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 POSTL'S LAKE II SITE
EGLIN AIR FORCE BASE, FLORIDA (OK-71)
William C. Lazarus
The location of the Postl's Lake II site in Oklaoosa
County, Florida is shown in Figures 1 and 2. It is situated
on the Eglin Air Force Base Reservation along the North
Shore of Postl's Lake, which is separated by a narrow neck
of land from Choctawhatchee Bay. Postl's Lake is fresh wa-
ter which drains by filtration thru sand into the bay at one
or more locations. It is fed by springs within the lake and
the swampy areas at the western and eastern ends. It's lev-
el remains approximately that of the bay. The Lake depth is
estimated at 12 feet. A narrow fringe of reeds is along the
lake shore on the northern and southern sides.
The land on the north side of Postl's Lake rises quite
sharply to an elevation of about 6 ft near its mid-length.
On this relatively high ground, the site was located during
a systematic archaeological survey of a 40 acre track which
the Air Force had selected for development as a trailer vil-
lage to house personnel and families of the British Royal
Evidence of the site was found thru a surface showing
of shell in which two sherds were discovered. The survey
was conducted prior to any clearing. A dense stand of trees
consisting of live oak, magnolia, hickory and short leaf
pine covered the entire 40 acres. Thick undergrowth of pal-
metto, myrtle and yaupon augmented by sea grape vines and
many fallen limbs and trees made the area difficult to ex-
amine for archaeological evidence.
It was tentatively concluded that the site consisted of
a number of small midden piles or pits scattered along the
lakeshore for about 400 feet with none more than 95 feet
back from the water. The average diameter of these piles
appeared to be about 5 feet. However, one larger row of
midden was discovered at that time which appeared to be ori-
ented North and South.
Geologically there is no evidence that the terrain has
changed in any major way for several thousand years. The
surface soil is the characteristic white beach sand which
extends down to a depth of 18 inches to three feet in this
area. Beneath it is the coarser yellow sand typical of the
subsoil in the Choctawhatchee Bay region. Along the north
shore of Postl's Lake the yellow sand outcrops occasionally
where the bank is steepest in a typical erosion pattern.
Nothing at this site suggests that the terrain was different
during the time of Indian occupation.
The location of the site is such that it would be a
safe place of abode except in the most severe hurricanes.
On rare instances Choctawhatchee Bay has had hurricane wind
tides over 10 feet which would inundate this site. However,
six feet of elevation is considered safe for modern homes in
the Bay area. No evidence was found on this site that it
had been flooded although it probably was under water brief-
ly one or more times since its occupation by the Indians.
A permit to conduct work upon lands of the United States
under Act for the Preservation of American Antiquities Ap-
proved June 8, 1906 (34 Stat. 225, 16 U.S.C. 432, 433) was
issued by the Office of The Secretary, U. S. Department of
Interior on June 28, 1962, which permitted excavation at
this site. By the time it was received and coordinated with
the Base Commander at Eglin Air Force Base, bids had been
advertised for the clearing of the site and building of the
trailer village. Less than a month was available for sal-
vaging materials from this site. Since no funds were avail-
able for labor, volunteers who were available on week-ends,
had to be used.
The area selected for the initial excavation was the
north end of the largest midden. A grid of nine 5' x 5'
pits was laid out as shown in Figure 3. The order for ex-
cavation for these pits was: II, III, IV, V, VI, I and VII.
Pits VIII and IX were not excavated. All material was pro-
cessed thru x screen mesh. Table I shows the sherd
and artifact count by Pit and type and lists the identified
Pit II was excavated in 3" stratigraphic levels in the
hope that some stratigraphy would become apparent. The mid-
den was found to be only 6" deep on the average with all 78
sherds from this pit being clearly identifiable as Pensacola
Plain or Pensacola Incised. In the east corner of Pit II,
the midden went down to a depth of 10" in a small area.
Pit III was then excavated as one 9" level since no
atratigraphy was apparent in the previous pit and digging
time was preciously short. It became apparent as the west
corner was excavated that we were dealing with a Pensacola
Incised vessel broken in situ. About 1/3rd of the vessel
m A A4
\NOT\EXCAVATED NOT EXCAVATED
\ \ \test
\\- o 1
I,,\\ \\ \\ \ "1 <,\\ -?
K\\\\\\ \ \K\\\> I
KNOT EXCAVATED NOT EXCAVATE' NOT EXCAVATED
\ \ \ \1 \\ \ I
.\\ \ .\ \\x/
Excavations (Pits I thru VI)
was uncovered and the sherds taken out in sequential order
to facilitate restoration. A review of the sherds from Pit
II showed that nearly all of them belonged to this same ves-
sel. The south corner of Pit III showed another part of
the feature which had been observed in the east corner of
Pit II. This feature was determined to be a shallow pit
about 4" deeper than the midden and about 2 feet in diame-
ter. Pits IV and V further corroborated this when they were
Since some sherds, which seemed to be parts of the Pen-
sacola Incised vessel, were apparent in the northwest wall
of Pits II and III; a rectangular pit, designated IIIA and
measuring 2 ft. wide and 8 ft. long was next excavated.
Seventeen sherds important to the restoration of the vessel
Pit IV was next excavated. The edge of the midden ran
across the middle of this pit. Two sherds were found.
Pit V produced 2 sherds and a feature in the south
quadrant showed midden extending to a depth of 24" from the
surface in a two foot diameter circle. It is suggestive of
a post hole tamped with shell midden -and similar to ones
found on the Ft. Walton Temple Mound by Fairbaks in 1960.
