of
split bones and many were charred. There was very little
shell. Strombus celts and bone points were found. In the 0-6
inch level of Pit 2 a small copper cone, 10 mm. wide at the
base and 38 mm. long, identical with one found a mile and a
half NW in Opa Locka No. 3 by Gordon Willey, was uncovered.
Gratification was felt at the uncovering of another sherd of
the Matecumbe variant consisting of a zig zag motif pendant to
the rim such as described in the article "Some Incised Pottery
From Cuba and Florida" by Bullen and Laxon (p.23). The sherd
found had two repeats of the pattern.
Pottery was predominantly Glades Plain, as usual. Pits
were shallow, and ended in marl at a depth of one foot.
SITE NO. TWO
This small habitation site is located seventy-five feet
east through the mangroves on the east shore of the Uleta Riv-
er directly opposite its intersection with the Royal Glades
Canal. The site is on an island part of Greynolds Park and is
about one and a half miles from the Atlantic Ocean.
The highest ground is about three feet above the river
and supports typical hammock vegetation including mastic, coco
plum, ficus, gumbo limbo, and hackberry. The soil is sand and
black dirt. The basal formation is a friable, laminated lime-
stone covered by the soil in depths varying from six to twenty-
four inches. The area surrounding the midden is swampy and
the river banks are debris-littered and pock-marked with num-
erous land crab holes. There are large deposits of Winged
Tree oysters along the river banks to the south near its en-
trance into the inland waterway system.
A series of test holes along the crest of the hammock
as unsatisfactory. Scraping through the dirt at the entrance
to the crab holes showed pottery fragments, so a place was
cleared for Pit 1. Along with pottery; shark vertabrae,sea
turtle bones, and conch shells were found. Pit 2 was dug fif-
teen feet west of Pit 1. More Glades Plain pottery, turtle
bones, and clam shells were found. Up to this point not a
single piece of incised pottery had been found.
A place was found thirty-five feet southwest of Pit 2
where the soil showed a depth of twenty inches. Pit 3 was ex-
cavated in this area. In the upper level were Glades Plain,
Belle Glade Plain pottery, and shark vertabrae. Numerous
conch and clam shells were also found. At the eighteen inch
level were found seven large sherds of Opa Locka Incised.
Several small holes were dug on the swamp side with no
results. It was decided that the site warranted no further
excavations so notes were made and the holes refilled.
In summation, the site seemed a typical coastal midden
and the shallow cultural deposits indicated a short occupation.
It is possible a change in salinity or depth of the water may
have been a contributing factor here. The numerous plain
sherds and the presence of Opa Locka Incised and Strombus
celts at the bottom level shows occupation during Glades II
times. Tests also showed conclusively that at or near the
ocean the percentage of ocean food has increased in comparison
with food remains at the more inland sites. This is undoubt-
edly explained on a geographical-ecological basis, but a defi-
nite correlation is interesting.
SITE NO. THREE
.This small midden is located in a wooded area in the cen-
ter of the SW t of Section 8, township 54 South, Range 40East
about half way between the Tamiami Trail and Coral Way fifteen
hundred feet east of SW 107th Ave. The hammock is for the
most part sand with a pot-holed limestone base that surfaces
at times. The midden is in the black dirt portion and its
center is marked by the remains of a sapling lean-to. The
quantity of charcoal, cement block, and pipe suggests an aban-
doned still location. The midden area is approximately fifty
feet in diameter, is several feet above the level of the road,
and the black dirt extends approximately one foot deep.
Four test pits were excavated. Pottery was for the most
part Glades Plain. However in the 0-6 inch level of the
second test pit another sherd of the Matecumbe variant with
the zig-zag motif below the rim was found. This type, al-
though found in small quantities, nevertheless has cropped up
in several Dade County middens. In the 0-6 inch level of Pit
1 a large iron spike was found. It was hand-forged, 10 3/4
inches long, with a head 3/8 inches thick and 1l inches in
diameter. Bone points and Strombus celts were also found.
The most interesting pottery was found in Pit 4. Samples
of Dade Incised and Miami Incised were located as well as a
unique Matecumbe, or possibly Glades Tooled, Variant. This
type had the usual "pie crust" Glades Tooled rim but a cross-
hatched Matecumbe body. With the exception of the unusual sh-
erds the midden was typical of others in the vicinity.Bone
fragments for the most part ran to turtle, deer, and rodent.
The shell fragments consist of clam, conch, macrocalista, and
oyster. Pottery and shell tool distribution tended to show
occupation back to and including the Glades II Period.
Appreciation is expressed to the following: Site 1, the
Commandant and Provost Marshall of the Marine Air Base for
permission to trespass and excavate on government property,
Noel Herrmann for help with the screening and digging; Site 2
Mr. A. D. Barnes, Director, and Mr. Yank Taylor, Public Rela-
tions Administrator of the Dade County Parks Department for
permission to dig, Noel Herrmann and Bob Masters for aidingin
digging and Mr. N. J. Winkleman who first located the site ;
Site 3, Mr. Clyde F. Dunn of Coral Gables, owner of the pro-
perty, for permission to excavate, Neel Herrmann for help with
screening and sketches. Richard Kotil located the midden.
Bullen, Ripley P. and D. D. Laxson
1954. "Some Incised Pottery from Cuba and Florida". The
Florida Anthropologist",Vol.VII, No.1, pp. 23-4.
Goggin, John M.
