2000 Florida Anthropological Society Inc.
The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
all rights to the
and shall be
and images of
The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.
The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.
Contact the Florida Anthropological Society for additional
information and permissions.
THE FLORIDA ANTHROPOLOGIST is published quarterly in March,
June, September, and December by the Florida Anthropological Society, Inc.,
at 3400 East Grant Avenue, Orlando, Florida 32806. Subscription is by mem-
bership in the Society for individuals interested in the aims of the Society.
Annual dues are $4.00; student membership $2.00. Requests for membership
and general inquires should be addressed to the secretary; subscriptions, dues,
changes of address, and back issue orders to the treasurer; newsletter items
to the president; and manuscripts for publication and books for review to the
editor. Second class postage paid at Orlando, Florida.
Volume XXIV, No. 4
The Primitive Solar Observatory at Crystal River and its Implications
Clark Hardman, Jr . .. .. . .... . . . . .135
Excavation of a Deptford midden burial, Destin, Florida
Jennings W. Bunn, Jr .. . . . . . . . . .. 169
Hafted Flake Knives
Ben I. Waller . . . . . .
OFFICERS OF THE SOCIETY
President Carl A. Benson
3400 East Grant Ave. Orlando, Fla. 32806
Ist Vice President William M. Goza
P. O. Box 246, Clearwater, Fla. 33515
2nd Vice President George Magruder
440 Tenth Ave. Indialantic, Fla. 32901
Secretary-Treasurer Sara B. Benson
3400 East Grant Ave. Orlando, Fla. 32806
Editor-Resident Agent Ripley P. Bullen
Florida State Museum, University of Florida
Gainesville, Florida 32601
Three years: Wilma B. Williams
Two years: Thomas Gouchnour
One year: Cliff E. Mattox
Cocoa Beach, Florida
At large, for one year
James Varner, Winter Park
J. Anthony Paredes, Tallahassee
. . . . . . 173
THE PRIMITIVE SOLAR OBSERVATORY AT
CRYSTAL RIVER AND ITS IMPLICATIONS
Clark Hardman, Jr.
Crystal River is a six miles long spring-fed tidal river located in
Citrus County on the west coast of Florida about 75 miles north of Tampa
Bay. The entire area is low and marshy with many slightly higher islands.
The town of Crystal River clusters around the springs that feed the river.
The major Crystal River site is located about four miles inland on the
north bank of the river. Other prehistoric sites dot the area. Shell mounds
are found on the south bank about a mile downstream and a large shell de-
posit is at the mouth of the river. However, the Crystal River site (Fig. 1)
is the only aboriginal ceremonial center in the region.
The earliest excavations at the Crystal River site were made by Clar-
ence B. Moore (1903, 1907, 1918) who investigated the main burial complex.
Hale G. Smith (1951) dug a stratigraphic test a little to the north of the South
Temple Mound. Further stratigraphic tests were made in the same general
area by Ripley P. Bullen in 1951 (Bullen 1953). Bullen excavated at several
locations in 1960 and 1964 (Bullen 1966) and directed the restoration of the
central mound and platform of the burial complex. His work indicates that
construction of the large burial mound started about 30 B.C. (Sample 1-1916,
1,980 100 years B.P.) and that the site was abandoned around A.D. 1200.
Since 1962 the site has been a Florida Historical Memorial and State Park.
The major features of the Crystal River site (Fig. 1) include the South
Temple or Spanish Mound, the Center Shell Mound, the North Shell Mound, the
Northwest or Second Burial Mound, the North Temple Mound, the Burial Com-
plex, and the Stelae.
The South Temple Mound (Fig. 2) is an immense shell structure located
on the north bank of the river and plainly visible from the river. According
to Moore' s (1903) original description it was 28 feet 8 inches in height, ob-
long in horizontal section and had a graded way or ramp 80 feet long and from
14 to 21 feet across. A causeway across a low place connected the graded
way or ramp with the South Plaza. Bullen (1966) gives a radiocarbon date
for this mound of A. D. 640 from a charcoal sample obtained 19 feet below the
Florida Anthropologist, vol. 24, no. 4, December 1971
SOLAR CRYSTAL RIVER
STELE 3 C"
4. CEN'.M R
( U 8HEL L'
^ STELE I
. - ai-
S 0 T.H
. PLAZA^ s.
Fig. 1. Map of Crystal River site with feature names used in this paper.
7.. - I' o *. |" -.,. I' ,,
Fig. 2. Relatively intact north end of South Temple Mound.