The midden covered Pit V at an average depth of 6 inches ex-
cept in the extreme west corner where it ran out into white
Pit VI produced no sherds. The midden extended only a-
cross the east corner of this pit and was again about 6"
Pit I was next to be dug. It was sterile except that a
small circular midden pile appeared in the west corner. It
was about 3" deep and produced no artifacts.
Excavation of Pits I thru VI are shown photographically
in Figure 4.
Pit VII was the last of the excavated pits. It had
been partially disturbed by a bulldozer. One residual plain
(sand tempered) sherd and one Pensacola Incised sherd were
the only artifacts found.
Since work on the trailer park was progressing and fur-
ther screening was not practical, a post hole digger was em-
ployed to check the existence of shell midden at the grid
.intersection -points to the south of the excavated area.
This method grossly confirmed the north-south orientation of
Figure 6 W
Pensacola Incised Bowl with 4
strap handles Recovered from
this row midden which had appeared in the excavated pits.
The midden was found to be about 55 feet long with a rather
uniform width of about 10 feet. Where sampled by the post
hole digger its depth was close to 6".
Archaeological work on the site was then discontinued so
as not to interfere with the contractor clearing and build-
ing the trailer park.
Since the pit excavations had clearly confirmed the o-
riginal survey impression that this was a site of very lim-
ited occupation, it appeared logical that the location and
size of the many small middens might yield additional infor-
mation as to the nature of the occupation.
The Air Force specifications for the land clearing re-
quired that all trees larger than six inches in diameter be
preserved. The clearing technique was to remove the under-
brush with as little disturbance as possible to the root
systems of the trees. The contractor used tractor drawn
disks to cut the underbrush and palmetto in the ground.
Then a bulldozer with a rake-like attachment would push the
loosened growth moving as little earth as possible under
these circumstances. Observing the machinery at work over
known midden areas it was apparent that the shell piles were
fot grossly displaced although some shell was dragged to
the brush piles. All midden that was displaced by the bull-
dozers was easily identifiable by its "softness" and the
loose roots in it. Midden remaining in situ was firm and
the roots in it were anchored into undisturbed soil beneath
After the clearing and burning was basically completed
a week-end was selected when the contractor would not be
working. A plane table was set up and all firm middens were
plotted as to location, size and shape. A total of 20 dis-
tinct middens were found. (figure 5). The existence of a
firm row midden parallel to the one in which the pits were
dug, was discovered. It was narrower but almost of equal
length. The distance between was about 25 feet. The larg-
est row midden, after bulldozing, measuring the same width
as was measured before the bulldozing.
Four other smaller middens were found to be longer than
they were wide. These also were oriented North and South.
Considerable rain had fallen since the bulldozers had
finished in this area, and an intensive search for sherds in
0 50 100' Midden
o EXCAVATED AREA 6
S o 191 sq. ft. 4
6 0 OK -710 o
each of the surveyed middens was conducted. Only one sherd
- a Pensacola Plain rim was found on the entire site at
this time. It came from one of the westernmost midden piles.
In addition to retrieving artifacts from the screen,
periodic checks were made as to the types of shells and
bones which were present in the midden. Shells of the East-
ern Oyster dominated by a wide margin (possibly 85%), next
most common was the Duck Clam and a poor third was the Moon
Shell. Fish bones, notably vertebrae ranging from 1/2" to
3/4" diameter, predominated among the fauna, with some bird
bones and occasionally shell fragments of small turtles
(probably box turtles). (See Table I) Deer bone fragments,
usually expected in village middens in this region, were no-
tably absent. No evidence of agriculture was found.
No tools such as stone projectile points, bone fish
hooks or shell hoes were present to provide clues to the
subsistence pattern. The one chert chip indicates that
stone tools of some sort were in use, but this is a tenuous
bit of evidence.
Probably Related Site
On the neck of land which separates Postl's Lake from
Choctawhatchee Bay and directly across the lake to the south
from the Postl's Lake II site, there is another site desig-
nated as Postl's Lake III (OK-72). No archaeological exca-
vations have been made into that site. It's existence be-
came known after the start of excavations of OK-71. A sur-
face collection from the land, beach and shallow water area
produced a total of 142 identifiable sherds and several
large broken projectile points. Table 2 below shows the
classification of this collection:
Archaic Projectile Points 2
Stillings Island Fiber Temper 2
Semi-Fiber Temper 4
Linear Stamped 6
Bold Check Stamped 32
St.- Marks Plain (tetrapod) 1
Santa Rosa-Swife Creek Complex
Alligator Bayou Stamped 1
Swift Creek Complicated Stamped (Early) 1
Weeden Island Complex
Carrabelle Incised 1
Carrabelle Punctated 2
Keith Incised 2
Weeden Island Incised 6
Wakulla Check Stamped 8
Weeden Island Plain 40
Ft. Walton Complex
Pensacola Incised 4
Pensacola Plain 20
Ft. Walton Incised 3
Lake Jackson Plain 7
Jefferson Ware 2
Total sherds 142
This site is in no immediate danger of destruction al-
though the area is used to a limited extent for recreational
activities by the residents in the Postl Point trailer parks.
It is not easily accessible by land. About 500 feet of wa-
ter separates OK-71 from OK-72.
Although there were a few modern trash piles west of
this site (OK-71) and a general litter of beer cans over the
area, all available evidence indicates that this site had
not been disturbed since its aboriginal occupation. The
dense growth had served to seal it off and conceal it.
Compared with the average village site around Chocta-
whatchee Bay which is solidly covered with layers of midden
in considerable depth, this site is unusual because a dis-
crete midden pattern of a single culture appears to be pre-
sent. The 20 midden piles fall into just two categories, i.
e. circular and row middens. There are 14 circular ones and
6 row middens. All row middens are oriented North-South.
