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 SILVER SPRINGS SITE, PREHISTORY IN THE SILVER
SPRINGS VALLEY, FLORIDA
E. Thomas Hemming s
In north peninsular Florida, especially in Marion County, artesian
springs of relatively great magnitude have strongly affected (and still
influence) human settlement and land use. According to the Florida Geo-
logical Survey, "... the primary vents of both Silver and Rainbow Springs
are believed to have larger average flows than any limestone spring in the
United States and possibly in the world...the total average flows down
Silver Springs Run and Rainbow River [are] 808 and 699 second-feet respec-
tively" (Ferguson and others 1947). Not surprisingly, the valley of Silver
Springs Run is an archaeological zone of great interest, where more-or-less
scientifically oriented investigations extend back nearly 80 years. The
purpose of this paper is to describe the results of a 1973 excavation by
Florida State Museum at the Silver Springs site (MR92), to review critically
earlier and ongoing archaeological projects in the Silver Springs valley, and
finally to suggest the potential for future work at Silver Springs.
Silver Springs Run winds five miles eastward from its headspring to
join the Oklawaha River, a major tributary of the St. Johns River system.
Above the mouth of Silver Springs Run the Oklawaha has essentially no
flow in drought periods, while below the mouth a substantial minimum flow
is assured by the nearly uniform, perennial, spring discharge (Mifflin 1970).
The valley of Silver Springs Run includes three geomorphic zones: (1) the
meandering channel itself, 100 to 200 feet across and reaching 30 feet in
depth, (2) a low quarter-mile-wide floodplain, and (3) the valley walls, some-
what steeply sloping, rising about 20 feet to upland terrain (Fig. 1). Major
vegetation types present today in the valley include aquatic communities
along the run, hydric hammock and swamp forest along the floodplain, mesic
hammock on valley slopes, and pine flatwoods or long-leaf pine-turkey oak
stands on upland terrain (Lugo and Carr 1970). The modern wildlife of the
Silver Springs and Oklawaha River valleys has been described as "...out-
standing in abundance [and] variety. ." (Layne 1970:35; also Neill 1958;
Reid 1970). For the sake of brevity this fauna will not be enumerated here
except where particular species occur as archaeological remains in the
Silver Springs area. A sequence of alluvial deposits is described below
with reference to Silver Springs and other archaeological sites in the valley.
The Silver Springs Site
The Silver Springs site (MR92) is located one-half mile east of the
headspring on elevated terrain ("Paradise Park") adjoining the south bank
of the run. The south bank in this area is characterized by upland terrain
Florida Anthropologist, vol. 28, no. 4, December 1975
Fig. 1. Upper Valley of Silver Springs Run showing locations of the headspring,
channel, floodplain, upland terrain, and selected archaeological sites.
closely bordering the channel, by little or no intervening floodplain swamp,
and by a concentration of archaeological sites (Fig. 1).
The discovery and initial testing of the Silver Springs site in the
early 1950's were reported by Wilfred T. Neill (1958), who was then assoc-
iated with Ross Allen's Reptile Institute. As a stratified occurrence of
Paleo-Indian and later artifact assemblages, Silver Springs was"...exceedingly
important because it provides one of the few undisturbed stratigraphic proofs
of the... assumed pre-Archaic status of fluted points in the eastern half of
the continent..." (Mason 1962:240). Neill s report indicated further that
essentially horizontal occupation levels, representing buried land surfaces,
occurred in deep eolian sand deposits at Silver Springs. In spite of these
A <-- permanent benchmark
3 50 10
borrow pit floor
* telephone pole
Fig. 2. Location of 1973 test excavations at MR92.
important indications no additional excavations were undertaken for 20 years,
chiefly because of the paucity of artifacts obtained in Neill's tests.
In late February-early March 1973 Florida State Museum carried out
a new exploratory project at the Silver Springs site designed to corroborate
or modify Neill's findings and seek additional evidence for the aboriginal
use of Silver Springs. Two stratigraphic tests, totaling 3800 cubic feet,
were placed along an old sand borrow pit face in the vicinity of Neill's
tests A and F (Fig. 2; Neill 1958:37). The specific objectives of this pro-
ject were to examine the stratigraphic sequence, to obtain artifact samples
for all levels, to investigate any horizontal occupation "floors, to locate
Paleo-Indian occupation levels (which Neill described at about 7 to 8 feet
below the modern surface), and finally to explore for even earlier remains
in the Silver Springs deposits.
Initially, slumped sand was removed from the borrow pit face in order
to examine and record undisturbed profiles (Fig. 3). Excavation was then
carried downward from the surface by horizontal 6-inch levels, locating
large artifacts or clusters of artifacts in place and recovering smaller
debris from a mechanical shaker screen (Fig. 4). After several annoying
collapses, the excavation walls were sloped markedly from vertical. The
areas excavated were mapped by transit and related to local topography, a
benchmark, telephone poles, and fencelines (Fig. 2). The schematic
geologic cross-section at MR92 (Fig. 5) is based on our excavation and on
reconnaissance of the Silver Springs area during extensive construction project
along the floodplain and on upland areas in 1973-74. In a site which en-
compasses several thousand years of human occupation the geological context
of remains must be thoroughly understood. The interpretation presented
here is a preliminary one, and will certainly be improved as additional
geological studies are completed.
The Ocala limestone (Unit C), the principal aquifer underlying this
area, is exposed at or near the surface in the vicinity of the headspring, alor
the upper bed of the run, and at the "toe" of the valley slope for at least a
mile downstream. Cherty boulders and tabular masses of chert at the
weathered upper boundary of the Ocala limestone were available for aboriginal
tool-making in these areas, especially when spring levels were lowered.
Overlying the Ocala is a varicolored clay (Unit B) which is largely absent
in the trough of Silver springs valley. This Hawthorne clay was exposed
recently by deep excavations in upland areas immediately southwest of
MR92, and minor lenses are visible in the banks of a new canal along the
outer edge of the floodplain. Relict chert boulders are frequent inclusions
in this clay at both exposures and were undoubtedly quarried in certain
areas near the run. Pleistocene and Recent sediments, which contain a
record of human occupation in the Silver Springs area, include two groups
of deposits underlying the valley slopes (Al and AZ) and floodplain (A3 and
A4) respectively (Fig. 5).
Unit A1 is a homogeneous eolian sand without cross-bedding or other
structure. The equivalent of this sand has been described by Hemmings and
Fia. 3. Profile in the lower sand (Unit A,) showing strong seepage lines.
Fig. 4, Excavation of Test 1 from the surface; Silver Springs floodplain
in the background.
Kohler (1974) at Lake Kanapaha in Alachua County. The maximum thickness
observed for Unit Al is 8.5 feet. Based on archaeological remains, this
deposit has accumulated over the past 10,000 years. Bones of white-tailed
deer were collected from the upper part of Unit Al, but for the most part
organic materials have not been preserved in the sand. Although Neill
(1958:35) describes charcoal in association with flakes at various levels
in the Silver Springs Site, this observation was not corroborated in our
excavation. Occasional charcoal flecks, dispersed and concentrated in
burrows, are believed to derive from the recently formed humic zone.
Fragments of a fulgarite (fused irregular tubes of silica resulting from a
lightning strike) were excavated at 6. 0. 5 feet in Unit A1, and presumably
represent a thunder storm several thousand years before present.
Site Mr 92
'.I borrow pit
Floodplain Silver Springs Rui
Unit A, Upper Sand
Unit A. Lower Sand Unit B Hawthorne Clay- Plio-Pleistocene
Unit A3 Peaty Muck and Recent Unit C Ocala Limestone-Eocene
Unit A4 Marl Sequence
Fig. 5. Schematic geologic cross-section across the valley of Silver Springs Run.
A lower sand or clayey sand, Unit AZ, contains a sparse admixture of
limestone gravel or rubble and is strongly marked by seepage lines (clay
and iron oxide partings), referred to as "laminations" by Neill. This sand,
exceeding 7 feet in thickness, was probably deposited partly by wind action
and partly by slope-wash from higher terrain to the south. The seepage
lines represent post-depositional effects of groundwater (Fig. 3). Where
Unit AZ occurs at lowest elevation, it intergrades slightly with floodplain
deposits. From such a location, immediately above Ocala Limestone,
fossilized bones of white-tailed deer, alligator, red-eared sunfish, and
(tentatively) extinct horse were collected. For the most part, however,
Unit AZ is not fossiliferous. The upper boundary of this sand is gradational
into Unit Al and the lower boundary is generally not exposed. The age of
this deposit is believed to be greater than 10,000 years B.P. and probably
Unit A3 is a dark-brown peaty muck at the surface of the floodplain,
presently exposed above spring level, which has formed in the recent past
(and therefore correlates with the upper part of Unit Al).