The Center Shell Mound (Fig. 3) is located on the edge of the marsh
north-northwest of the South Temple Mound (Fig. 1). This is a rather
small mound built on top of midden deposits. On the summit it measures
44 feet in length with a width variation of from 26 feet near the south end
down to 19 feet near the north end. The base length is approximately 75
feet and the height about 10 feet. Traces of a ramp can be detected on the
The North Shell Mound (Fig. 4) is a short distance north and slightly
west of the Center Shell Mound (Fig. 1). It is about the same height as the
Center Shell Mound with a summit 12 feet above the marshy land immediately
to the north. This area is extremely puzzling--the shell slopes up gradually
from the south but drops off abruptly on the north side. The surface of this
south slope is unusually rough and uneven with a large hole directly south of
the North Shell Mound proper. On the west side the shell slopes downward
from the summit of the mound to the level marsh west of the hole. The
summit of this mound is small and its area difficult to determine. The
roughness of the area strongly suggests something was destroyed here.
The Northwest Mound is an oval shaped sand mound about 150 feet long,
100 feet wide and 5 to 6 feet high (Fig. 1). A part of the southwest section
of the oval has been removed.
The North Temple Mound (Fig. 5) is about 12 feet high with a graded
SOLAR CRYSTAL RIVER
way, has a flat top 180 feet long and 30 feet wide. The ramp is about 20
feet off center towards the east. The base of the ramp connects with a
shell causeway presently covered by an asphalt walk, that leads to the
Northwest Mound (Fig. 1).
According to Moore (1903), the Burial Complex (Fig. 6) is "about 115
yards in a northerly direction from the great shell-heap. He further des-
cribes it as "a circular embankment of sand, of irregular height and width,
the maximum, respectively, being 6 feet and 75 feet. Within this circle is
certain territory on the general level and an artificial elevation of sand, ir-
regularly sloping. This elevation culminates in a mound of sand (Fig. 1).
While difficult to determine where the artificial elevation ended and the
mound proper began, to call the diameter of the base of the mound 70 feet,
would be a fairly correct estimate. The height of the mound proper from
the east, where it bordered the level ground, was 10 feet 8 inches. "
The stelae system will be described in later sections. The two stand-
ing stelae--the only two known for eastern United States--are located, as
shown in Figure, 1, to the west and southeast of the Burial Complex.
Except for the formation of plazas, the major features of the Crystal
River site seem to reflect random placement. The purpose of this is to
point out that there is nothing random about this arrangement. Features
are located in alignment with the solstitial and equinoctial positions of the
sun. There are also north-south alignments. This important ceremonial
site can be interpreted as a giant calendar of sand, shell and stone.
Method of Study
The alignments described here are based upon almost daily observa-
tion over a period of about eight months when the author was living near the
site, plus several subsequent periodic observations. Observed positions of
the sun and the features of the site were photographed. Vegetation inter-
fered with this photographic operation and some desired photographs were,
hence, impossible to obtain. Many were difficult and leave something to be
desired as a finished product. However, over 1,000 exposures in black
and white and in color were made.
A lensmatic pocket compass mounted with a front sight on a wide
twelve-inch ruler (a make-shift alidade) was used to determine azimuths.
This instrument was highly portable--it could be slipped into a hip pocket--
and was reasonably accurate when used on a plane table.
The base map (Fig. 1) used was a photographic enlargement of Moore' s
(1903) map. Features not indicated thereon were located by drawing inter-
secting rays from two or more points. Moore' s map was checked and proved
accurate except for a discrepancy in the alignment of the North Temple Mound
^2i* -. ^
Figs. 3-6. Center Shell Mound from east; North Shell Mound from east; west end
of North Temple Mound with ramp on right; main Burial Mound and Platform.
SOLAR CRYSTAL RIVER
and the placement of the North Temple Mound ramp perpendicular to the mound
and centered. The closer approximation for alignment is a 290-110 degree
line. This is a clockwise rotation of about five degrees from Moore's orien-
tation. The reconstructed burial mound summit seems accurately placed.
Fig. 7. Apparent path of sun at equinox and solstices for 23.5 degrees north
latitude, from Strahler (1960), courtesy Johns Willey and Sons.
The sun, due to a 23 1/2 degree inclination of the earth's axis away
from a perpendicular to the plane of the ecliptic, apparently moves north and
south (Fig. 7) in the sky. This apparent movement is greatest along the
horizon varying from a total of 47 degrees at the equator to 153 degrees along
the horizon at 65 degrees north. Changes in the altitude of the sun at noon
are readily observable at different times of the year. However, it is along
the horizon that the apparent movement is most noticeable. Almost any hori-
zon will have a landmark or reference point. The trouble is that this land-
mark is not likely to be in the right spot to line up with a solstitial or equi-
noctial position of the sun.
Once it has been determined that the sun moves along the horizon from
Point A to Point B and then back to Point A, and something significant hap-
pens, say a change in the weather, it becomes important to keep track of the
sun' s movement as a basis for prediction.