1950. "Florida Archeology -1950". The Florida Anthro-
pologist, Vol. III, Nos. 1 & 2, pp. 9-20.
Goggin, John M. and Frank Sommer.
1949. "Excavations in Upper Matecumbe Key, Florida".
Yale Univ. Publ. in Anthropology, No. 41.
Griffin, John W., ed.
1949. The Florida Indian and His Neighbors. Rollins
College, Winter Park.
Figure 1. Map Showing Locations of Three
Small Dade County Sites
Depth In Inches
n-R A-192 12-1
n-R R-1 1-1ia
BELLE GLADE PLAIN 1 2 1 3 1
ST. JOHNS PLAIN 4 7 4 7
ST. JOHNS CHECK STAMPED 4 1_ 4 1
UNIQUE INCISED 1 1
KEY LARGO INCISED 4 2 6__
GLADES PLAIN (BODY) 106 96 306 190 57 50 7 469 336 7
GLADES PLAIN (RIMS) 11 15 46 29 9 5 66 49
OPA LOCKA INCISED 2 1 7 2 1 7
GLADES TOOLED 5 11 16
DADE INCISED 1 1
MIAMI INCISED 2 1 1 2 2
MATECUMBE (VARIANT) 2 1 1 2
Zig-zag motif pendant
MATECUMBE (VARIANT) 1 1 1 1
Glades Tooled rim,
BONE POINTS 2 3
STROMBUS CELTS 2 2 1 1 4
THE BARNHILL MOUND, PAUL BEACH COUNTY, FLORIDA
Ripley P. Bullen
The Barnhill mound ( PB-13 ) is located, immediately east
of Route 1, about half way between Delray Beach and Boca Raton,
Florida. During November, 1954, the author directed excava-
tions in this mound at the request of Mr. E. G. Barnhill of
Ft. Lauderdale, owner of the mound. Mr. Barnhill sponsored
labor and field expenses.
As shown on the sketch map ( Fig. 1 ), the mound is about
a hundred feet across, not allowing for recent erosion. Exam-
ination of the contours on this map indicates an upward slope
of the surface of the mound towards the west. Toward the
southeast the mound has a straight edge or side ( Fig. 1, near
6A) suggestive of the side of a ramp as found on temple mounds.
Extensions to the north, northeast, and southwest are rela-
tively recent features. Eliminating them, the Barnhill mound
has somewhat the appearance of a temple mound although the
present surface is not suggestive of such an edifice.
Disturbance of the surface of the mound has resulted from
four activities. A Japanese, who grew vegetables nearby, lev-
eled the top of the mound and built a house on it about forty
years ago. Mr. Gates of Boca Raton is of the opinion sand was
removed from the top of the mound during boom times in the
1920's. About fifteen years ago, Boy Scouts used to camp on
the mound and dig up skeletons, according to local residents.
The extension to the southwest has permitted automobiles and
light trucks to climb to the top of the mound.
These disturbances have kept ground cover to a minimum
although the sides, where steep, are well protected with trees
and bushes. The mound is built of fairly homogeneous, reason-
ably fine, whitish sand. Prevailing winds come from the
0 > 10 20 30 40
Figure 1. Sketch Map of the Barnhill Mound
Atlantic Ocean only a few miles to the east. They have moved
sand from the eastern part of the top of the mound and piled
it toward the west. As a result, the mound has much the ap-
pearance of a sand dune (Fig. 2).
Figure 2. Uncovering a shallow burial at the Barnhill Mound.
NOTE: increase in elevation of surface with dune appearance
toward upper right hand corner of picture.
Archaeological work at the mound-consisted of a series of
tests whose locations are given in Figure 1. Of these, only
the main excavation in the area of Square 9C reached the base
of the mound. Significant features discovered by these tests
and having to do with the construction of the mound will be
mentioned next. Squares will be referred to by their south-
eastern corner stakes.
The test in Square 16F was dug to a depth of five feet.
Roots and white sand were all that were found. The test
northwest of Stake 10G uncovered only white sand and occasion-
al flecks of charcoal down to a depth of 41 feet when excava-
tion there was discontinued. Test 12B was also sterile, at
least to a depth of 3 feet.
Several small tests were made along the edge of the pos-
sible ramp ( Fig. 1, near 6A ). From west to east in these
tests, a shallow charcoal zone was found at depths of 16, 18
to 24, and 24 to 28 inches below the present surface. This
Path o X.
-------------------- ~-- -
9K 9J 91 9H 9G 9F 9E 9
*<- 10 F ->-
Figure 3. Profile from the Barnhill Mound
charcoal impregnated zone is not definitive but it is sugges-
tive of a buried mound surface or surface of a ramp.
Excavation near the center of the mound ( Fig. 1, west of
Stake 10E ) uncovered evidence of an old mound surface. The
evidence consisted of a charcoal impregnated zone, 3 to 4
inches thick, which dipped downward toward the west from a
depth of 2 1/2 feet to a depth of 9 feet over a distance of 17
feet. Along the western face of Square 10E this charcoal zone
sloped downward toward the north 8 inches in 4 1/2 feet. Here
and at Stake 10E, where it joined the disturbed surface sand,
this zone was brownish in color and contained more humic mate-
rial than further west. It was suggestive of an old sod-line.
It was evident from the slopes of these zones that the
western and northern portions of the present mound had at some
time in the past been added to a previous mound. This addit -
ion, particularly to the west, was very substantial and ap-
pears to have about tripled the size of the mound. The north-
ern addition is shown as III in Figure 3.