Unit A4 includes a sequence of shelly marls which are localized in
occurrence along Silver Springs Run and outcrop below spring level. As
these marls are the context of the Guest Mammoth site (MR130), currently
being investigated, they will not be described in detail here (Brooks 1973;
Hoffman and Brooks 1974). The marl sequence reaches 12 feet or more
in thickness and correlates with the lower part of Unit Al and AZ (greater
than 10,000 to about 5000 B.P.). These marls were presumably deposited
in meander cut-offs and sloughs along the middle and lower reaches of
Silver Springs Run at times of lower spring levels. The search for underwater
sites with associated artifacts and extinct animal remains is of necessity
directed to marl outcrops in eroding locations (cut-banks) along the modern
channel. Figure 5 is a composite cross-section inasmuch as the marl
Table 1. Excavated Levels and Artifact Counts in Tests 1 and 2.
Levels* Test 1 Test 2
(feet) Sherds Stone Tools Debitage** Sherds Stone Tools Debitage**
0-1 0 1 ------ 0 1
192 1 0 ------ 0 0 -----
2-3 3 2 73 0 1
3-4 5 3 186 0 0 -----
4-5 1 17 361 7 3 395
5-6 0 6 204 0 2 152
6-7 0 0 52 0 0 47
7-8 0 0 12 0 0 4
8-8.5 0 0 2 0 0 0
Totals 10 29 890 7 7 598
*Excavated levels combined into even feet to simplify tabulation.
**All waste flakes and fragments from stonemaking reciverable on 1/4-inch screens,
debitage counts omitted for unscreened levels.
sequence does not occur at the upstream location of MR92.
Ceramic and Stone Tool Assemblages
The 1973 excavations at Silver Springs provided some new evidence
and corroborated Neill's report in some respects but not in others. Sampling
error apparently affects our comparison in that relatively more sherds and
stone tools and fewer projectile points were recovered (Table 1). Moreover,
the density of remains was indeed low in the areas selected for excavation on
the basis of this earlier report.
The ceramic zone at MR92 consists of approximately the upper 4
feet in Test 1 and 5 feet in Test 2, with only a few sherds recovered below
these depths. The superposition of ceramic types agrees well with known
chronology in Florida. Orange Plain and Orange Incised sherds were
present 3 to 5 feet below the surface, sand tempered plainware (probably
a Deptford type) from 2.5 to about 4.5 feet, and Wakulla Check Stamped
and St. Johns Plain sherds in the upper 2.5 feet. In general the small
ceramic collection tends to corroborate the undisturbed nature of the strat-
igraphic sequence. The stone tools recovered in the ceramic zone were
relatively few and not distinctive (Table 1; Fig. 6a, b).
A preceramic Archaic zone was well represented below the ceramic
zone in Test 1 and was probably mixed with the earliest ceramic levels
in Test 2. In general this zone extended from 4 to 6 feet below the modern
surface while a clearly defined occupation "floor" was present at about 4.3
feet in Test 1. Tools and flakes were lying flat along an otherwise unmarked
buried land surface at this level. Although these remains were undoubtedly
in "primary archaeological context, the horizontal distribution (when plotted)
did not elicit a meaningful pattern other than loose clusters of tools and
flakes. Presumably, if a large excavation were opened up horizontally
at this level the pattern of an Archaic living floor could be distinguished
(Hemmings 1969). A large proportion of the stone tools and debitage from
the excavation was recovered in the preceramic Archaic zone, especially
from the floor at 4.3.feet. These tools are, for the most part, common
Late Archaic types and need not be described in detail here (see Hemmings
and Kohler 1974). A selection of tools is shown in Figure 6, a-n, and in
Figure 7. The latter specimens (a heat-treated microcore, two retouched
mocroblades, and a bifacial mocrotool) are of special interest because
they are known for certain areas of Florida, but have not previously been
excavated in stratigraphic context. At MR92 two of the microtools were
found from 4 to 5 feet and the other two from 5 to 6 feet below the surface
in Test 1. All are believed to date to Late Archaic occupation (ca. 5000-
2000 B.C.) at Silver Springs (Bullen 1973).
Not well represented in these tests were the Paleo-Indian levels
reported by Neill nor any recognizable Early Archaic levels. Only a
minor proportion of debitage and no stone tools were recovered below 6
feet. However, the flakes in these lowest levels were flat-lying and
apparently in situ, as in overlying levels. It seems probable that the
earliest occupation levels are very sparsely represented and that only the
periphery of a greater concentration was penetrated by deep levels of
Tests 1 and 2. There is no particular reason why Paleo-Indian or Early
Archaic campsites could not occur at many locations along the Silver
Springs valley, but these would be deeply buried and difficult to locate except
by extensive testing. At MR92 the deepest cultural occurrence was two
flakes at nearly 8.5 feet in Test 1. Despite conscientious deeper testing
no older remains were located here, reasonable negative evidence for
anything pre-dating Paleo-Indian occupation in this area.
A single fragment of a fluted projectile point was recovered at 5
feet in Test 1 (Fig. 61). This small midsection fragment is recognizable
from the flute scar on one face and edge-grinding near the base. If Neill' s
Paleo-Indian level at 7 to 8 feet is accepted, then this fragment was out
of context in the stratigraphic sequence, presumably carried upward by
natural disturbance of dropped by later inhabitants of the site.
Debitage or the waste flakes and fragments resulting from flintknapping
comprise the largest artifact category in all levels of the site. This
collection was analyzed for several kinds of information: (1) the density
of debitage by level as an indication of intensity of occupation, (2) the
proportions of kinds of debitage by zone as an indication of the nature
of flintknapping activity, and (3) measurements of patination rinds by level
to examine the age/weathering relationship in local flint. The results of
these analyses are summarized in Tables 2 to 4, except the patination
Fig. 6. Stone tools from MR92. a, f-g, j, o, projectile point tips; b,
notched or tanged knife; c, irregular flake scraper; d, p, side scrapers;
e, notched adze;h, end scraper; i, large biface fragment; k, drill base
1, fluted point midsection; m, large unifacialtool; n, bifacial preform;
q, unifacial beaked tool; r, projectile point midsection. a-b, from ceramic
zone; c-r from preceramic Archaic zone. Length of a is 65 mm.
Fig. 7: a, microcore; (b-c), retouched microblades, and (d) bifacial microtool.
Table 2. Densities and Ratios for Stone Tools and Debitage by Levels, Tests 1-2.
1 Avg. debitage
:f 100 cu. ft.
per 100 cf
100 cu. ft.
0-1 ** .* .....
1-2 ** 4* ^ ** ** 4^
2-3 0.7 26 1:32 ** -- ,-
3-4 1.0 61 1:62 ** .. *.
4-5 5.2 109 1:21 2.6 344 1:132
5-6 1.7 57 1:34 2.1 160 1:76
6-7 0.0 14 ---- 0.0 63 ----
7-8 0.0 3 ---- 0.0 7 ----
8-8.5 0.0 1 ---- 0.0 0 ----
*Excavated levels combined into even feet to simplify yabulation
**Densities and ratios omitted for unscreened levels.
study which was unsuccessful (most flakes from all levels were
patinated through their thickness and had no measureable rind).
Comparatively low densities of debitage (Table 2) indicate that the
Silver Springs Site was at all times primarily a campsite with minor
flintknapping activity. Presumable a large part of stone toolmaking was
accomplished at a workshop location on or near a quarry, and local knapping
per 100 c
Table 3. Analysis of Debitage from Test 1.
Percent Totals (N=892)
Cultural Zones Flakes of bi- Hard hammer Blocky Percent heat Total
(feet) facial retouch flakes fragments altered debitage
Ceramic zone 83.2 4.2 12.6 28.7 261
Preceramic 86.2 2.9 10.9 21.0 617
Paleo-Indian 100.0 0.0 0.0 0.0 14
Table 4. Analysis of Debitage from Test 2.
Ceramic zone 90.3 1.0 8.7 14.9 423
Preceramic 84.0 4.0 12.0 20.6 199
Paleo-Indian 100.0 0.0 0.0 0.0 4
was concerned with finishing and refurbishing tools. The nature of debitage,
largely flakes of bifacial retouch (Tables 3 and 4), and the tool assemblages
support this inference. The highest tool and debitage densities occur in
the 4 to 5 fool level in both Tests 1 and 2. In Test 1 the Late Archaic
occupation floor accounts for much of this density. In Test 2, however,
greater density of debitage (and relatively greater numbers of flakes of
bifacial retouch) occurred at the base of the ceramic zone (Table 4).
This 4 to 5 foot level in Test 2 has, therefore, the character of a camp-
site-workshop area with a small tool-debitage ratio (1:132) where bifaces,
possibly projectile points, were manufactured.
Heat alteration of flakes, presumably intentional heating to enhance
flaking quality of the raw material, was recorded by zone for both tests,
and appears to be relatively uniform in occurrence (15-30 Percent). How-
ever, the heat alteration characters (discoloration, lustrous surfaces,
crazing, etc.) are practically obscured in heavily patinated flakes. The
debitage from supposed Paleo-Indian levels is indeed heavily patinated,
and comprised too small a sample for any other comparison.