(Figure 5). The total midden deposit is estimated at 20 tons.
The existence of a North-South row midden was previous-
ly recorded as a feature of the La Casa Site in Santa Rosa
County (SA-12) Fla. At La Casa, the site was a pure Ft.
Walton Culture occupation which ran 97% shell tempered Pen-
sacola Series sherds. Here at Postl's Lake II the same gen-
eral situation appears to exist with a thinner midden con-
taining less artifacts but with the shell tempered component,
based on 129 sherds, higher than 99%.
Two possible explanations have been advanced as to the
significance of the row type of midden the Ft. Walton Cul-
First, it has been proposed that these rows of midden
are man-made garden plots. The coastal sands along the bays
of Northwest Florida are not in the least conducive to pri-
mitive agriculture. The Ft. Walton Culture has been identi-
fied as agricultural at many inland sites along the Gulf
Coastal plain. Domesticated corn has been found in a Ft.
Walton Mound site in Houstin County, Alabama and at other
locations. Dr. J. M. Goggin has proposed that the Ft. Wal-
ton people farmed the middens of previous cultures where
these occurred on a sandy soil on the Northwest Florida Coast.
(personal communication). Extrapolating this hypothesis to
sites like OK-71 and SA-12 where no ancient middens are a-
vailable for farming, it is conceivable that the Fort Walton
people spread their own midden into rows to create gardens.
Why such rows would be oriented North-South is not clear.
A second possible explanation, and the one favored by
the author, is that the row middens are a manifestation of
the village pattern. This is based on the theory that mid-
dens will occur at the nearest convenient location to the
living area. Middens rich in shell would be uncomfortable
to walk upon and further there would be an odor and insect
problem initially. Village pattern, prevailing wind, and
terrain features would influence midden locations in the or-
der given. The absence of land depressions and other criti-
cal terrain features at this site eliminate this latter as-
pect as an influencing factor at this site. The prevailing
winds would have a southern component much of the year at
There are several references in diaries of early ex-
plorers to the fact that the doorways of Indian houses were
to the south thus preventing cold northern winds from blow-
ing in, and affording the maximum of sunlight into the door-
Assuming a southern orientation of the doorways on the
houses at this site, and assuming that midden would build up
initially along the side of the houses to the right or left
of the doorway, it seems that the location of five or six
dwelling units at this village can be identified, generally
they are on the west side of the row middens. The two larg-
est middens obviously indicate the location of the oldest or
largest dwelling or dwellings. Perhaps the spacing between
the two largest middens on the eastern half of the site pro-
vides clues as to the shape and width of dwelling units.
The opinion is that at least one of these houses was rectan-
gular with a maximum width not exceeding 25 feet. This sug-
gests a "long house" configuration as illustrated by the
French artist LeMoyne in 1564 in his drawing titled "Con-
struction of Fortified Towns among the Floridians".
The most significant artifact recovered in this project
is a Pensacola Incised bowl with high collar and four strap
handles (Figure No. 6). Only under rare circumstances have
restorable vessels been found in Fort Walton Period middens.
Of the 129 sherds recovered, 120 were identified with this
This vessel, coming from a midden may provide seme in-
sight into the ratio between decorated and plain sherds as
they relate to whole vessels of the utility type at Ft. Wal-
ton period village sites. If just 13 sherds are removed
from the 120 identified with this vessel, one would call
this bowl a plain vessel. Only about 10% of the sherds be-
longing to this vessel contain any clue as to its decoration.
The adornment of this vessel consists of a single in-
cised line in 6 scallops with a double row of punctations
above it plus four strap handles. Decoration is close to
the minimum, although the finished product is clearly dis-
tinguishable from plain forms.
The identifiable residue of foods found in the midden
indicates an economy largely oriented to the sea and its
bays. The shell and small fish bones suggest gathering as
the principal technique used. The bird bones indicated some
hunting. The absence of deer bones in this midden is diffi-
cult to explain. They are common at most middens in this
area. Agriculture would have had to be confined to the mid-
dens. In no case could it have figured prominently in the
subsistence pattern of this small village.
Possible Relationship with Adjoining Site (OK-72)
The site (OK-72) located 500 ft south across Postl's
Lake shows a substantial Ft. Walton period occupation being
exceeded only in sherd count by Deptford and Weeden Island.
These latter two cultural periods are much longer in time
span than Fort Walton, so the concentration of people on the
site must have been higher in Ft. Walton times than during
any other period. The area of the site OK-72 is limited by
the Bay and Postl's Lake. Therefore if it became crowded or
if incompatibility developed during Ft.Walton times it would
be logical that a suburb would develop on nearby desirable
land. OK-71, the site upon which this report is based, may
well have been a suburb of a main village located to the
south across Postl's Lake.
1. This site, OK-71, Postl's Lake II, indicates a small
Ft. Walton period village consisting of not over six dwel-
ling units, based on the location, size, shape and orienta-
tion of the middens.
2. The larger row middens suggest at least one rectang-
ular type of house with maximum width not exceeding 25 feet.
3. The Pensacola Incised vessel provides a good example
concerning the ratio of decorated to undecorated sherds
which can represent whole decorated vessels in a Ft. Walton
midden. Only 10% of the recovered sherds need to be decora-
ted to indicate a high percentage of decorated vessels in a
Ft. Walton village midden.
4. This site is probably a suburb of the adjoining site
southward across Postl's Lake. If the main village is clos-
er to the bay it further substantiates the evidence found at
OK-71 that the subsistence pattern of the coastal Ft. Walton
culture was oriented toward the sea rather than agriculture.