9C 9B 9A 9AA 9BB
X Bundle burial
0 Isolated skull
'- Prone burial
It was noted during excavation that the sand of mound
was whiter and "cleaner" that that of the burial mound proper
(Fig. 3, Mound II). It may be proper to mention here that the
possible ramp points directly towards an Indian village area
about a half mile to the southeast. While it is not suscept-
able of proof, it seems very likely this substantial addition
(forming Mound III) was made to convert an older burial mound
into a temple mound.
That there had been a still earlier mound ( Fig. 3, mound
I ) was indicated by evidence found in the lower parts of the
main excavation (Fig. 1, Sq. 9C). Here was found the charcoal
impregnated base and charcoal impregnated southern slope of a
third mound. This situation has been illustrated in Figure 3
between stakes 9B and 9D. The charcoal zones were about 8
inches in thickness. The relative amount of charcoal was
greater toward the north and east than toward the south and
west. This zone disappeared completely near Stake 9B. We
only penetrated a small corner of the original mound (mound I).
That portion of it was sterile.
Figure 3 presents a profile along the "9" line. On it
III~ I-- 0L
0 x 0
X X )x x 0 x
x x X X s 70
0 a* X 7
OD U1 103 X X>
I X x
I-.X Bundle burial
V 0 O Isolated s1ull
tO E- Prone burial
Figure 5. Prone burial at Barnhill Mound
*/* ,I* -; -.^...:. -
1 -. i ) -" ^. ---Mk **
Figure 6. Three isolated skull burials at the Barnhill Mound
Figure 7. Bundle burial at the Barnhill Mound
"C''! LBJ~iiC~,a~ep$i;~nrIs~s~qr b~i
c ir -.,-
the space between Stakes 9E and 9F is a projection of that
found between O1E and 1OF (where digging did not penetrate
very deeply ). On this profile the relationships between the
three mounds are indicated as well as the relative positions
of burials. Presumedly, the primary mound ( mound I contains
one or more burials. We did not excavate enough of the pri-
mary mound to determine that point. All the burials we un-
covered were in the secondary mound (mound II) or intruding
into the top of the charcoal cap of the primary mound (mound
I) with the exception of Burials 12 and 13 (Fig. 4). They
appeared in the lower part of mound III just above the sloping
charcoal zone which separated mounds II and III. Otherwise,
as far as we could determine, mound III contained no burials.
Below the base of mound I was about 3 feet of grey sand
and then, to an undetermined depth, brown sand. A test made a
short distance away from the mound produced only white san d
and roots but no humic zone. The elevation of the base of
the primary mound is essentially the same as that of the sur-
rounding land. Apparently, sand from the surrounding land was
utilized in the construction of the mounds.
Relative vertical position of many of the burials has
been plotted in Figure 3 and the horizontal location of seven-
ty-five of them indicated in Figure 4. In Figure 4 only the
area within dashed lines was dug to base of mound. Condition
of skeletal material was extremely poor. Ninety-eight occur-
rences of human bones were recorded in the field. Analysis of
the data with the elimination of isolated bones, probable dup-
licates (bones nearby each other and at the same elevation or
found slightly below a bundle burial), and other questionable
occurrences reduced the number of interments to seventy-five.
Of these three are inconclusive as to type because we did not
dig far enough into walls to ascertain completely the kind of
burial represented. This leaves a total of seventy-two buri-
als which will be briefly discussed.
Burials were of three types: Isolated skull, bundle, and
modified prone interments. The last, represented by seven
skeletons in lower levels (Figure 3), was an ordinary prone
burial with the body lying on its back except that in every
case the lower legs had been bent backward at the knees (Fig.
5). This should place ankle bones near pelvis but in no in-
stance did we find an ankle or heel bone.
These burials were identified by a badly decayed skull, a
scrap or two of upper arm or shoulder bones, possibly a verte-
bra near the skull, bits of the two inominate bones, shafts of
the femora (noticably knock-kneed), and the shafts of the ti-
biae and fibulae lying a little outside and a bit deeper than
the femora (Figure 5). Sometimes the long bones of the legs
were only represented by dark yellow-brown stains in the sand.
Hips were lower than skulls and knees usually slightly
lower yet. This is interesting when the locations of these
burials on Figures 3 and 4 are noted. All are oriented ap-
proximately northeast-southwest and located-in mound II a
short distance above the sloping side of mound I. They may
have been lain upon a sloping surface covering mound I or
placed in small pits dug into such a surface. All had their
heads to the northeast except Burial 96 which was reversed.
Following the argument presented above, his legs were higher
than his skull.
Bundle burials were of usual forms. The skull might be
on top of, at one end of (Fig. 7), or even below the long
bones. Never did there seem to be enough bones present to re-
present a complete skeleton. However, this might be a func-
tion of preservation. When the skull was on top of long
bones, only the lower part of the skull was present as bone.
The upper part being represented by dark brown sand.
Isolated skull burials were represented by spherical
masses of dark brown sand (Fig. 6). Frequently all that was
left as bone were the mastoid processes and a few teeth.
It is obvious from the above comments and from the illus-
trations (Figs. 5-7) that identification of the form of buri-
al in the field was difficult. The same applied to identifi-
cation of age and sex. The following generalizations are made
with some reservations but they should present an approxima-
tion of the correct situation.