The Silver Springs site was located on favorable high terrain border-
ing the run. This site was occupied and re-occupied over about 10,000
years, and has aggraded by some 8 feet of windblown sand (Unit Al)
during this interval. The stratigraphic sequence is largely, but not en-
tirely, undisturbed. Occupation floors preserved in the sand deposit are
horizontal accumulations of stone tools and other inorganic debris. One
of the best defined floors (Late Archaic) was present at 4.3 feet below
the surface in Test 1. A microcore and microblade tools were recovered
at several levels dating to Late Archaic occupation. Paleo-Indian (?)
debitage is barely represented in 7 to 8.5 foot levels of both tests, and
one fluted point fragment was recovered out of contest in an Archaic level.
Throughout its span of occupation the Silver Springs site was a campsite
whose inhabitants were exploiting the abundant plant and animal resources
of the area. Stone tools and ceramics employed in extracting and processing
of these resources or in manufacture of other tools are the only direct
evidence remaining for this exploitation. Flint debitage from all levels
of the site indicates that knapping was not extensive and was directed
toward finishing or refurbishing tools.
Other Sites in the Silver Springs Valley
No intensive archaeological survey of the Silver Springs valley has
ever been undertaken, although sporadic archaeological excavation in this
area has continued for 80 years. The following summary of a few known
sites (in addition to MR92) is based on published information, archaeol-
ogical site files and collections in the Florida State Museum, and numerous
field trips in the area during 1973-74.
Silver Springs Midden (MR53). Although there is some confusion
about the identification and location of this shell midden, it appears to be
a disturbed site located 3 miles downstream from the headspring on high
terrain bordering the south bank. The midden is restricted to about one-
quarter acre and a foot in thickness, consisting largely of snails (Viviparus
and Pomacea). Ceramics from the midden include the linear check stamped
version of St. Johns ware and sand tempered plainware. The occupation
of this site was approximately at the time of Silver Springs Mound. Goggin
(1952:96) lists the site, but gives no further information.
The location of MR53 is geomorphically identical to the Silver Springs
site, and there is good reason to believe that early lithic levels in deep
eolian sand could be located here as well. Scattered flakes are present
on the surface over an area of 10 acres or more overlooking the run.
The Cavern and Headspring (MR59). The underwater occurrence of
sherds, flakes, stone tools, other artifacts, and extinct animal remains at
the head of Silver Springs has been described by Neill (1952, 1964). Florida
State Museum collections from the headspring contain a number of sherds
from heavy net or fabric impressed globular jars, representing a ceramic
type presently undescribed. Regarding earlier remains, Neill has reported
the apparent association of a fluted Suwannee point, a few other worked
flint pieces, a human skull fragment, and mammoth and mastodon bones
from the mouth to 45 feet within the cavern at the headspring (Neill 1964:26).
This cavern, the main vent of Silver Springs (Fig. 1), has an essentially
horizontal floor 35 feet below present spring level. Because no systematic
excavation has been carried out here, and insufficient data are available,
the association reported should be regarded as unproven. Clearly, however,
the cavern would have been habitable at times of lower spring level, and
the site has great potential for further study.
Unnamed Sites (MR83 and MR93). Two sites known only from surface
reconnaissance are located on the south bank of Silver Springs Run near
"Site MR93, just upstream (Fig. 1), was exposed extensively by
construction work in the area during 1973-74, but has now been largely
backfilled and landscaped. A surface collection from this site (deposited
in the Florida State Museum) contains a number of interesting specimens,
including portions of a St. Johns Linear Check stamped vessel with tetrapods
(Fig. 8), somewhat like Moore's find at Silver Springs Mound, and several
large exceptionally will-made chert blades (Fig. 9). This material was
recovered in areas deeply disturbed by bulldozing. The mantle of windblown
sand (Unit Al) is present throughout the area, and debitage was observed
at 4 feet below the modern surface.
MR83 is a sherd and flint area just downstream from the Silver Springs
Site (not shown in Fig. 1), partly disturbed by a dredged run. This is
probably the location where Neill (1964:26) retrieved a "Suwannee knife,"
and which has been briefly described by Bullen (1958:28). A ceramic zone
over-lies a lithic zone in the familiar deep windblown sand. Here again
there appears to be great potential for early lithic occupation levels.
The Guest Mammoth Site (MR130). An underwater excavation of
mammoth bones and stone artifacts at a location about two miles downstream
on Silver Springs Run (not shown in Fig. 1) was begun in 1973 and continues
at the present time (Brooks 1973; Hoffman and Brooks 1974). The major
find to date has been a small fluted Suwannee like projectile point associated
with the somewhat disarticulated carcass of a juvenile Columbian mammoth.
These remains occur on a marked erosional unconformity in the marl
sequence (Unit A4), and have been dated just under 10,000 years B.P.
The carcass and projectile point occurred about 10 to 12 feet below modern
spring level. As additional work is in progress here, no further description
will be given. Clearly, this site is of major importance to the study of
Early Man in Florida and the eastern United States.
Fig. 9. (above) Blade tools
Fig. 8. (to left) Check stamped
vessel from MR93: a, rim ex-
terior; b, junction of base and
wall; c, tetrapod. Height of
c is 40 mm.
Miscellaneous Underwater Finds. A number of artifacts have been
found along the bed of Silver Springs Run including a fluted Suwannee
Point near MR83, a Bolen Beveled Point near MR130, and bone points and
point fragments in several areas. Remains of one other mammoth were
recovered in marls just upstream from the Guest Mammoth site several
years ago, apparently not in association with artifacts. Pleistocene horse
and tapir bones, including concentrations of parts from single individuals,
have also been reported in marl outcrops downstream from MR130.
Finally, in addition to the mammoth and mastodon remains reported by
Neill (1964), Florida State Museum collections include a number of sloth
bones from the headspring.
The valley of Silver Springs Run contains exceptional resources for
the study of prehistory. Few other areas of comparable size in the Florida
peninsula combine such a long record of human occupation, uniquely favor-
able preservation, and definable geochronological context. By "geochrono-
logical context, I mean that the various aspects of stratigraphy, dating,
and paleoecology are all susceptible to study. The promising work in these
areas has merely begun. It is also true that the inventory of known
archaeological sites in the Silver Springs Valley, both above and below
spring level, is markedly incomplete. It is desirable to complete this
inventory in the near future so that comparisons can be made on a local
and regional basis, and so that very significant sites can be identified
Silver Springs is especially important for the study of Early Man
since the potential, at least, of this area for certain basic requirements
has now been demonstrated: (1) stratified sites in which projectile point
styles and associated tools can be isolated and related to other lithic
complexes, (2) firm evidence of man's contemporaneity with and exploitation
of extinct animals, and (3) reliable dates for early lithic occupation sites
(Hemmings 1972). These are striking initial results from several invest-
igations in the earliest habitation and "kill" sites in Silver Springs Valley.
The archaeological record for Late Archaic, Orange, and Deptford-
St. Johns I peoples is also well represented, and certainly not unique to
Silver Springs. However, the occurrence of buried living floors in "primary
archaeological context" can provide a new dimension in the interpretation
of site use and social groups. Presumably, these peoples were hunter-
gatherers adapted to wild plants and animals of closely-spaced microenvir-
onments, nearly the same as today, in Silver Springs Valley. When does
plant domestication appear in the record for central Florida? Perhaps
the pollen-bearing sediments of Silver Springs Run can contribute to this
problem as well.
Excavations at the Silver Springs site were made possible through
the cooperation and assistance of Florida's Silver Springs, a division of
ABC Scenic Attractions. I am most grateful to Mr. W.C. Buck Ray,
Jr., and Mr. Donald F. Keck, officials of that organization, for their
continued interest in Silver Springs archaeology. The field crew at MR92
consisted of Gerry Evans and Malinda Stafford, Florida State Museum,
Tim Kohler and Nina Thanz, University of Florida, Carol Spindel, Bennington
College, and Jim Michie, University of South Carolina. All of these
individuals gave unstinting effort to remove and screen several thousand
years' accumulation of sand (in much less time).
I am also grateful to my colleagues, Dr. Charles A. Hoffman, Jr.,
Northern Arizona University, and Dr. H. Kelly Brooks, University of
Florida, who allowed me to dive at the Guest Mammoth Site as work
proceeded, and who shared their findings from this interesting project.
Finally, Mr. Ben Waller, Ocala, Florida, and Dr. David S. Webb, Florida
State Museum, contributed firsthand information about underwater site in
Florida and, in particular, Silver Springs.
1973 Late Pleistocene and Holocene Climatic Change in
Peninsular Florida. Paper presented at the Annual meet-
ing of the Geological Society of America, November 12-14,
Bullen, Ripley P.