The entire project was carried out by volunteers. With-
out their efforts this report would not have been possible.
The author is indebted to the following members of the Flo-
rida Anthropological Society for their assistance: Mr. Ger-
ald A. Spence, Mr. Jack Webb and Mrs. Yulee W. Lazarus. The
labor contributed by Mr. Paul Deglau, Mr. James Kendrick and
Chief Warrant Officer Jack Elliott is also gratefully ack-
The full cooperation of the Commander, Eglin Air Force
Base and the Commander, Air Proving Ground Center, Air Force
Systems Command is sincerely appreciated.
The total effort expended on this project is divided as
Initial Site Survey (40 acres) 12 man hours
Excavation (7 Pits) 102 "
Surface collection at nearby site 6 "
Processing artifacts 6 "
Restoration of Pensacola Incised vessel 75 "
Plane Table Survey of middens 10 "
Preparation of report 40 "
Total 251 "
The restored Pensacola Incised vessel, which is the
property of the United States Government, is under the cus-
tody of the Department of Anthropology,Florida State Univer-
sity, Tallahassee, Fla. It is planned to place it on tempo-
rary display in a suitable case at the Elementary School for
British Children on Eglin Air Force Base, Fla., since it was
recovered in the area where they are making their temporary
home while in America. Later it will be deposited for safe
keeping with the Temple Mound Museum of the City of Fort
Walton Beach, a publicly owned museum, with facilities for
proper storage and safe keeping, or placed in the University
A popular account of the work accomplished under this
project appeared on page 3 of the Eglin Eagle, the Base
newspaper, under date of August 3, 1962. Four photographs
accompanied the full page feature story written by Sgt. Mike
Lucas of the Eagle Staff. Appropriate warning is contained
in this article against the pilfering of any archaeological
sites on the Eglin Reservation.
Fairbanks, Charles H.
1960 Excavations at the Fort Walton Temple Mound, Re-
port to the City of Ft. Walton Beach, Fla.
1961 The Fort Walton Culture East of the Appalachicola
River, Fla. Newsletter, Southeastern Archaeologi-
cal Conference, Macon, Georgia.
Lazarus, William C.
1961 Ten Middens on the Navy Live Oak Reservation,
Santa Rosa County, Fla., Florida Anthropologist,
Vol. XIV Nos. 3-4, Tallahassee, Fla.
1961 The Ft. Walton Culture West of the Appalachicola
River, Fla., Newsletter, Southeastern Archaeologi-
cal Conference, Macon, Georgia
Neuman, Robert W.
1961 Domesticated Corn from a Ft. Walton Mound Site in
Houston, Co., Ala. The Florida Anthropologist,
Vol XIV, Nos. 3-4, pp 75 Dec 1961, Tallahassee,
Fundaburk, Emma Lila
1958 Southeastern Indians Life Portraits, Illustration
No. 30, Luverne, Ala.
Bullen, Ripley P.
1958 Six sites near Chattahoochee River in the Jim
Woodruff Reservoir Area, Fla. Bureau of American
Ethnology Bulletin 169 River Basin Surveys Papers
No. 14 pp 315-357, Washington, D. C.
Willey, Gordon R.
1949 Archaeology of the Florida Gulf Coast, Smithson-
ian Miscellaneous Collections, Vol. 113, Washing-
ton, D. C.
Sherd Pit No.
Classification I II III IIIA IV V VI VII Total
Pensacola Incised 0 10 7 0 1 0 0 1 19
Pensacola Plain 0 63 26 17 1 2 0 0 109
Residual Plain 0 0 0 0 0 0 0 1 1
Totals 0 73 33 17 2 2 0 2 129
Other artifacts: 1 chert chip from Pit II
Shells identified in the midden: Eastern Oyster (85%); Duck
Clam (12%) Moon Shells & Misc. (3%)
Other identified materials in midden in order of frequency :
fish bones, bird bones, box turtle shell.
Ft. Walton Beach
THE ASSOCIATION OF SUWANNEE POINTS AND EXTINCT ANIMALS
Wilfred T. Neill
Some Suwannee points closely resemble Clovis
points, which have been found in association
with the remains of extinct animals, especially
elephants. Suwannee points have been found in
similar association. Several considerations
suggest a possibility that the makers of the
Florida points were also elephant hunters.
The Cavern Site, submerged beneath the waters
of a large spring in Marion County, Florida,
yielded a Suwannee point of a Clovis-like vari-
ety, other artifacts including a utilized flake
and spalls with worked edges, a thin sheet of
elephant ivory, and remains of man, mammoth,
mastodon, and large horse. It is suggested that
the animal bones represent the prey of early In-
dians who frequented the cavern in Paleo-Indian
times, when the water table was much lower than
at present. However, alternative explanations
of the association cannot be ruled out entirely.
In parts of the western United States, Clovis projec-
tile points have been found associated with the remains of
extinct elephants, especially around water-holes which later
dried up (Haury et al., 1959). Sellards (1952: 17-46) rec-
ognized a "Llano Complex," diagnosed mainly by Clovis points
and similar but unfluted points, made by "elephant hunters."
Sellards' action appears justified in the light of more re-
cent work, and Llano Complex artifacts have been found in
eastern United States (Haury et al., op. cit.; Mason, 1962:
In Florida, some very Clovis-like points have been
found (Bullen, 1962: fig. 2, three specimens at left, bot-
tom row; Griffin, 1952: 21, upper figure, center specimen;
Neill, 1958: pl. 3, A, C; 1961: 13, left figure, bottom row
of specimens; J. C. Simpson, 1948: fig. 3, A, C). In recent
years these Florida artifacts have generally been placed by
local workers in a rather broad category called "Suwannee
point" (e.g., Bullen, 1958: 28-29; Goggin, 1949: 20; 1952:
64-65, footnote), although Mason (op.cit.: 240) evidently
would extend the term "Clovis" to include some Florida spec-
imens, and restrict the term "Suwannee" to much less Clovis-
like specimens such as those figured by Goggin (1950: fig.