Examination of Figures 3 and 4 suggest certain conclu-
sions. Bundle burials seem to be concentrated at shallow
depths in the near Suqare 9D. This would seem to represent a
concentration of bundle burials at or near the top of mound II.
In Square 9D these burials were found from immediately below
the present surface downward. In Squares 10D and O0E burials
were not encountered until below a depth of 6 inches. This
may be because of the beginning of mound IIIat about this lo-
cation or may reflect Boy Scout activities mentioned earlier.
South of the "D" line burials were not found above a depth of
2 feet below the present surface. Again, this may be the re-
sult of Boy Scout or other activities but there was good
ground cover over Squares 9B and 9C. This should eliminate
the question of very recent disturbance in these areas.
Horizontally, there seems to be an avoidance for more or
less shallow burials of Square 90 or the area above the modi-
fied prone burials (Fig. 4). However, we did not reach the
base of mound II in Square 9D so that it is entirely possible
more prone burials might have been found in the square Burial
51, shown for Square 9D was at a depth of only 3 feet to the
highest part of the skull. The highest part of Burial 103,
shown beside Burial 51, was at a depth of 6 feet 9 inches.
It has been suggested in Figure 3 that bundle burials had
the shallowest average depth and prone burials the greatest
with isolated skulls intermediate in this respect. Plotting
our seventy-two burials vertically substantiates this. Bundle
burials were found from the surface downward to a depth of 60
feet, isolated skulls from a depth of 6 inches to one of 7
feet, and prone interments between depths of 3 feet and 6 feet
9 inches. Using a depth of 3 feet as an arbitrary reference
plane, thirty-two bundle burials were above and only 10 below,
eleven isolated skulls were above and twelve below, while all
seven prone interments were below. This suggests a change in
burial habits during the useful life of mound II.
A similar change has been suggested by the author for the
Tampa Bay area of Florida (Bullen, 1952, pp. 82-3). If this
is correct and if such cultural changes occurred more or less
simultaneously on both the east and west coasts of Florida, it
would seem proper to correlate mound II with the Weeden Island
period of the Florida gulf coast. A time span of roughly A.D.
700 to 1200 might be suggested for mound II at the Barnhill site
Due to the poor condition of skeletal material very lit-
tle can be said of the individuals whose remains were excavated
from mound II. Teeth of very young children as well as adult
teeth showing substantial wear were noted. In one instance,
an upper and lower jaw contained deciduous teeth being pushed
out by permanent ones. One of the most complete bundle buri-
als (B-12) appeared to represent a female about 20 years age
at death. Skulls were large, small, and medium sized. The
relative size of long bones varied similarly. All of this
suggested a normal population with burials of all ages and
Burial 93, the interment (?) of an isolated skull was
unique in certain features. It was first encountered at a
depth of 7 feet just north of Stake 90 (Fig. 4). The skull
was upright in the ground with the face to the southwest. The
upper part of this skull had completely eroded away leaving
only dark brown sand to show its location. The bottom of the
basal portion exhibited a yellowish color not noted in other
cases. Directly below was a brownish area, 4 to 5 inches in
diameter. We traced this stain vertically downward a distance
of 21 inches during which it tapered a little (Fig. 8). At
the bottom it suddenly widened 6 inches to the west and then
This phenomenon is hard to explain. The skull was at a-
bout the same relative depth as the prone burials. It was al-
so a short distance above the edge of mound I. Possibly a
skull had been placed at this point on a post--perhaps a small
Figure 8. Discoloration below isolated skull burial, Burial 93
(Lower part of skull replaced for photograph.)
tree had been cut off to mount the skull. Maybe the skull was
painted or contained paint which, over centuries, produced the
stain found below it. While the answers to these questions
remain unknown, some ceremonial implication seem unavoidable.
VILLAGE SITE (PB-12)
About a half mile southeast of the Barnhill mound are the
remains of an Indian village. We made a small test which un-
covered a black dirt midden containing clam, oyster, conch,
and "rock snail" (probably Livona, sp.) shells. Other food
remains included turtle, deer, and fish bones. At a depth of
30 inches water entered our test. No sherds were found in the
test but undecorated gritty sherds are known for the site.
This fairly heavy occupational debris may well represent
the village of the builders of the Barnhill mound. It is now
located in a fairly low and damp area but such may not have
been the case when the site was occupied.
Unfortunately we were unable to completely excavate the
Barnhill mound. If we had, some of the questions raised by
the data from our excavations might have been answered. As it
is many must wait until further work is done at this or a
In all our digging in the Barnhill mound we did not find
a single artifact--not even a sherd, a chert chip, or a scrap
of shell--in the mound fill. Because of roadways, the only
place we could test for a special pottery cache was to the
southeast. Our work in that area was in no sense complete but
several tests (not shown in Figure 1) failed to disclose such
Our excavations revealed the Barnhill mound to consist of
three superimposed mounds. The first, with a specially pre -
pared charcoal-impregnated base and a similar cap, was probab-
ly built during Glades II times (Goggin, 1950, p. 10) for an
important interment (Fig. 3, 1). Our work was not sufficient-
ly extensive to determine the purpose of this mound.
The second mound, which enclosed mound I and extended a
considerable distance to the south, was also constructed dur-
ing Glades II times. In it we found primary burials, lying on
their backs with their lower legs bent backwards, isolated
skulls, and bundle burials. Prone burials had the deepest ,
isolated skulls the next deepest, and bundle burials the shal-
lowest average vertical distribution (Fig. 3). We believe
this represents a change in burial habits during the life of
mound II. However, there seems to have been an avoidance of
the area over the prone burials for later interments. This
avoidance may have ceremonial implications.