1958 The Bolen Bluff Site on Paynes Prairie, Florida.
Contributions of the Florida State Museum, Social Sciences,
No. 4. Gainesville.
1973 Introduction. In Archeology of the Florida Gulf Coast, by
Gordon R. Willey, pp. vii-xi. AMS Press, New York.
Ferguson, G.E., C.W. Lingham, S.K. Love, and R.O. Vernon
1947 Springs of Florida. Florida Geological Survey, Bulletin
No. 31. Tallahassee.
Goggin, John M.
1952 Space and Time Perspective in Northern St. Johns Archeology,
Florida. Yale University Publications in Anthropology,
No. 47. New Haven.
Excavation and Analysis of Living Floors in Alluvial Sites.
Proceedings of the 26th Annual Southeastern Archeological
Conference, Bulletin No. 11, pp. 41-45. Morgantown
Early Man in the South Atlantic States. South Carolina
Antiquities, Vol. 4, No. 2, pp. 1-13. Columbia.
Recent Excavations at Silver Springs, Marion County, Florida.
Paper presented at the Annual Meeting of the Florida Anthro-
pological Society, March 17-18, St. Augustine.
E. Thomas, and Timothy A. Kohler
The Lake Kanapaha Site in North Central Florida. Bureau
of Historic Sites and Properties, Division of Archives, History
and Records Management, Bulletin (in press). Tallahassee.
Hoffman, Charles A., and H.K. Brooks
1974 The Guest Mammoth Site in North Florida. Paper presented
at the Annual Meeting of the Society for American Arch-
aeology, May 3-5, Washington, D.C.
Layne, James N.
1970 Terrestrial Wildlife of the Oklawaha Regional Ecosystem.
In Environmental Impact of the Cross-Florida Barge Canal
with Special Emphasis on the Oklawaha Regional Ecosystem,
pp. 35-42. Florida Defenders of the Environment, Gainesville.
Lugo, Ariel, and Archie Carr
1970 Vegetation of the Oklawaha Regional Ecosystem. In Environ-
m mental Impact of the Cross-Florida Barge Canal with Special
Emphasis on the Oklawaha Regional Ecosystem, pp. 29-33.
Florida Defenders of the Environment, Gainesville.
Mason, Ronald J.
1962 The Paleo-Indian Tradition in Eastern North America.
Current Anthropology, Vol. 3, pp. 227-246. Chicago.
1970 Effect of the Cross-Florida Barge Canal on Hydrology of
the Area. In Environmental Impact of the Cross-Florida
Barge Canal with Special Emphasis on the Oklawaha
Regional Ecosystem, pp. 65-72. Florida Defenders of the
Moore, Clarence B.
1896 Certain Sand Mounds of the Ocklawaha River, Florida.
Journal of the Philadelphia Academy of Natural Sciences,
Vol. 10, Part 4, pp. 517-543. Philadelphia.
Neill, Wilfred T.
1952 Unusual Rattles from Silver Springs, Florida. Florida
Anthropologist, Vol. 5, Nos. 3-4, pp. 33-35. Tallahassee.
1958 A stratified Early Site at Silver Springs, Florida. Florida
Anthropologist, Vol. 11, No. 2, pp. 33-52. Tallahassee.
1964 The Association of Suwannee Points and Extinct Animals
in Florida. Florida Anthropologist, Vol. 17, No. 1, pp. 17-32.
Reid, George K.
1970 The Native Aquatic Fauna of the Oklawaha River. In
Environmental Impact of the Cross-Florida Barge Canal
with Special Emphasis on the Oklawaha Regional Ecosystem,
pp. 43-46. Florida Defenders of the Environment, Gainesville.
Gaine sville, Florida
July 1, 1974
FORT WALTON TEMPLE MOUND (80K6M):
FURTHER TEST EXCAVATIONS, DePAUW 1973
Yulee W. Lazarus and Robert J. Fornaro
The Temple Mound at Fort Walton Beach in the Florida Panhandle
has been observed, excavated, and tested through the past ninety years.
The latest testing produced some ceramics not easily classified except as
variants of recognized types and an extended burial showing evidence of
a disease not previously recognized in the area.
The mound was first recorded in the literature by S. T. Walker. His
report to the Smithsonian Institution in 1883 stated the mound was so over-
grown with bushes as to be not easily measured. Major excavations by
C. B. Moore in 1901 proved the importance of the mound by producing ample
ceramic and burial evidence in a cultural structure with recognition of the
culture as one of a distinct social group. Although Willey (1949) excavated
in the village midden area between the mound and Santa Rosa Sound, and not
in the mound itself, he classified the ceramic vessels recovered by Moore
as of the Fort Walton period, with the mound as the type site of that culture.
Throughout these years the mound had been privately owned until 1959 when
it became the property of the City of Fort Walton Beach.
Excavations in 1960 (Fairbanks 1965) established the manner of mound
construction. Evidence of post holes was also found supporting the use of
the structure as a temple mound. In 1963 work in a limited area document-
ed an important Deptfort period midden below the mound base and (Wing 1968)
produced a diet list for that group. The site was placed on the National
Register as a National Historic Landmark by the United States Department
of Interior in 1964. Moore's (1901:436) map shows the areas he excavated and
also a circular shell deposit he reports as 55 feet in diameter and about 2
feet thick. If this shell deposit is assumed to be the foundation for the last
temple structure, any evidence of post holes should define structure size,
shape and location. Willey makes no reference to this shell; neither was it
present in 1960. Figure 1 shows the excavated areas reported above. Fig-
ure 2 shows later test areas with dates of work. Post holes defined in the
Fairbanks work are also included in Figure 2.
Subsequent years (1971, 1972) have seen additional minor excavations,
particularly to locate additional post holes which might relate to the temple
structure or to recover additional ceramic material which might fit earlier
finds and possibly produce complete vessels. This later work has been un-
dertaken by Temple Mound Museum personnel on a volunteer basis. Test
excavations by DePauw University in 1973 provided additional information
about the type site (Fig. 2). Mound fill to restore to original shape had
Florida Anthropologist, vol. 28, no. 4, December 1975
FORT WALTON MOUND
Fig. 1. Plan of Fort Walton Temple Mound locating pre-1965 excavations.
been brought by the City in 1966 from a marshy area a mile away which
was being cleared for construction of the City's Safety Building. Because
the added fill was of dark humic materials it was clearly discernible in
the upper stratum of all tests.
LAZARUS AND FORNARO
Fig. 2. Plan of Fort Walton Temple Mound locating 1966-75 excavations.
Field Tests, August 1971, Museum Personnel
It was hoped the 1971 tests would show more evidence of post holes
of late aboriginal structures on the mound. Areas selected for testing in
FORT WALTON MOUND
1971 were those not tested in 1960. Results were disappointing. One area
had been occupied by a modern cottage removed in 1960 by City crews. An-
other was a section probably previously excavated by Moore. A third area
did produce a possible post hole as a circular deposit of shell 8 inches in
diameter and thirteen inches vertically. This shell may possibly have been
tamped around a post for stability in the soft sand. Test depths to 30 and
40 inches were judged sufficient for locating the post holes of any recent
aboriginal structure. Among additional material recovered were 873 sherds
of all local cultures. Description of these sherds is omitted in this report
as they had no distinct or unusual distribution specially or stratigraphically.
Field Tests in 1972, Museum Personnel
The datum point used in 1960 was removed in highway construction in
1966. In 1972 a permanent datum point was established on the southwest
corner of the new museum building 85 feet north of the Highway 98 right-of-
way line. A temporary datum, referred to the permanent one, was then
placed on the southeast corner of the top of the mound 62 feet directly west
of the permanent one. Elevation of the temporary point is 25 feet above
sea and 10 feet above the mound base at the southeast corner.
In September 1972 a 5- by 5-foot test square was excavated in the
extreme southeast corner of the platform top 4 feet south of the temporary
datum. This test was undertaken because of surface finds of numerous fit-
ting Fort Walton sherds. The first 4 inches were of recent fill dirt but
the underlying soil was dark gray and relatively uniform to the final 30 inch
depth of this test. Sherds from 9 vessels were recovered from the square
at varying depths, some fitting the surface sherds which had led to the test
in the first place. Four vessels were restorable -- two of Pensacola Incis-
ed, one Lake Jackson Plain, and one of Fort Walton Incised (Fig. 3). Four
vessel segments were one of Englewood design but shell tempered, one Fort
Walton Incised, and two Pensacola Incised. The ninth vessel was an in-
verted complete Fort Walton Incised bowl which had been broken in situ.
It is a shallow casuela of very friable, red, sand tempered paste bearing
a shallow, incised design covering the entire bowl (Fig. 4). There was a
single quartz flake under but not within the bowl. There were 508 sherds
recovered. One sherd represented a six pointed plate shape and is includ-
ed in the discussion and tables regarding ceramics and plate shapes later
in this report. The presence of the complete bowl as well as the unifor-
mity of soil color suggests this extreme corner on the top of the mound
had not been excavated before this test.