The Clovis-like specimens mentioned above, and some
much less Clovis-like ones (e.g., Goggin, 1950: fig. 21, P),
represent opposite ends of a typological continuum, other
Florida specimens falling between these extremes (e.g.,
Bullen, 1958: pl. 1, L; 1962: fig. 2; Goggin, 1952: pl. 7,
R-S; Griffin, loc. cit., first and third specimens; Neill,
1958: pl. 3, G,J). A specimen (Neill, 1958: pl. 3,G), cited
by Mason as exemplifying his concept of a Suwannee point,
actually is a knife. Elongate, roughly worked knives, with-
out basal (lateral) grinding, occur on Suwannee campsites
along with the basally ground projectile points, which lat-
ter are of different, and generally somewhat more Clovis-
A small series of artifacts, assembled by the late Dr.
John M. Goggin at the University of Florida's Department of
Anthropology, was considered by Mason to be a "type series"
for the name "Suwannee point." These artifacts, recently
shown me by Dr. Charles H. Fairbanks, are not close to the
modal condition of early lanceolate points from peninsular
and northeastern Florida.
For the present I continue to use the name "Suwannee"
for a considerable variety of lanceolate points, some much
more Clovis-like than others. Figure. 1 compares true Clovis
points, from localities in the High Plains and Midwest, with
some exceptionally Clovis-like Suwannee points from Marion
County, Florida. Of the Florida specimens, the second from
the left has a repair on the tip and on one basal "ear;" the
others are as they were found. The first two in the row of
Florida specimens are from the deepest occupation level of
the Silver Springs Site (Neill, 1958); the other three were
found in spread sand dug from the hillock on which this site
It may be mentioned that Suwannee points, unlike Clovis
points, are never fluted on both faces, although a weak uni-
facial fluting is often present on specimens from the Silver
In Florida, Suwannee points have been found associated
with the remains of extinct elephants, in water-holes which
did not dry up. (I use the term "elephant" to designate
both mammoth and mastodon, a convenient procedure here al-
though the two belong to separate families of the Suborder
Elephantoidea.) It seems desirable to explore the possibil-
ity that the Florida associations, like those of the South-
west, might represent Paleo-Indian kill-sites; and to review
the evidence for a contemporaneity of Suwannee points and
extinct animals in Florida; as well as to describe an inter-
esting new Suwannee point site in Marion County.
1. Suwannee points are strongly associated with aquat-
ic situations, especially large springs and spring-fed riv-
These points have been found on the bottom of springs
or streams at the Withlacoochee River, Citrus-Marion County
line; Ichetucknee River and several of its tributary spring
runs, Columbia County; mouth of the Santa Fe River, Dunne-
gan's Mill on the Santa Fe River, 2 miles west of High
Springs on the Santa Fe River, and Lilly Springs, Gilchrist
County; Suwannee River at and near White Springs, Hamilton
County; Silver Springs, Marion County; Silver Glen Springs,
near Marion-Lake County line; and River Springs, Suwannee
County. Suwannee point sites also occur on higher ground
overlooking springheads, streams, large swampy basins, and
lakes along streams. These sites include Biven's Arm, Darby
Springs, Hornsby Springs, and Payne's Prairie, Alachua Coun-
ty; end of Orange Lake near Alachua-Marion county line;
Crystal River, Citrus County; high ground overlooking the St
Johns River valley near Paisley,Lake County; Natural Bridge,
Leon County; Wekiva Springs, Levy County (not to be confused
with Wekiwa Springs, Orange County); Salt Springs, Silver
Springs (3 sites on high ground bordering the spring run),
and near Lake Weir, Marion County; and Watson's Landing, Vo-
lusia County. Suwannee points have also been found either
in or around Wakulla Springs, Wakulla County. Other Suwan-
nee point sites, such as Lake Helen Blazes in Brevard County
and two or three localities in Alachua County, probably
should be added to this list; but in these cases I lack data
as to site location with reference to water.
The above list of Suwannee point localities is based on
published accounts (Bullen, 1958, 1962; Goggin, 1950, 1952;
J. C. Simpson, op. cit.); personal observations, some pub-
lished (Neill, 1958, in press); and specimens in the respec-
tive collections of V. J. Allen, Foster Barnes, C. E. Burk-
hardt, Gordon Coates, L. H. Pharmer,Jr., and W. A. Franklin,
Of 51 Suwannee point localities tabulated by me, at
least 31 are in water or overlooking it. (This tabulation
does not include specimens dredged from the Gulf of Mexico.)
No other type of flint projectile point is so strongly asso-
ciated with aquatic situations in Florida.
2. Elephant remains also occur abundantly in springs
and spring-fed rivers of peninsular and northeastern Florida.
G. G. Simpson (1929: 269-277) listed a number of mam-
moth and mastodon finds from peninsular springs and streams,
J. C. Simpson (op. cit.: 13) stated, "Most of the findspots
Lof Suwannee points have been instream beds, or in the im-
mediate vicinity of streams. /One specimen/ was found in
the Ichetucknee River with fossil vertebrate remains and
fossilized artifacts of bone. Lt7 this locality were also
found three fossilized ivory points, similar to those found
at Clovis, New Mexico, and ... a chert scraper was found in
place below a partly articulated mastodon skeleton. /Another
Suwannee point/ was found in the Santa Fe River with fossil
vertebrate remains and fossilized artifacts of bone." (One
supposed ivory point has since proven to be of bone.)