The change in burial forms with depth is similar to that
found in the Tampa Bay area of the west coast of Florida (Bul-
len, 1952). In that area such changes seem to have transpired
during Weeden Island times. Weeden Island times are correlat-
ed with late Glades II and early Glades III times of south-
eastern Florida (Goggin, 1950, p. 10). If this line of argu-
ment is correct, mound II should date sometime between A.D.700
and A.D. 1200.
Finally a very extensive addition was made to the mound
(Fig. 3, III) to form what may have been a temple mound. Such
an edifice should date to the Glades IIIB period of southeast
Florida. The corresponding periods on the Gulf coast, when
temple mounds are common, are the Safety Harbor period of the
Tampa Bay region and the Ft. Walton period of West Florida.
This period should be post-A.D. 1200--more likely A.D. 1300 or
a little later in east Florida (Bullen, 1956, p. 33). Such
mounds are also found in the lower St. Johns River area during
Due to the lack of any artifacts we know nothing of the
everyday life of the makers of the Barnhill mound. Undoubted-
ly some of them lived in the village located about a half mile
southeast of the mound.
Bullen, Ripley P.
1952. "Eleven Archaeological Sites in Hillsborough
County, Florida." Report of Investigations, No.8,
Florida Geological Survey. Tallahassee.
1956. "Some Florida Radiocarbon Dates and their Signifi-
cance." The Florida Anthropologist, Vol. IX, No.
2, pp. 31-36.
Goggin, John M.
1950. "Florida Archeology -- 1950." The Florida Anthro-
pologist, Vol. III, Nos. 1-2, pp. 9-20.
Florida State Museum
THE NORMAN MOUND, McINTOSH COUNTY, GEORGIA
Lewis H. Larson, Jr.
During the winter months of 1953-1954, the Georgia Histor-
ical Commission carried out a series of excavations on the
Georgia coast in an effort to define the cultural position of
the pre-Spanish, Guale Indians. An archaeological survey made
by the Commission indicated that the central area of the Geor-
gia coast had been occupied immediately before the European
contact period by a group of Indians whose ceramic complex was
similar to, but not identical with, that found at the Irene
site to the north on the Savannah River. On the basis of docu-
mentary evidence and typological similarities to the known
Guale pottery of historic times, this ceramic complex has been
called the Pine Harbor Complex, using the designation of the
type site located near Pine Harbor in McIntosh County. The
Pine Harbor Complex includes the Irene Plain, Irene Filfot
Stamped, Irene Incised, and McIntosh Incised pottery types.
With the exception of McIntosh Incised, all of the types named
occur at the Irene site during the last occupation, and are re-
cognized by Caldwell and McCann as constituting the Irene
Complex (Caldwell and McCann 1941: 1-3).
A comparison between material recovered by Clarence B.
Moore from mounds in McIntosh County ( Moore 1897: 15-73 ),and
similar items from the excavations by the Georgia Historical
Commission at the Pine Harbor site led to the belief that fur-
ther data concerning the mortuary practices and religious com-
plex of the Guale could be obtained by the excavation of a
burrial mound. Accordingly, a small mound on the property of
Mrs. Roscoe Norman, Belleville Point, McIntosh County was se-
lected (Fig. 1). The excavation was undertaken for the Georgia
Historical Commission under the direction of the author.
Figure 1: Map of the Sapelo River area, McIntosh Co., Georgia
(1) Norman Mound, (2) Pine Harbor Site, (3) site
on the north end of Creighton Island.
The mound stood on a slight rise some ten feet above the
high tide mark. The ground level sloped gently toward the
marsh, one hundred and fifty yards to the north. This marsh
lay between the high ground on which the mound stood and the
Sapelo River, a tidal stream. In a line due north of the mound,
but on the opposite side of the Sapelo River, is the Pine
...:.:..;..:. .... .-.
Harbor site. Small scattered marine shell middens cover the
area in the vicinity of the mound, offering ample evidence of
an aboriginal occupation.
.: .. ...... .....
The mound itself, was small and conical in shape, only
fifty feet in diameter and three feet high. In all probability
it had suffered much erosion over the years because of its san-
dy construction. If one judged by the size of the live oak
trees which stood around and on the mound, the area had never
been under cultivation. Considerable surface disturbance to
the mound had taken place in recent years due to the hogs which
ran loose in the area. A small hole in the top of the mound
had been made by treasure hunters.
The excavation of the mound was controlled by the conven-
tional five foot grid system. An area fifty feet square was
laid out to include all of the mound, and this area was com-
pletely excavated, except for a small portion on the south-
eastern margin of the mound where a tree stood. At the request
of the owner of the land, the mound was not restored after ex-
It was found that the mound was constructed of fine local
sand placed over a core mound of marine shell. The sand was
concentrated on the flanks of the shell mound with only a thin
layer of it covering the summit. Presumably, much of the sand
covering the top of the mound had washed or blown away. The
shell core was resting on an old occupation level composed of a
light sand. The burials lay over the shell core, and in the
sand which covered it.