Field Tests, January 1973, DePauw University
In January 1973 three test areas were plotted for DePauw University
students to excavate. One area, a 10-foot square designated Pit 1, was
LAZARUS AND FORNARO
Fig. 3. a, c, Pensacola Incised; b, Lake Jackson Plain; d, Fort Walton Incised.
Fig. 4. Fort Walton
Incised casuela (broad
FORT WALTON MOUND
situated adjacent to the 5-foot square excavated in 1972. Its west boundary
was on the temporary datum and the south boundary overlapped the 1972
square by one foot (Fig. 2). The other two areas, Trench 1 and Trench 2,
were test trenches on the east side of the mound top and were oriented east
to west. They cut into the platform top of the mound and extended partially
down the slope. The width and location of these two trenches was governed
somewhat by the large trees and the avoidance of their destruction. Test
Trench 1 was 1 foot north of Pit 1, 3 feet wide and originally 15 feet long.
Lying parallel to Trench 1, Test Trench 2 was 62 feet north of Trench 1
and was 7 feet wide and originally 10 feet long. Both trenches were extend-
ed to a final length of 26 feet 8 inches and 31 feet, respectively. Trench
1 was extended both east and west and Trench 2 was extended west (Fig.
2). In addition to the whole and partial bowls described in the following
sections, the DePauw tests produced 7925 ceramic sherds and miscellaneous
stone, bone, and shell artifacts (Tables I, II, III).
Table 1. Vertical distribution of sherds and historic items, DePauw 1973, Pit 1
Levels Fort Weeden Santa Rosa- Dept- Residual Unclas- Total Historic
(inches) Walton Island Swift Creek ford plain sified sherds items
0-5 14 -- -- 18 --- 32
5-12 133 6 9 7 181 5 341 2
12-15 203 47 34 1 334 2 621 1
15-20 173 5 10 7 56 -- 251 4
20-26 385 2 5 4 309 1 706
26-36 508 11 16 19 307 2 863 1
36-45 113 1 5 12 146 1 278
45-56 102 18 34 51 211 1 417
56-69 174 60 63 77 239 -- 613 --
Totals 1805 150 176 178 1801 12 4122 9
DePauw University Test Pit I.
In the south half of Test Pit 1 at a depth of 26 inches was the City's
3-inch gas pipe running in an east-west direction. Nine pieces of late his-
toric material were from above the 28 inch level; none was from below this
level. The soil below appeared undisturbed by 19th or 20th century activi-
ties and contained clam shells, a small quantity of animal bone fragments,
and a few sherds. At the general level of 41 inches the floor of the pit
was of uniform black humic material with fish and animal bone fragments,
clam shells, and some oyster shell. Since the upper surface of the test area
was sloping down the mound corner and a horizontal floor was maintained,
the average depth measurement is used. No stratification of soil was ap-
parent in the walls. From 50 to 74 inches the humus stratum was so solidly
uniform in character, black humus with fine roots and vegetal matter, that
a flotation sample was taken. One inch (one-half bushel) of the floor surface
was scraped out and washed through window screen (Sample No. 1, 80K6M-
LAZARUS AND FORNARO 165
Table 2. Vertical distribution of sherds and historic items,DePauw 1973, Trench 1
Santa Rosa- Dept-
Swift Creek ford
Table 3. Vertical distribution of sherds and historic items, DePauw 1973, Trench 2
Levels Fort Weeden Santa Rosa- Dept-
(inches) Walton Island Swift Creek ford
All flotation and soil samples were sent to Florida State
identification. The final depth was 90 inches below present
with no change in soil characteristics.
FORT WALTON MOUND
DePauw University Test Trench 1.
This 3-foot wide, 15 feet long, trench was located 1 foot north of Pit 1
and extended 5 feet west of the temporary datum. It was ultimately extended
5 feet more to the west and 6 feet 8 inches to the east. Trench length on
the platform top was 18 feet. In addition, an oblong extension was cut north
and the 1-foot control wall was removed on the south in the same area (Fig.
2). These irregular extensions became necessary for the removal of a bur-
ial. Mottled topsoil was evidently the recent City fill as contour of stratum
was clear (Fig. 5). From just below recent fill to the final depth of 80 in-
ches artifactual material was present. Random bits and pieces of bone were
recovered from 24 to 51 inches. In the eastern half several sections of Fort
Walton bowls were found broken in situ. At 35 inches a small globular casuela
bowl of Fort Walton Incised (Fig. 6) was found upright but broken in situ.
Some human bone fragments were at the 43 inch level. At 47 inches and 7 feet
from the east extension stake (on the crest of the slope) human long bones
were found in diagonal placement across the floor. It was obvious the rest of
the body extended northwest and southeast so the necessary extensions were
made. The skeleton, esposed before removal, was articulated, lying on its
back and oriented northwest to southeast, arms at the sides, cranium to north-
west. The material was taken to Florida State University for examination.
There was no funerary were in direct association. No distinct intrusive pit was
defined for the burial but mottled soil extended to 55 inches in the eastern half
of the trench. The following report was provided by R. C. Dailey of Florida
Using suture closure and appearance of the skull, it is suggested
that this individual was a female, 40-45 years of age. The teeth
show moderate to marked attrition with mild alveoloar resorption.
Examination reveals a roughening over the outer table of the entire
cranial vault, some of which may be post mortem. The cranium is
misshapen, compressed anteriorially and posterially, undoubtedly
due to earth pressure. There may also be some cradle-board defor-
There are several circular holes in the skull of various sizes
with minimal bone condensation at the borders. Most these are
located at the base of the skull. There are several additional
holes through the outer table only, located in the parietal and
occipital bones. These appear to have originated in the outer
table and not from diploic space. Examination of the lower jaw
shows no similar lesions; the palate and the maxillary teeth
appear to be normal; nor are there any pathological changes in
the vertebral fragments. X-rays of the cranium show many roughly
circular areas of decreased density ranging in size from one to
LAZARUS AND FORNARO
Fig. 6. Fort Walton Incised globular
Fig. 5. Trench 1 at 12 inch level. casuela from depth of 35 inches, Trenchl.
twenty mm. in diameter located in the basal portion of the skull.
The reaction around these holes is minimal to zero. The top of the
skull and the lower jaw appear to be normal with no evidence of
Likely possibilities are that this is an instance of multiple myeloma
or osteolytic carcinomatosis. However, the periosteal reaction on
the outer table on the cranium does not belong to either one of the
Because pathologies of this kind should occur throughout the skel-
eton the remainder was requested in order to perform a gross
assessment. The additional parts proved to be in a poor state of
preservation with most of the bones in fragments. None are meas-
ureable. Included in the fragments was the petrous portion of the
right temporal bone. As the diseased cranium had both left and
right portions present the existence of an additional right portion
suggests that there was a second skeleton in this burial. And since
the infracranial bones are normal it is suggested that these should
be associated with the petrous portion of the right temporal and not
with the diseased cranium.
The problem of the possibility of a second partial burial is not resolv-
ed. The extraneous bone referred to in the report might have been present
in fill dirt. Moore's report had indicated that other remains were found in
very close association (Moore 1901:438). The presence of multiple myeloma
among peoples in the Panhandle extends the area as reported found in the
FORT WALTON MOUND
Tallahassee area (Morse 1973). Some of the sherds from Pit 1, Trench 1,
and the 1972 square, all on the southeast corner of the mound, were found
to fit and form several partial sections of Fort Walton time period vessels.
The extreme southeast corner had apparently not been disturbed by any 19th
or 20th century activities below 30 inches.
Historic material, Trench 1.
Of the full length of 26 feet 8 inches of this trench, nearly 18 feet
were on the platform top. Moore's map (Fig. 1) shows his excavations to
have been over the full southeastern corner of the platform top (Fig. 1).
The western extension was made to locate the 1901 excavations. While no
clear line of demarcation was evident on the walls of the trench, the mater-
ial indicated the area had surely been used for deposit of trash. This sec-
tion should have extended into the 1901 excavations. The west half of Trench
1 was more heavily concentrated with trash as deep as 50 inches from pres-
ent surface. An open air general store was operated for some years at
the base of the mound just prior to 1959 when it was removed. Trash from
this operation could have been deposited in this area. There was sporadic
digging in the mound from 1860 to 1959. The account of the Walton Guards'
digging in 1861 (McKinnon 1968:68-70) reports a quantity of bones recovered
from "mixed up" burials, but there is no mention of historic trash. In
the course of 70 years with settler activities and weathering, changes could
occur in the final configuration of the mound shape. Material recovered
from Trench 1 was distinctly different in qualitative count from one end to
the other. Late historic items greatly exceeded 19th and 20th century "trash"
from all previous areas.