Olsen (1958) described the finding of numerous mammoth
and mastodon bones, with bone projectile points, in Wakulla
Springs, Wakulla County. From in or beside these springs I
have also seen three Suwannee points.
Artifacts and elephant remains from Silver Springs are
discussed in another section of the paper.
3. The large springs of peninsular Florida probably
were water-holes in Suwannee times.
By "water-holes" I mean sources of water, permanent or
fairly so, in country where other water sources are limited,
at least in some seasons. Water-holes attract wild animals,
and primitive people with a hunting economy. It is not in-
tended to suggest that the rainfall of Suwannee times was
less than at present; I am here concerned with the effect of
a water table that was far lower than the modern one.
At Warm Mineral Springs (= Warm Salt Springs), Sarasota
County, the bones of seven human beings were found 80 feet
deep in water, in a cave from which the spring waters now
flow. The cave had stalactites, which do not form under wa-
ter. Projectile points were not found. A burnt log from
the cave was dated by radiocarbon determination. It is re-
alized that the log and the skeletal remains may not be of
the same age, and that radiocarbon dates are often chal-
lenged. At any rate, the date of the log, 10,000 I 200
years B.P., conforms to expectation if the deeply submerged
cave burials were those of Suwannee point people.
Shepard and Suess (1956) showed evidence that, about
10,000 years ago, sea level was 80 to 135 feet lower than at
present. The water table in Florida should therefore have
been much lower than at present. Even with abundant preci-
pitation, the ponds and swamps, today innumerable, would
then have been few, except perhaps in flatwoods where the
hardpan prevented the percolation of rainwater into the
ground. Under such conditions, most of what is now the Flo-
rida peninsula would have been "central highlands" with wa-
ter sources restricted to the larger springs and a few riv-
'In connection with rise of sea level since Suwannee
times, it is interesting to note that Suwannee points have
been dredged up from beneath the waters of the Gulf of Mexi-
co, at several localities in the Tampa Bay region (Ripley P.
Bullen, personal communication based on information from Dr.
Lyman 0. Warren).
4. At two larger springs, artifacts and elephant bones
have been found together in submerged caverns, in positions
hardly attained by the action of water.
At Wakulla Springs, enormous tusks, bones of mastodon,
ground sloth and deer, charred wood, and over 600 bone pro-
jectile points, were found about 300 feet back in a horizon-
tal cavern from which water now flows (Olsen, op. cit.).
The remains and artifacts were concentrated in a small area
at depths of 200 to 220 feet. Olsen thought the cave could
never have been dry at these depths, and could account for
the location of the objects only by the assumption that they
had arrived through a sinkhole, since vanished.
At Silver Springs, artifacts and elephant remains have
been found in a submerged cavern, under somewhat less equiv-
ocal conditions. The find is discussed in another section
of the paper.
Bone points have been found in several springs, but are
extraordinarily abundant at Wakulla, where they are concen-
trated in the deepest spots of the springhead floor. A se-
lection of Wakulla specimens is illustrated (Fig. 2). All
of these are completely mineralized, of stone-like consis-
tency. Usually, one end of the artifact is sharp, and
smoothly polished; the other end is bluntly pointed, and may
show marks of both cutting and abrading. Evidently this
type of bone point was hafted by inserting the blunt end in
a socket. However, the smallest illustrated specimen is
sharply pointed at both ends, and transversely scored a lit-
tle above the middle; presumably it was bound across the
scored area to the beveled end of a shaft (Fig. 2, upper
right). Finally, V. J. Allen (personal communication), who
Figures 1 & 2
amassed a large collection of bone points from Wakulla,
found three with a beveled base which presumably fit against
a reverse bevel at the end of a shaft.
Of course the bone points from Wakulla and other Flo-
rida localities at present cannot be dated, but they call to
mind the bone points of Clovis sites such as Blackwater No.
1 in New Mexico (Sellards, o2. cit.: fig. 14).
5. There is some other evidence of a contemporaneity
of man and elephants in Florida.
In the bed of the Ichtucknee River, J. C. Simpson found
Suwannee points, remains of elephants and other extinct ani-
mals, and beveled ivory points resembling bone points from
Clovis sites in New Mexico (Jenks and Simpson, 1941). J. C.
Simpson (op. cit.) mentioned three ivory points from the
Ichtucknee, but Cotter (1962: 251) stated that one of these
specimens is of bone, while two are of ivory. It might be
argued that these two latter points were made from fossil
ivory; but in Florida, fossil ivory is chalky, inclined to
split into sheets, and unsuitable for manufacture into
points (Fig. 4, no. 2). As Cotter was impressed by the ty-
pological similarity of the Florida specimens to New Mexican
ones excavated by him, and as the latter artifacts were made
by people who preyed upon elephants, it is not unreasonable
to ascribe a similar origin to the Florida objects.
The Suwannee point level of the Silver Springs Site
an unidentified object (Neill, 1958: 42). Just possibly
this is a small fragment of a point resembling those found
by Simpson; I figure it here for the first time (Fig. 4, no.
6. Actual Suwannee habitation sites have yielded a var-
iety of artifacts. The Silver Springs Site, and the Snyder
Site in Lake County, between them have produced elongate
knives, snub-nosed end-scrapers, choppers, retouched flakes,
utilized flakes, flake-based gravers, graver-perforators,
sandstone abraders, and possible ornaments, as well as pro-
jectile points. On the other hand, the stream-bed local-
ities, such as those discussed by J.C. Simpson, have yielded
projectile points of flint, bone and ivory, along with the
remains of animals, but nothing else except one scraper and
a few stone "clubheads" not certainly of Suwannee origin.