When first encountered, it was thought that the shell core
was a central tomb or a covering for an earlier mound construc-
tion phase, such as Moore encountered in the Hopkins and Walker
mounds ( ibid., 43-5 ). This did not prove to ue the case, for
further investigation determined that the shell core was simply
a shell midden, the surface of which had been used for mortuary
purposes. No burials were found in the shell, or intrusive in-
to the shell, and all the ceramic evidence indicates that the
shell heap was constructed as the result of village activity
before it was used a.s a depository for burials.
The burials themselves were numerous, considering the rel-
actively small area in which they were contained. Almost every
type of burial was found in the mound, including: primary, bun-
dle, and urn burials, along with several cremations. These in-
terments were placed on top of, or around the shell core, with
little regard to orientation or burial position ( Figs. 2, 3A).
Almost without exception, the burials, when placed either on
the side or top of the shell core, were put in a shallow dep-
ression which had first been lined with clean sand. Burials
made beyond the margins of the shell core were commonly placed
directly on the surface of the ground, which had been prepared
by the removal of two or three inches of sandy humus. All of
the burials, whether on the shell core, or on the surrounding
ground, were then covered with clean sand.
From the complete lack of order in arranging the burials
with respect to one another, it is evident that all of the
bodies were not buried simultaneously. Rather, it appears that
they were added to the mound over a period of time. The mound
is not a deliberate construction effort, but the gradual accu-
mulation of graves built up around an old shell heap. It
should be noted, that in several places in the mound, a number
of bodies were found together, and it is assumed that they were
interred at the same time.
Most of the primary burials were lying on the right side
and tightly flexed, perhaps having been wrapped in mats or
cloth. One burial ( no. 24 ) had tiny fragments of charred or
decayed organic material scattered through the grave, probably
the remains of such wrapping. 6ome of the burials were so
tightly flexed as to suggest partial decomposition before
burial, e. g. burial 11 (Fig. 3B).
The cremations were of three types: (1) cremation of the
body or bones in the place where they were found, (2) the buri-
al of bone cremated elsewhere and deposited in the mound, ( 3 )
the burial of cremated bone in pottery vessels.
There was one primary infant burial in a pottery jar
( burial no. 7 ), and at least one definite bundle burial was
found (burial no. 16). One of the primary burials (no. 12) was
Figure 2B: Excavation plan of the Norman Mound, with the
locations of burials in relation to the shell core.
exceptional in that it had been placed on a small bed of oyster
shell rather than the usual sand. It was covered with a small
heap of the same shells. Burial 7, the urn burial, and burial
13, the primary burial of a child, had been made on small lay-
ers of oyster shell and then covered with sand.
With the exception of the pottery vessels associated with
cremations and the one infant urn burial, grave goods were
found in only six other burials. This material included: two
knobheaded shell ear-pins found with burial 6, a plain pottery
elbow pipe with burial 18, a Savannah Burnished Plain bowl with
burial 17, a plain pottery elbow pipe, shell disc beads, a bone
"dagger", and a, small crude, steatite, human effigy with burial
24, and shell disc beads with burials 25 and 28 (Fig. 4A).
Burial 17 contained some charred wood, apparently the re-
mains of a log which lay across the body.
In all, there were some thirty-one burials, containing the
remains of at least thirty-six individuals. These burials are
summarized in Table I.
POTTERY (Fig. 4B)
The pottery vessels found with burials were of three
types. The most popular type was Savannah Check Stamped. In-
complete vessels of this type were found with burials 14 and 23,
and an example occurred under burial 20, but not in association
with the burial. In all, four partial Savannah Check Stamped
vessels were recovered. Two broken Savannah Burnished Plain
vessels were found, one each with burials 14 and 17. A single
example of an Irene Filfot Stamped vessel was found with burial
7, the primary urn burial.
Sherds were recovered from three distinct areas in the ex-
cavation, the sub-mound sand, the shell core mound, and the
sand fill over the burials. There was a total of 533 sherds
recovered from these three locations. Of this number, 431
sherds were identified and classified as to a particular type.
There were 103 sherds or 19.2 percent of the total that were
unidentified. There were 67 sherds found in the sub-mound sand,
254 sherds found in the shell core, and 110 found in the sand
Figure 3: A. General view of burials grouped around the shell
core. View is towards the east.
B. Burials 11 and 12. Burial 11 is in the upper
C. Burial 9.
BURIAL NUMBER 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
NO. OF BODIES IN GRAVE 2 11 1 2 1 1 2 1 2 1 1 1 1 2 1 1 1 7 1 1 1 1 1 1 ? 1
flexed on right side x x x x x x x x x x x x x x x
flexed on left side x
flexed on back x x x x
position indeterminate x x x x x
PRIMARY URN BURIAL x
BUNDLE BURIAL x
In pottery vessel x x
in place x
deposit of burned bone x
adult x x x x x x x x x x x x x x x x x
adolescent x x
infant x x
GRAVE GOODS (other than pottery) x x x x x
POTTERY VESSELS IN GRAVE x x x x
*A field error omitted a burial 21.
T AliFlB TRAITS
fill over the burials which could be assigned to a specific
pottery type. The types recognized were a plain fiber tempered
ware ( St. Simons Plain) ( Caldwell and McCann 1941: 51 ),Wil-
mington Heavy Cordmarked, Savannah Fine Cordmarked, Savannah
Check Stamped, Savannah Burnished Plain, Savannah Complicated
Stamped, Irene Plain, Irene Filfot Stamped, Irene Incised, and
McIntosh Incised. The percentages of each type, with respect
to their location within the mound, are summarized in Table II.