If the horizon contains no landmark in the right spot, then it is a sim-
ple matter to set up a marker--a rock, a post, or a mound--perhaps on the
horizon or a mountaintop as in the case of the Inca, but a marker only a few
hundred feet away will serve the same purpose and this is particularly true
if the purpose is an approximation for a ceremony at solstice times. These
are very simple observations and associations. The next step is a more
difficult one. If the marker is on a distant horizon, any reasonable general
area will serve as an observation point, but if that marker is only a few hun-
dred feet away, the observation point becomes critical. To be accurate,
the observer must stand at one spot.
Fig. 8. Fixed observation point and
lines of sight over markers. When sun
is behind a fixed point, it marks a par-
The essentials of a primitive solar observatory then becomes a fixed
observation point and a marker to sight over (Fig. 8). Once the solstice
SOLAR CRYSTAL RIVER
Fig. 9. Essentials of a solstitial observation system with central observa-
tion point and a marker for each solstice point.
points are determined, the cycle from one solstice point to the other and
back to the first will establish the year. Any point between the solstices
can also be used to establish the year. If an in-between point is used, the
cycle becomes a complete circuit from the starting point through both sols-
tices and back to the starting point.
If the horizon is level, like a plain or ocean area, the angles of sunrise
and sunset are for practical purposes the same angular distances north or
south of east and west and starting from a sunrise summer solstice, the sun-
set at winter solstice is 180 degrees away or a straight line. A line-of-sight
can be established in either direction. From a particular point, the azimuths
of the sunrise and sunset system at solstice time becomes some variant of
that illustrated in Figure 9.
Fig. 10. Minimum essentials of a solstitial observation system when
observation points and marlers interchange (M, markers; OP, obser-
Again, with level terrain, it is necessary to have only three points es-
tablished (Fig. 10). This is the probable stela system at Crystal River.
The markers and observation points interchange depending upon the solstice
point under observation. This develops into a problem--was the whole pos-
sible system used or was only one set of solstice points important?
The solstice points are very poor points for observation if accuracy is
a consideration. It is doubtful that a day-to-day difference in azimuth can
be detected with a relatively short line or sight and no instruments. The
movement of the sun in the week before and the week after the solstice is
only about twenty minutes of an arc on the horizon at Crystal River' s lati-
tude. Approximations or accuracy within a period of a few days is no
problem (Figs. 11-13).
Figs. 11-13. Paths of setting sun near summer soltice, June 1, June 8, and
June 18 respectively. Main bush in lower right hand corner is marker.
Accuracy is easily obtainable with a line-of-sight system if the obser-
vation system is at or near the equinox. At Crystal River's latitude, the
sun is moving along the horizon 12-13 degrees per month (Figs. 14-15) near
the time of the equinoxes. Detecting day-to-day differences in the angles of
SOLAR CRYSTAL RIVER
Figs. 14-15. Movement of the sun near the equinox, March 9 and 13 respec-
tively. High pole is marker.
sunrise or sunset requires only the observation point and marker carefully
used. This procedure has a potential accuracy of plus or minus a day for
fixing a point in time consistently. The two main methods for establishing
the year or calendar based upon the sun available to primitive peoples are
(1) the line-of-sight to the horizon discussed above and (2) the use of the
gnomen or shadow stick. Both of these methods were used at Crystal River.
Considerations and Limitations
This section might be more descriptively called troubles or problems.
1. Primitive instrumentation was not the precise kind we know.
The instruments were limited to lines-of-sight and markers such as posts,
rocks, or mounds and the observation of shadows. There was no knowledge
of degrees of an arc or the equivalent that we know of. If carefully used,
line-of-sight and shadow systems could have been very accurate. If they
were carelessly used, they would still have a practical accuracy.
2. We do not know what instant was considered sunrise or sunset by
the Indians. Was sunrise the appearance of the upper edge of the sun and sun-
set the instant it disappeared in the evening? Or was the lower edge the re-
ference point? The truly nasty thought is, was the reference point some
fixed distance above the horizon--a diameter of the sun--four fingers at arms
length or some other standard? Fog and cloud banks are very uncooperative
at sunrise and sunset.
3. Vegetation creates a problem now in observing at sites such as
Crystal River. In early times, there may have been virgin timber. We do
not know the height of the trees. We do not know the size of the clearing.
Were lanes cleared along the line-of-sight? Tall, dense vegetation close
to the markers could make it rather hopelessly difficult to determine now that
an alignment was once used for determining a position of the sun. The error
due to vegetation or terrain will always be toward a more southerly direction.
4. We cannot depend upon present vegetation patterns as guides. The
mean sea level has varied and was about eight feet lower in 200-300 A. D. than
it is now (Lazarus 1965). This would cause changes in the vegetation patterns,
for example, a few feet of change in mean sea level (lower) would allow trees
to grow in the marsh on the down river or west side of the Crystal River site.