The time period of use of this area would ordinarily be difficult to
identify by the 371 pieces. Of this total, 332 fragments were of common
glass. It was possible to conclude that at least 53 glass containers or
shapes were represented, including drinking glasses, an electric light bulb,
and a Mason jar (late 1800's, Austen 1971:89) as well as a number of wine,
milk, and soda bottles.
One bottle (Figs. 7 and 8) in almost intact condition was clearly a
bottle from the Coca Cola Bottling Company in Mobile. Two contradictory
references are used for time period of this bottle. A study by Craig Gil-
born was published in MUSEUM NEWS in December 1968 (p. 12). Criteria
used in this study places the bottle in use between 1894 and 1915. The Coca
Cola Bottling Company in Mobile states this type bottle was used to package
soda water flavors from the mid-twenties to 1959 (M.C. Geron, personal
communication). Using the first criteria we could assume the bottle fit the
time of Moore's work. Coca Cola was bottled in straight-sided bottles from
1894 to 1915 and in verying designs. Our bottle, with the lacy design, is
LAZARUS AND FORNARO
Figs. 7-8. Coca Cola bottle.
FORT WALTON MOUND
clear glass showing a double mold seam on opposite sides along the raised
rib. The base ring is worn and stamped to show contents of 6 /2 fluid
ounces. It is 7 3/4 inches high. The "type" then fits rather well for the
early time period. However, the Bellingrath Beverage name could apply sol-
ely to the soda water flavors distributed in bottles made by the Coca Cola
Bottling Company in nearby Mobile at any time period, as reported by Geron.
Presence of glass fragments terminated at the depth of 55 inches.
Data recovered from the three test areas on the southeast corner of
the temple mound (the 1972 square, Pit 1 and Trench 1 of 1973) suggest the
tests slightly overlapped on the west some prior disturbances of the mound
structure. The in situ material east of the disturbed area was in place be-
low the top of the mound and within the eastern slope.
DePauw University Test Trench 2.
This trench and its extensions yielded only about one-third the cult-
ural material as was found in the other excavated areas. However, there
were some compensating features. Poised on the steep slope as it was, it
soon became apparent that an extension in length was necessary so digging
could be done in "steps". The 7-foot north-south width was confined by
large trees, bushes, and wooden steps down the side of the mound to the
museum building. Digging began midpoint of the original length, cutting 5
feet to the west, forming a step, and 5 feet to the east, forming a second
step. Trench 2 was then extended west for 7 feet. Topsoil was black
humus devoid of sherds and was the City's mound fill used to rebuild the
mound to the near 45 degree angle slope of its original shape. The humus
was progressively more shallow in depth toward the west in the 7-foot ex-
tension (Fig. 9). Old mound fill was brown and grainy and its surface or
top was distinct under the City's humus. Soil Sample No. 3 was taken at
the 20-inch level in the extension, 8 inches below late humus fill. At 27
inches a flotation sample was taken of the same soil (Sample No. 2, 80K6M).
The trench was again extended west for 14 feet. The mottled appear-
ance of soil was evident but without sharp line of change. Three narrow
shell column rings, one of 71/2-inch diameter and the other two of 5-inch
diameter, were exposed in an east-west line. They were from 6 to 23 in-
ches in depth and separated by 47 and 29 inches respectively (Fig. 10, SH).
Soil sample No. 5 was scooped out of the deepest column center. These
might be considered possible post holes of some use.
The total length at this time was 31 feet. At this point it was hoped
that a greater depth could be attained by limiting the digging to a 7-foot
square in the western end. The walls were shored up with scrap lumber.
A permanent measuring string was staked along the surface and then 2-foot
LAZARUS AND FORNARO
Fig. 9. Trench 2,
looking up slope at
L I I I I
FIGURE /. TRENCH 2 SOUTH WALL
depths of mound surface were removed to 3 feet back from trench walls for
safety. As greater depth was attained in the 7-foot square, the soil became
lighter in color turning to dark gray, tan, and then to light gray. There
was a slump appearance as though an addition to the old mound had sloped
east to the lower surface. At 86 inches there was some ceramic material,
but only 1 Fort Walton sherd out of 68 came from the 71- to 92-inch level.
The remaining 67 sherds were Weeden Island, Santa Rosa-Swift Creek, and
Deptford in almost equal numbers. One particular sherd was considered
very unique. It is only 3 mm. in thickness and high-fired to a hardness of
4. It is black with a curved shape suggesting a bottle. Uniquely, its rocker
decoration was excised after firing (Fig. 11). At 92 inches a dark stain area
in the floor of the trench was removed (Sample No. 4, 80K6M-DePauw 1973)
and shortly thereafter the shoring gave way collapsing the walls.
There have been thousands of ceramic sherds recovered from the
Temple Mound over the 15 years of work in this site. In general, these
FORT WALTON MOUND
Fig. 12a. Shell tempered, buff slip,
charcoal encrusted. Lamar influence?
Fig. 13a. Shell and grit
tempered with three line
Fig. 12b. Shell tempered, red slip,
black core. Aucilla-like.
Fig. 13b. Shell tempered
buff, smoke clouded.
Fort Walton-like ?
Fig. 14. Three micaceous paste
sherds. Lamar or Jefferson Ware.
LAZARUS AND FORNARO
Fig. 15b. Sand tempered, dark gray,
smoothed check stamped.
Fig. 15a. Sand tempered, buff,
over stamped check stamped
Fig. 16a. Deptford Linear Check
Stamped and variants.
Fig. 16b. Shell tempered, dark gray paste, unclassified.
174 FORT WALTON MOUND
AC Dc. E F_
Fig. 17. Projectile points: A, 63-71 in. deep in Trench 2; B, 73 in. deep in Pit 1; CrG, al
from Trench 1, 41-47,0-12, 30-36, 41-47, and 0-17 in. deep respectively.
Fig, 18. Fort Walton rim effigies.
sherds have easily fallen into culture classifications covering the time per-
iods in this area. Examples are numerous of so-called diagnostic traits
so that typology sherd boards have been assembled for institutional uses.
Additional Fort Walton designs have occurred here which have not been
found in other Fort Walton areas further from the type site. The same
has been true regarding Swift Creek complicated stamped paddle elements
occurring here but not found in its original area in Georgia. Some of the
variant examples from the 1973 work have been examined for cultural class-
ification and placed in problematic categories. Though some may appear
familiar to areas outside of the Panhandle, these have been selected to show
examples not readily classifiable with ceramics ordinarily found here (Figs.
12-16). Some are not typical Fort Walton. All may demonstrate either an
influence from another cultural group or degenerative craftsmanship on the
part of the potter, or both. There were also a few projectiles (Fig. 17)
and few interesting rim effigies (Fig. 18).
Interest in the six-point plate shape (Fig. 19) grew as study revealed
the concentration of this style to be in a rather limited area around Choc-
LAZARUS AND FORNARO
Fort Walton incised punctated plate.
sand temper, gray paste.
Table 4. Sites with plates having six points
Sites Number 6- Total all Six other sites (Okaloosa, Walton,
pointed plates other bowls and Bay Counties) are recorded as
having one 6-pointed plate each
80K6M 13 64 but the total number of other vessels
80K35 8 39 is not on record. Sizes range from
8WL30 13 59 9- to 18-inch diameters and 3- to
8WL33 4 102 4-inch depths.
8WL50 12 73
FORT WALTON MOUND
tawhatchee Bay. Available published reports, site records in museum files,
and the museum ceramic collection were searched. The first examples,
found in the mound by Moore (1901:436), were considered by him as unique.
Eight plates are represented in the total of 30 Fort Walton vessels represent-
ed in the 1973 work. Five sites have 50 of these plates present. The same
5 sites have a total of 337 vessels of other shapes (Table IV). All but the
temple mound have evidence of European contact. The obvious discrepancy
in ratio of six-point plate shapes to all other known vessels in the one site
(8WL33) is curious when compared to the other 4 sites. The six-pointed
plate may have had a status or special significance in the Fort Walton culture.
Whether or not the temple structure was round or six-sided, it seems
probable it was not square or rectangular if the shell area was the foundat-
ion. One the other hand, if the shell area was used as ceremonial ground
the concentration of post holes clustered near the center of the area might
be considered locations of ceremonial poles of some type.
Ploughing by early settlers in the early 19th century, subsequent devel-
opment of the downtown area, and the intersection of US Highways 98 and
85 around the mound have probably destroyed most of the evidence of late
aboriginal use of the area.