The unidentified sherds have been omitted from the table.
On the basis of sherds found in the sub-mound sand, it
appears that the site had been occupied from the Deptford,the
succeeding Wilmington, Savannah I and Savannah II ceramic
periods as they have been defined for the coastal area of
(Caldwell 1952: 315-19 ). The few fiber tempered sherds fonmd
in the shell core extend the occupation of the site back even
further. This type has been found elsewhere on the Georgia
coast in an earlier stratigraphic position than the Deptford
wares ( ibid., 312-14 ), and undoubtedly the location of these
sherds in the Norman Mound belies their true chronological
significance. The plain sand tempered ware makes up the third
highest percentage of sherds from the sub-mound level. This
type, while it has yet to be recognized as a type, is probably
related to the Deptford wares. The Irene sherds of the Pine
Harbor Complex, that were found in the sub-mound sand, are so
few in number ( 7 sherds ), that their presence in this lower
level may be regarded as an excavation error.
The marked increase in the percentage of the Savannah
Check Stamped sherds, and the decrease in Savannah Fine Cord-
marked sherds in the shell core, would place the construction
of the shell core during the Savannah II Period.
In the sand fill over the burials, we have significant in-
creases in the percentages of the Pine Harbor wares, and a
corresponding decline in the percentages of all of the earlier
ceramic types, with the exception of the Savannah Check Stamped
type. This latter ware decreases only some 5 percent under the
SHELL CORE SAND FILL
---------------- ----- ----
........... t. ------ ----- .-- ----- 4- .......
PLAIN FIBER TEMPERED
nEPTFORD SIMPLE STAMPED
PLAIN SAND TEMPERED
WILMINGTON HEAVY COPDMARKED
SAVANNAH FINE CORDMAPKED
----- SAVANNAH CHECK STAMPED
SAVANNAH BURNISHED PLAIN
SAVANNAH COMPLICATED STAMPED
IRENE FIIFOT STAMPED
amount found in the shell core.
The period during which the burials were made, and the
sand fill was added to the mound, is not easy to determine.
With the exception of the burial urn with burial 7, all of the
pots found in the mound can be assigned to the Savannah II
Period. In addition there is the Inrge percentage ( 53.6% ) of
Savannah Check Stamped in the sand fill. On the other hand, if
the burials had been made during the Savannah II Period, we
would not expect the sand fill to contain any Pine Harbor
sherds, and yet about 33 percent of the total sherds from this
level can be attributed to that period (some of these Pine Har-
bor sherds were found in the sand fill immediately around buri-
als ). However, if the burials had been made during the Pine
Harbor Period would we find Savannah II pots with the burials,
or so high a percentage of Savannah Check Stamped sherds in the
sand fill over the burials?
The artifacts found associated with the burials are of
little value in the assignment of a temporal position for the
burials. The shell disc beads, the steatite effigy, and the
bone dagger have little diagnostic value for this situation.
Shell ear pins were found with both Savannah II and Irene buri-
als at the Irene Mound ( ibid. 1941: 54), as were pottery elbow
pipes ibidd., 53)
There are a number of ways to explain the chronological
position of the burials. It is possible that the burials were
made during the transition between the Savannah II and the Pine
Harbor Periods. Typologically there is evidence that the cer-
amic forms of the Irene and Pine Harbor Complexes developed out
of the earlier Savannah II forms ( ibid., 42-3). At the Norman
Mound the elements which Caldwell and McCann recognize ( ibid.,
42, Fig. 16 ) as criteria of this transition at the Irene Site
are not present in the ceramics.
It should be noted, however, that the transition between
Savannah I and the Irene Complex at the mouth of the Savannah,
and its equivalent the Pine Harbor Complex at the mouth of the
Altamaha, is at present inadequately known, and there are many
Pottery found in the mound. From left to right,
upper row: Deptford Simple Stamped, Wilmington
Heavy Cordmarked, and Savannah Fine Cordmarked;
center row: Savannah Burnished Plain and Irene
Filfot Stamped; bottom row: Savannah Check
Stamped (burial 23) and Irene Filfot Stamped.
Figure 5A: Pottery pipe found with burials.
Figure 5B: Shell beads, shell pin, and pottery effigy.
The most plausible explanation for the burials seems to
be simply that the mound was used for burial purposes during
both the Savannah II and the Pine Harbor Periods. During the
excavation it was impossible to determine any stratification
of burials, because of the nature of the sand in which they
were buried. Nevertheless, the multiplicity of burial forms,
the lack of orientation, and the disregard for the placement
of the burials indicates that the burials were made over a
period of time, sufficient in length, to have extended from
Savannah II into the Pine Harbor Period.
In many respects the Norman Mound recalls the archaeo-
logical situation encountered in the burial mound at the Irene
site. At Irene the burial mound "..... was very low, circular,
and composed of sand and shell. The diameter was about fifty-
five feet and the height about two and one-half feet. Structu-
rally it consisted of a central shell deposit and a series of
flanking shell layers separated by sand fills. ..... The pre-
mound surface and the central shell deposit contained only
cremated secondary burials and pottery of the Savannah complex,
while the flanking shell layers contained chiefly flexed pri-
mary burials and some pottery of the protohistoric Irene com-
plex as well as that of the Savannah." ibidd., 22)
In both mounds there was a central shell core flanked by
burials. In the Irene burial mound burials were below, and in
the shell core, such was not the case with the Norman Mound.