The circumstance implies that the stream-bed localities were
kill-sites, where the Indians lost some weapons but little
else. The situation parallels that obtaining in the West,
where Clovis habitation sites produced a rich inventory as
compared with kill-sites which yielded only projectile
points and a few cutting or scraping tools.
I I -,
.' *A'tr ^^
; ~. .. ~~il-~
Ripley P. Bullen (personal communication) stated that
a search of the Ichtucknee area revealed no habitation sites
from which Suwannee points might have washed into the stream.
This observation is consistent with the idea that J.G. Simp-
son's stream-bed localities, with Suwannee points and animal
bones, were kill-sites.
A Submerged Suwannee Point Site
In 1949, great numbers of animal bones, especially mas-
todon and Columbian mammoth, were brought up by divers at
Silver Springs, from the floor of the headsprings. Exten-
sive construction work around the headsprings has revealed
no fossiliferous stratum from which the bones could have
eroded. Thus it is believed that the elephants died in the
In 1953, divers began to explore a roughly horizontal
underwater cavern at Silver Springs. The mouth of this cav-
ern is the opening of the largest spring in the group. The
bottom of the cavern mouth is about 35 feet below the sur-
face of the water. Animal bones, especially mastodon but
also Columbian mammoth and a few other species, were abun-
dant on the floor of the cavern,from the entrance to a point
about 45 feet back from the mouth. Farther back, the cavern
narrowed to a series of small tunnels and fissures, too
saall for entrance. The remains included an enormous ele-
phant tusk which could hardly have arrived through one of
the fissures. Some of the bones from the cavern are shown
in Figure 3. I regret that I cannot supply precise ident-
ification of all bones from the cavern; but I was only a by-
stander during these operations, and did not have chance to
examine any of the remains closely.
Subsequent diving in this cave also brought up a frag-
ment of a human skull (Fig. 4, no. 3), and a flint object
(Fig. 4, no. 4). This latter may be a fragment of a Suwan-
nee knife, comparable to one from the Suwannee point level
of the Snyder Site (Fig. 4, no. 5) and to one from the Su-
wannee point level of a site about a half-mile below the Sil-
ver Spring Site on high ground overlooking the run (Fig. 4,
no. 6). This latter site is the one described briefly by
Bullen (1958: 28-29).
The identification of the cavern fragment as a knife
might be questioned, for it is about 15 mm. thick; but the
specimens figured for comparison are respectively about 13
and 14 mm. in thickness. At any rate the cavern specimen is
artifactual, with two worked edges.
At least on piece of charcoal of charred wood was
found in the cavern.
The topography of the springhead floor is such that, if
the water table dropped no more than 45 feet, the cavern
would be a dry and roomy cave; and further, at the cavern
mouth there would be a small ledge on which one could sit
and look out over the springhead basin. Thus, in 1960, with
the assistance of a local diver, I instituted a search for
small objects that might have remained in situ in the sand
of the ledge. This was done by raking with the fingers
through the sand. The procedure yielded the following:
1. A Suwannee point, of a distinctive variety (Fig. 5,
no. 1). I have previously illustrated this specimen in a
semi-technical article (Neill, 1961: 13), as exemplifying an
approach to the Clovis point. (For illustrative purposes at
that time a small repair was made to a broken basal "ear.")
Recently, Bullen (1962: fig. 4) figured a very similar Lee
County specimen, which he called "Suwanneelike" rather than
Suwannee, and which he feels (personal communication) may
have closer Clovis affinities than the usual Suwannee point.
The specimen from the cavern ledge is weakly fluted on
one face. The fluting was, however, produced by the removal
of two or perhaps even three spalls, not by the removal of a
single channel flake.
2. An object resembling a utilized flake, but seeming-
ly made of animal material (horn?)rather than flint (Fig. 5,
3. One thick spall with a worked bifacial edge (Fig. 5,
4. One thin spall. with a finely worked unifacial edge
(Fig. 5, no. 4). The serration of the edge seems too reg-
ular to have been produced simply by use.
5. One utilized flake (Fig. 5, no. 5).
6. One fragment from the outer surface of a mastodon
tooth (Fig. 5, no. 6).
7. One molariform tooth of a large horse (Fig. 5, no.
7). Identification was supplied by Dr. Clayton E. Ray, of
the Florida State Museum.
8. One thin sheet of elephant ivory (Fig. 5, no. 8).
Two edges of the sheet, forming an approximate right angle,
r d Ld ,- al F9l 5' 7 A ,b i As Ap
IN gure 5
are straight, vertical or slightly rounded; the other edges
are concave, clearly broken and not cut. All the edges re-
semble the flat surfaces in being brownish and somewhat
eroded. When a piece was snapped off the object, the newly
exposed edges were white and smooth. Therefore the original
breaking (and cutting?) of this piece of ivory did not take
place in recent times.
Also found were 10 spalls of flint (8 of them shown in
Fig. 5, unnumbered). Some of this flint is identifiable as
a material that crops out as inclusions in the local lime-
stone. Several spalls have a jagged appearance, like many
from the Suwannee point level of the Silver Springs Site.
Items from the ledge show no sign of abrasion or stream-
rolling. Patination of the flint is but moderate, as might
be expected; the chemical processes of flint patination,
whatever they may be, evidently do not take place with any
rapidity beneath the waters of Florida's large calcareous
It is believed that the material from the ledge is
attributable to Suwannee occupation of the cavern at a time
when the water table was much lower than at present,and that
subsequent rise of water table did not significantly disturb
the position of these objects. The cavern and ledge are
thought to constitute a "site" in the archeological sense,
hereafter called the Cavern Site.