In acknowledgment, we are particularly fortunate to have the labor-
atory analysis of the skeleton by Florida State University. The following
DePauw University students worked from January 4 to 24, 1973: John
Pochette, Indianapolis; Libby Forsyth, Washington, D.C.; Debi Simpson,
Aurora, Colo.; Julie Niehaus, Vincennes, Ind.; Suzy Porter, Logansport,
Ind.; Philip Pochon, Kokomo, Ind.; Steve Vickery, Indianapolis; Gary
Garaffolo, Lebanon, Ind.; Steve Jenkins, Massillon, Ohio. Although some
of these students had had field experience, none had ever worked in a tem-
ple mound. For the interim tests the museum is grateful to those princi-
pally involved: Marshall Cartledge, John Glover, John Morgan, and Jack
Sturm, -and to Pat Balanzategui, Carol Hawkins, and Bill Grigsby for man-
Fort Walton Beach, Florida
October 25, 1974
LAZARUS AND FORNARO
1971 Poor Man's Guide to Bottle Collecting. Doubleday.
Fairbanks, Charles H.
1965 Excavations at the Fort Walton Temple Mound, 1960. Florida
Anthropologist. Vol. XVIII, No. 4, pp. Z39-64.
1968 "Pop Pedagogy: Looking at the Coke Bottle". Museum News,
Vol. 47, No. 4.
McKinnon, John L
1968 History of Walton County. Palmetto Books. Gainesville, Fla.
Moore, Clarence B.
1901 Certain Aboriginal Remains of the Northwest Florida Coast.
Journal of the Academy of Natural Sciences of Philadelphia,
Vol. XI, Part 1, pp. 435-454
Morse, Dan F.
1973 Prehistoric Multiple Myeloma. Paper given at Annual Meeting,
Florida Anthropological Society.
1885 Mounds and Shell Heaps on the West Coast of Florida, Annual
Report of the Smithsonian Institution for 1883, pp. 854-68.
Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast. Smithsonian Miscellaneous
Collections, Vol. 113.
Wing, Elizabeth S.
1968 Animal Remains from a Midden at Fort Walton Beach.
Quarterly Journal Florida Academy of Sciences, pp. 57-58.
THE SENATOR EDWARDS CHIPPED STONE WORKSHOP SITE
(MR-1ZZ), MARION COUNTY, FLORIDA: A PRELIMINARY
REPORT OF INVESTIGATIONS
Barbara A. Purdy
Outcrops of high quality chert throughout north central Florida were
exploited by prehistoric peoples. One such area, the Senator Edwards
Chipped Stone Workshop (Mr-122), Marion County, Florida is located in the
Fairfield Hills region. This is an area of erosional hills raised during the
Ocala uplift of the oligocene and capped by Miocene sands and gravels.
The sands and gravels supplied the silica for the formation of the chert
which is replaced limestone.
Several years ago approximately an acre of the Senator Edwards site3ite
was bulldozed by amateur collectors to a depth of 15 cm. The 5,000 stone
tools recovered were loaned to me for study. In order to obtain a more
representative sample of the chipping debris, utilized flakes, discardedcores,
and crude implements--which were of no interest to the collectors--a back-
hoe trench 48 m long and 1. 5 m wide was dug in an undisturbed area of the
site (Fig. 1). After recording the stratigraphy (Fig. 2), the stone material
was recovered as the trench was refilled (Fig. 3). Every effort was made
to collect all the material but many small flakes passed through the 95-
by 1.9-cm screen. These flakes could have been shatter flakes, pressure
flakes, collapsed remnants of longer flakes, or flakes flipped off during
tool use. More than sixty thousand stone remains were recovered from the
trench. These specimens as well as the collectors' tools have been ana-
lyzed. Each piece of stone was observed under magnification in order to
determine if it had been used and how. Four categories were created from
the initial sorting (Table 1).
1. Tools: The stone had been intentionally altered. If it were a
flake, it had been worked subsequent to its removal from the core. If it
were a core tool, it had to have been used. There were 567 tools recov-
ered from the trench.
2. Utilized flakes: A 4X hand lens was used almost exclusively to
determine that chipping on the edges had resulted from utilization. There
were 4,602 utilized flakes recovered from the trench.
3. Blades: Many triangular and trapesoid blades were recovered
typical of a polyhedral core and blade technique, but the cores were not
found. A possible explanation for the absence of blade cores is that only
the material directly under the cortex in Florida cherts was suitable for
the manufacture of an implement if a sharp edge were desired. This out-
er area, therefore, was stripped from the chert nodule and the nodule was
then discarded or used for another purpose. These blades are of interest
because they represent a blade tradition separate from projectile point man-
Florida Anthropologist, vol. 28, no. 4, December 1975
Fig. 1. Backhoe trench at the Senator Edwards Chipped Stone Work-
shop Site, (a) looking southeast, (b) looking northwest.
Table 1. Distribution of stone remains by 3-meter sections of the trench.
Blades Utilized Flakes
Totals 567 50,326 6,071 4,602
*Stone from Sections 13-16 came from clayey deposits which were diffi-
cult to screen without a water supply. Increased artifact count in Section
16 resulted from two crew members spending an entire day working by hand.
Fig. 2. Stratigraphic profile of 48 m trench.
ufacture. The typical point is 6-8 cm long, at least 2-3 cm wide, and
not usually very thin. Through the reduction process involved in stone
tool production, the projectile points from this site could not have been
manufactured from these blades (Fig. 4).
Many of the blades were snapped (Fig. 5) and resemble burins des-
cribed in various publications. It appears as if the technique described
by Movius et al. (1968:23) and Crabtree (1972:49) was used to truncate
some of the blades, but many could have snapped accidentally or through
use. Many of the blades have been utilized. There were 6,071 snapped
blades recovered. To my knowledge these have never been specifically
reported upon for Florida. They will be the subject of another paper.
4. Waste flakes: The remaining 50, 000 pieces of stone from the
trench were classified as waste but many had probably been used. Knife
use is not easily detected. These waste flakes form a valuable collection
available for study of manufacturing and chipping techniques.
The Senator Edwards Chipped Stone Workshop was occupied during
that part of the Florida Archaic which ranged from 5000 to 8000 or 9000
years ago. It might be possible to correlate the occupation of the site
with climatic changes in Florida during the Holocene. Ten thousand years
ago when sea level was 30 m lower than present, the watertable in the
limestone terrain which is graded to the sea, was also considerably lower.
This probably means that the sinkhole which is present at the site had no
available water supply, nor is there evidence that man was utilizing the
site at that time. There is always the possibility, of course, that man
hadn't discovered the area yet but this is unlikely since many Florida
paleo-points are found in Marion County. Sea level rose rapidly during
the Holocene and changes in vegetation and a fluctuating watertable occurr-
ed also. Sometime before 5000 years ago, well water was probably avail-
able in the sinkhole at the site.
There has been no pottery found at the site and no projectile points
which date more recently than 5000 years ago. It was puzzling why the
site was not utilized after this date but, again, a climatic solution (Brooks
1973) can be suggested. Sometime between 4500-5000 years ago, the lakes
went dry in northcentral Florida because of a drop in the watertable.
This meant that there may, once more, have been no available water at
the site. After this fairly short dry spell, the area became forested and
has been characterized by a hardwood forest ever since, that is, for about
4000 years. The muck and humus deposits at the site fall within this
period and contained no cultural materials. The watertable has fluctuated
slightly since thattime but has been fairly close to present for 4000 years.
A persistent wet period, however, plus the hardwood forest probably made
human habitation uncomfortable. It is not known whether this explains why
the workshop site was abandoned but it probably was a contributing factor.
Projectile points vary considerably in size but generally average 6-8
cm long, and are characterized by stemmed bases (Fig. 6). In addition,
there have been a number of Bolen points recovered. Bolen points report-
edly date around 9000 years ago (Bullen 1975:50-51). They are side-notched
points which are often bevelled. The majority of the Bolen points from the
Senator Edwards Site were bevelled (Fig. 7).
A breakdown of the other tools collected from the site includes: 750
broken projectile tips, 600 broken unstemmed bases, 350 broken stemmed
bases, 650 unifacial scrapers, 400 bifacial scrapers, 250 reforms, 25
hammerstones, and 12 large anvils. There are other cutting, piercing,
pounding, and scraping implements which apparently were multipurpose
tools. The knife category was difficult to classify; many of the implements
categorized as reforms may have been knives. In addition, there were
1600 "tools" which defied classification. These will be discussed below. No
drills were recovered from the site which is interesting since nicely made
stemmed drills are known from the Florida Archaic. This situation may
be easily explained, perhaps, if Witthoft (1968:13) is correct in saying that
these "drills" show no rotary wear and are probably knives that have been
"resharpened and resharpened. In an area where chert is plentiful, it
would not be necessary to use an implement so long since it could be easily
replaced. In this regard, there have been no stemmed scrapers recovered
from the Senator Edwards site.
Because it was not possible to collect the stone remains by natural
stratigraphic levels using the backhoe, another trench 9 m long and 1. 5 m
wide was excavated using controlled stratigraphic methods. The stone re-
mains from the second trench were similar to those recovered earlier.
The cultural materials were recovered from a sand zone but since there
was little quartz sand deposition in the area during the occupation period,
no chronological differences in the stone material could be detected via the
I have had considerable experience with thermal alteration of silica
minerals, especially Florida cherts (Purdy 1971; Purdy and Brooks 1971).