The flanking layers of shell, though present at the Irene
mound, were absent at the Norman site. On the other hand, in
both mounds,the cremations were in direct association with the
shell core. At Irene all of the cremations are assigned to
the Savannah II Period. At the Norman Mound we can assign at
least two of the four cremations to this same period onthe
basis of the pottery vessels which contained the burned bone.
At Irene, the Irene Period burials were confined to the
flanks and to the margins of the mound. The one burial from
the Norman Mound which we can definitely attribute to the Pine
Harbor Period, was in Irene Filfot Stamped vessel, and was at
margin of the mound.
While the data is not entirely clear about the situations
encountered, C. B. Moore seems to have found one mound in
McIntosh County which also revealed burials from.the Savannah
II and the Pine Harbor Periods, e. g. the mound at Bourbon,
Sapelo Island ( Moore 1897: 55-65). In the mound at Bourbon,
Moore found "a vessel with incised decoration below the margin
and a faint complicated stamp on the body" ( very likely an
Irene Incised bowl) ibidd., 59), and "a small bowl ..... bear-
ing the checked, stamped decoration" (Savannah Check Stamped?)
The Norman Mound can be characterized as a low, conical
burial mound which developed around a shell midden constructed
during the Savannah II Period. Previous to the build-up of
the shell midden, the site had been occupied more or less con-
tinuously from the fiber-tempered ceramic period through the
Deptford, Wilmington, and Savannah I ceramic periods.
Following the construction of the shell midden, the site
was utilized for mortuary purposes. Some thirty-one burials
were deposited on and around the shell midden. Little or no
formal arrangement of burials took place, and there was con-
siderable variation in the type of burial. Burial goods were
scanty and occurred in very few graves.
In view of the available data it is felt that the burials
were made during the final two ceramic phases at the site, the
Savannah II Period and the Pine Harbor Period. This conclusion
is borne out by similar archaeological situations at an-
other mound in McIntosh County and by the burial mound at the
Caldwell, J. R.
1952 The Archeology of Eastern Georgia and South
Carolina. In Archeology of Eastern United
States, ed. by J. B. Griffin, pp. 312-21,
Univ. Chicago Press, Chicago.
Caldwell, J. R. and Catherine McCann
1941 Irene Mound Site, Chatham County,
Univ. Georgia Press, Athens.
Larson, L. H. Jr.
Moore, C. B.
Unusual Figurine from the
The Florida Anthropologist,
pp. 76-81, Gainesville.
Certain Aboriginal Mounds of the Georgia
Coast. Journal of the Academy of Natural
Sciences of Philadelphia, second series,
Vol. 2, pt. 1, pp. 4-138, Philadelphia.
1. The author is indebted to a number of in-
dividuals for their cooperation and assis-
tance, particularly Mrs.Roscoe Norman for
permission to excavate on her property, Mr.
and Mrs. William Barnett for their gracious
hospitality on some particularly cold days,
and to Miss Bessie Lewis for her constant
willingness and ability to solve so many
Contributors To This Issue
D. D. Laxon is one of the busiest members of the Florida Anthropological Society
and Vice President of the South Florida Chapter. In this issue he presents the
results of salvage archeology he has been doing in the Miami area for the past
Ripley P. Bullen, Curator of Social Sciences at the Florida State Museum, Gaines-
ville, is probably known to more members of the Society than any other member.
For seven years he labored long and hard as Treasurer but managed to conduct
a number of excavations as is indicated by this report.
Lewis Larson for a number of years has been archeologist for the Georgia His-
torical Commission. The present paper indicates some of the archeological
material north of Florida along the Georgia coast. He is at present digging at
the well-known Etowah Mounds, near Cartersville, Georgia.
Florida Anthropological Society
Membership is open to all interested in the aims of the society; regular dues,
$3.00 a year; student dues, $1.50 a year. Members receive The Florida An-
thropologist, the Newsletter, and Publications as issued. Applications and orders
for back issues should be sent to the Treasurer (each single number to members,
$.50; each double number, $1.00; to non-members, $.75, and $1.50 respectively;
Newsletters Nos. 1- 35, $.15 each). General inquiries should be sent to the
Secretary, manuscripts to the Editor, and Newsletter items to the President.
1st Vice President:
2nd Vice President:
William J. Armistead, 2413 Watrous Ave.,
John W. Griffin, St. Augustine
Irving Rouse, New Haven, Conn.
Marvin J. Brooks, 805 N.W. 15th Court, Miami 33
Hale G. Smith, Box 3051, Florida State Univ.,
Charles H. Fairbanks, Box 3051, Florida State Univ.
Hugh N. Davis, Miami
H. James Gut, Sanford
William H. Sears, Gainesville
PUBLICATIONS OF THE SOCIETY
No. 1. "Two Archaeological Sites in Brevard County, Florida," by Hale G.
Smith. 32 pages, 4 plates _-...--.---------..---------------- 0.50
No. 2. "The Safety Harbor Site, Pinellas County, Florida," by John W. Griffin
and Ripley P. Bullen. 42 pages, 4 plates .. ------..- 0.50
No. 3. "The Terra Ceia Site, Manatee County, Florida," by Ripley P. Bullen,
48 pages, 7 plates -. .......----- ----------- ---------------------- 0.50
No. 4. "The European and the Indian," by Hale G. Smith. 150 pages,
frontispiece, 6 maps ........------................---------- -----------------------. 2.00