Almost surely the abundant animal remains were in the
cavern in Suwannee times. One cannot rule out the possibil-
ity that they were brought in long before Suwannee occupa-
tion, by some huge predator such as dire-wolf, extinct bear,
extinct jaguar, or saber-tooth. Nor can one rule out the
possibility that Suwannee people found fossil bones in the
springhead, and brought them into the cavern.
As Florida has many large calcareous springs, it is
likely that other submerged caverns will yield artifacts and
the remains of extinct animals. It is to be hoped that fu-
ture divers, making such a find, will not hasten to bring
the material to the surfact, but rather will first metic-
ulously record the position of the bones with reference to
each other, the cave walls, the substratum, and other cave
contents such as artifacts; and further, will handle the
bones with extreme care, so that they may be examined for
possible marks made by flint knives or projectile points or
heavy Stones, as well as for possible tooth-marks of car-
nivores. It should also be noted that submerged bones com-
monly fall to pieces as soon as they begin to dry out; and
adequate preparation should be made for the preservation of
the bones as soon as they are brought to the surface.
As noted, there are several possible explanations of
the situation at the Cavern Site. But again it must be re-
called that, in the West, lanceolate points and the bones of
extinct animals are found in caverns and former water-holes
under circumstances demonstrating that the beasts had ac-
tually fallen prey to early Indian hunters.
Nearly all the extinct animal species, preyed upon by
Llano hunters of the West, were once present in Florida, or
represented in the latter area by closely related species;
and most of the artifacts of the Llano Complex have close
counterparts in Suwannee sites.
Mastodon remains have predominated at the Florida spring
and stream-bed localities that have yielded Suwannee points
in close proximity to the bones of extinct animals. A num-
ber of radiocarbon dates for the mastodon in the East are
well within what should be Suwannee times. These dates have
ranged from about 13,000 down to 6,000 years B. P., with
several in the 8,000-9,000 bracket (Byers, 1962; Williams,
The contemporaneity of man and extinct animals in Flor-
ida has not been demonstrated beyond dispute; but a growing
body of evidence suggests that the makers of Suwannee points
were actually in contact with at least a remnant of a Pleis-
tocene fauna that has since vanished.
Bullen, Ripley P.
1958 The Bolen Bluff Site on Payne's Prairie, Florida.
Contr. Fla. State Sus., Soc. Sci., No. 4.
1962 Suwannee Points in the Simpson Collection. The
Fla. Anthropologist, Vol. 15, No. 3, pp. 83-88.
Byers, Douglas S.
1962 Comments. Current Anthropology, Vol. 3, pp. 247-
Cotter, John L.
1962 Comments. Ibid., pp. 250-252.
Goggin, John M.
1949 Cultural Traditions in Florida Prehistory. Pp.
13-44 in John W. Griffin ed., The Florida Ind-
ian and His Neighbors. Winter Park, Fla.
1950 An Early Lithic Complex from Central Florida.
Amer. Ant., Vol. 16, No. 1, pp. 46-49.
1952 Space and Time Perspective in Northern St. Johns
Archeology, Florida. Yale Univ. Pubs. Anthr.,
Griffin, John W.
1952 Early Hunters of Florida. Fla. Wildlife, Vol. 5,
No. 10, pp. 20-21, 34-35.
Haury, E. W., E. B. Sayles, and W. W. Wasley
1959 The Lehner Mammoth Site, Southeastern Arizona.
Amer. l. 25, No. 1, pp. 2-30.
Jenks, A. E., H. H. Simpson
i941 Beveled Artifacts in Florida of the Same Type as
Artifacts Found near Clovis, New Mexico. Amer.
Ant., Vol, 6, No. 4, pp. 314-319.
Mason, Ronald J.
1962 The Paleo-Indian Tradition in Eastern North Amer-
ica. Current Anthropology, Vol. 3, No. 3, pp.
Neill, Wilfred T.
1958 A Stratified Early Site at Silver Springs, Flori-
da. The Fla. Anthropologist, Vol. 11, No. 1, pp.
1961 Giant Rattlesnakes -- Past and Present. Fla.
Wildlife, Vol. 15, No. 1, pp. 10-13.
In press Trilisa Pond, an Early Site in Marion County,
Florida. The Florida Anthropologist.
Olsen, Stanley J.
1958 The Wakulla Cave. Natural History, Vol. 67, No.
7, pp. 396-398, 401-403.
Royal, William, and Eugenie Clark
1960 Natural Preservation of Human Brain, Warm Mineral
Springs, Florida. Amer. Ant., Vol. 26, No. 2, pp
Sellards, E. H.
1952 Early Man in America. Univ. Texas Press, Austin.
Thepard, F. P., and H. E. Suess
1956 Rate of Postglacial Rise of Sea Level. Science,
Vol. 123 (3207), pp. 1082-1083.
Simpson, George G.
1929 The Extinct Land Mammals of Flori Twentieth
Ann. Rept. Fla. Stat --ol. Surv., 19Lz-1928, pp.
Simpson, J. Clarence
1948 Folsom-like Points from Elor-o Fla. An-
thropologist, Vol. 1, Nos. 1-2, pp. -14.,
1957 The Island 35 Mastodon: Its Bearing on the Age
of Archaic Cultures in the East. Amer. Ant., Vol
22, No. 4, Pt. 1, pp. 359-372
New Port Richey, Florida
I nt., .