I began observing areas on many of the utilized flakes that looked as if
they had been subjected to fire. Only a portion of the flake or tool, how-
ever, had been heated (Fig. 8). Some of these implements had been used
as scrapers, others as perforators. There are ethnographic accounts des-
cribing stone tools being used with fire to accomplish a task--shaping fire-
hardened sticks, hollowing out logs for canoes, piercing holes in hides--but
I know of no account which describes the effect on the stone tools. Color-
ation like this will not occur in Florida cherts unless they have been ex-
posed to fire. The color of the specimen illustrated in Figure 8 is between
very pale orange and pale yellowish brown (10 YR 6. 5/2) except for the tip
which is between pale reddish brown and dark reddish brown (10 R 4/4)
(Munsell 1946). The stone must have been subjected to temperatures of at
least 2500C for color change to occur and temperatures of 3500C for thermal
alteration to occur. Some of the specimens exhibit the vitreous luster in-
dicative of thermal alteration.
About 1600 of the tools from the Senator Edwards Site were imposs-
ible to classify. This posed a frustrating problem until it occurred to me
that we cannot and should not type temporary or convenience tools the same
way or by using the same criteria that we use for the classification of fut-
ure use tools. We need to distinguish techniques from types. The shape
of the future-use tool is determined by the use to which it will be put. The
shape of the convenience tool is often determined by the use to which it was
put. In fashioning a future-use tool, the flintknapper follows a preconceived
trajectory of manufacturing techniques. But the final shape of the conven-
ience tool might result from (1) its original size and shape, (2) the use to
which it was subjected, (3) the length of time until the task was completed,
(4) removing flakes to rejuvenate an edge and how many were removed,
(5) or removing flakes where the flake itself is utilized, not the lump of
stone which may never have been used at all. Thus, the "tool" which re-
sults is not only the end product of a trajectory of manufacturing techniques
with a preconceived finished product in mind but results from a variety of
fortuitous events. Consequently, some specimens cannot be classified into
categories even though they may look fairly similar because of the nature
of flint fracture. Many were used as choppers, scrapers, cutting tools,
pounders, and some were used in combination, but others show no detect-
able use even under magnification. We should not, therefore, classify con-
venience tools on the basis of shape. They may resemble each other mor-
phologically but may have functioned quite differently (Fig. 9).
The foregoing discussion could apply to Clactonian flakes and chopper-
like cores (Oakley 1972:50) or the Chopper and Flake tradition (Bordaz 1970:
19) described for the Lower to Middle Pleistocene in the Old World. The
occurrence of a similar technology in Florida can be explained as a techno-
logical continuum from a time period when none of the stone remains could
be classified into true types; that is, before man had developed a trajec-
tory of manufacturing techniques. After man developed techniques to produce
definable types, he did not abandon a functionally advantageous way of getting
a job done. Progress is seldom accomplished through replacement but by
accretion. For that reason, it is not strange to find a chopper and flake
tradition in Florida side by side with the Levallois technique, blade manu-
facture, and projectile point production. What defines a time period in the
Old World becomes only one aspect of the stone industry in Florida.
A collection of this size recovered from the same site which falls with-
in a well-defined time period, the Florida Archaic, provides a rare oppor-
tunity to discover a great deal about man's use of stone.
Acknowledgments. The Division of Sponsored Research, University
of Florida, provided funds for the backhoe operation. I wish to express
my appreciation to the many students who volunteered their time to re-
cover the artifacts; to the Florida State Museum for providing technical
assistance and space to aid this project; and to Drs. E. T. Hemmings and
J. T. Milanich for their assistance in recording the stratigraphy. I would
especially like to thank Messrs. George Neal, Robert Keene, Alvin Hendrix,
and Senator L. K. Edwards without whose generous cooperation these ex-
cavations would not have been possible.
1970 Tools of the Old and New Stone Age. The Natural History Press.
Garden City, New York.
1973 Personal communication. Department of Geology, University of
Bullen, Ripley P.
1975 A Guide to the Identification of Florida Projectile Points, Revised
Edition. Kendall Books. Gainesville.
Crabtree, Don E.
1972 An Introduction to Flintworking. Occasional Papers of the Idaho
State University Museum, No. 28. Pocatello.
Movius, H.L., Jr.
1968 The Analysis of Certain Major Classes of Upper Palaeolithic Tools.
American School of Prehistoric Research, Peabody Museum, Bulletin
No. 26. Cambridge.
1946 A Color Notation, Tenth edition. Munsell Color Co. Baltimore.
Oakley, Kenneth P.
1972 Man the Toolmaker, Sixth edition. British Museum of Natural
Purdy, Barbara A.
1971 Investigations Concerning the Thermal Alteration of Silica Miner-
als: An archaeological Approach. Unpublished Ph. D. dissertation,
University of Florida, Gainesville.
Purdy, Barbara A., and H. K. Brooks
1971 Thermal Alteration of Silica Minerals: An Archeological Approach.
Science, Bol. 173, pp. 322-25. Washington.
1969 Lithic Materials and Technology. Southern Archaeological
Conference, Bulletin No. 9, pp. 3-8. Morgantown.
November 15, 1974
Fig. 3. Recovering stone remains from the backhoe trench,
Fig. 4. Blades and projectile point, relative sizes show former could not
have been blanks used to manufacture such projectile points.
Sene see s eers
Fig. 5. Snapped blades, note evidence of use on many of the illustrated examples.
Fig. 6. Archaic stemmed points from Senator Edwards site.
Fig. Bolen points from Senator Edwards site.
Fig. 7. Bolen points from Senator Edwards site.
Fig. 8. Flake from the site with dark area at tip, note burned area and utilized
edge in enlargement to right.
*i 'te ?
e e imier i
Fig. 9. "Tools" and, below, flakes similar in shape to flake scars on "tools.
: '' ;
PAINT ROCKS OF NORTHWEST FLORIDA
David C. Reichelt
Over the years this author has recovered many lithic artifacts from
the numerous prehistoric Indian sites to be found in the northwest Florida
area. While he was familiar with the various worked stone artifacts such
as points, tools and polished items there were a great number of amor-
phous rocks for which the intended use was uncertain. While most were
shapeless bits of rock there were a few that bore evidence of ware. The
author was compelled to call them hones and abraders, for lack of infor-
mation, and many of them may have seen such use.
Some of these rocks when recovered from water were found to give
off some color when rubbed on the hand. Also the author found one large
flat rock that had a ground-out cupped-shaped area that was heavily dis-
colored a dark red in the cupped area. It was not until recently with the
finding of a deposit of amorphous rocks of various kinds that also included
a well worn abrader that the author undertook to learn the possible use
of such rocks.
The above abrader was ground and shaped to be grasped on the side,
thereby leaving a smooth platform on which to rub whatever its owner was
working on. When rubbing one of the amorphous rocks on the platform only
produced a discoloring of the surface, the author dipped the rock in water
and placed a few drops of water on the platform. Rubbing again, the re-
sults were most amazing. Color literally flowed from the rock. It had com-
pletely mixed with the water and stayed mixed until it dried later. Each
of the rocks in the group gave off a different color or shade of yellow, red,
brown, orange, off-white, and black.
A question that now remained was whether or not the paint could
actually be used. Would it adhere and would there be sufficient consistency
to cover? To test this, the author acquired a mounted artist canvas and
using some twenty-five rocks began painting. The colors were applied with
the finger and each allowed to dry before the next was applied. The results?
As good as any commercial watercolor paints. In fact, maybe some could
be called better, due to the richness and naturalness of the shades.
It is suggested here that many of the various amorphous rocks from
our excavations should be given a similar test. It has been known, of
course, that Indians used limonitic and hemititic lumps as well as graphite
Florida Anthropologist, vol. 28, no. 4, December 1975
(where available) as paint materials. Apparently there are other rocks
and minerals that they might have used.
Santa Rosa Beach, Florida
November 26, 1974
192 OWNERSHIP STATEMENT AND CIRCULATION REPORT
In accordance with U. S. Postal Regulations No. 132. 622, the follow-
ing statement of OWNERSHIP, MANAGEMENT AND CIRCULATION for THE
FLORIDA ANTHROPOLOGIST is included in this issue.
Florida Anthropological Society, Inc.
c/o Ripley P. Bullen, Resident Agent
Room 102, Florida State Museum
Gainesville, Fl. 32611
Officers and Executive Committee
See inside front cover for names and addresses.
Total no. copies printed Average no. copies Sept., 1975 issue
Varies from 1000 each issue Vol. 28, no. 3
to 1150 1105 1028
Sold, U of F Library-Foreign 220 230*
Mail Subscriptions 724 637
Total Paid Circulation 944 867
Free Distribution 5 5
Total Distribution 949 872
Office Use, Left-Over 156 156
Total 1105 1028
*10 copies sold author, this issue only.