5. Precession will prevent a perfect alignment of features with the sun.
Precession is the wobble of the earth on its axis--something like a wobbling
top. This wobble amounts to 47 degrees in 13,000 years for stars in the
polar regions. The variation in the position of the sun is much less amount-
ing to 1 degree in 7,000 years (Lockyer, 1894) or in the time span of the
Crystal River site something like 10-15 minutes of an arc.
6. Erosion is another unknown factor. Over the centuries, weather
has produced changes. There has been the occasional freeze in the winter to
help work shell and sand down a slope; the winds of hurricanes and thunder-
storms to uproot trees; and torrential rains to level and redistribute the earth-
works. The features will have changed and this becomes important in at-
tempting to determine an accurate orientation.
7. The observation point is critical for the short distances at Crystal
River and this observation point would need to be consistent and marked to ob-
tain what is reasonable accuracy to us. The assumption here is that the ob-
servation point is at a feature. It is quite possible that the observation point
was only near a feature (something like three paces or some other unit away)
or a position marked with a rock, a small bed of shell, dirt of a different color,
etc. How many ways are there to mark a spot?
There is only one highly probable observation point at Crystal River.
The rest assumes a point near a feature or within a small area. It should be
pointed out again that from a practical standpoint or for some ceremonial pur-
SOLAR CRYSTAL RIVER
poses accuracy is relatively unimportant. If the winter solstice is what we
want to pinpoint and our purpose is to awe, generally impress and control the
community with a forecast of things to come, an approximation is sufficient.
The Stele System
Stele 1 (Figs. 16-18) was discovered during clearing operations at the
site in 1964 (Bullen 1966). This stele is an irregular boulder that extends
five feet six inches above the present surface level. The northeastern side
of the stele (Fig. 16) has a pecked and incised head and shoulders with hair
or plume, right ear, earspool, eyes, and mouth. (Lines below the shoulders
have been added recently.) It is this side of the stele that faces sunrise dur-
ing summer solstice.
This orientation was determined by butting the square end of a compass
against a piece of plywood placed across the incised figure just below the
shoulders. The reading was 65 degrees. The sunrise compass azimuth for
the summer solstice was 63 degrees and the computed azimuth is approximately
62 degrees. Irregularities on the incised surface will cause orientation figures
to vary 2 or 3 degrees depending upon the placement of the piece of board. The
orientation of the incised face is essentially and practically a perfect facing of
sunrise at summer solstice. "The Little Sun God" label this writer acci-
dentally attached to this incised figure seems decidedly appropriate.
Bullen (1966), in his report of Stele 1 and the excavations around its
base, describes a curious "pathway of cobblestones" (Fig. 19) a little over
two feet wide and nearly four feet long on the northeastern or incised face
side of the stele. The pathway ends under Rock D. Rock D is a flat-topped
rock about 16 inches across and set 15 inches in the ground. Food remains,
charcoal and sand covered the cobblestones. Mottled brown sand and yellow
gray sand covered the food remains and charcoal level. Charcoal from this
cobblestone area (Bullen 1966) was dated at about 440 A. D.
If an observer stands on the cobblestone area facing "The Little Sun
God", he will be in a position to see the sunset over the right side of the stele
(the incised figures left side). This suggests a sunset ceremony and observa-
tion of the sunset. As the sun moves north from the winter solstice sunset,
it will pass behind the Center Shell Mound around the first of March. If at
this time the priest-astronomer stood on Rock D (Fig. 19) of the cobblestone
area (Bullen' s suggestion) and looked west or sighted over the apex of Stele 1,
he would be able to align the sun, some aspect of the Center Shell Mound such
as an end, and the top of Stele 1 (Figs. 19, line d-f; 20). The purpose for
this off solstice-off equinox alignment would be to establish a reference point
in time. This use of Stele 1 accounts for the features of the Stele 1 complex.
The lower northwest shoulder is to give a clear view of the setting sup. The
pointed apex (Fig. 16) is pointed to serve as a rear sight. The height of five
feet six inches is about the right height for an observer standing on Rock D.
Figs. 16-18. Front, side, and detail view of Stele 1.
SOLAR CRYSTAL RIVER
Rock D and the cobblestone area constitute the ceremonial and observation
With a front sight, a rear sight, a fixed point to stand upon and a 600-
800 foot line between the front and rear sights there isn't much need for
cross-hairs, small tubes and verniers. This system could have been re-
markably accurate and could easily have determined the length of the year
within a day or two. If clumsily used, for example, the sun setting behind
the mound marks a festival or ceremony and/or the highly unlikely idea that
there was no concept of a year then the accuracy is within about a week--
still a practical determination in a primitive society. The probable use of
Rock D and the peaked shape of the top of Stele 1, however, suggests accuracy.
Why not an equinoctial alignment? The equinox may also have been
used but this Stele 1--Center Shell Mound line is the longest line in the site
and therefore has the capability of being the most accurate. There is no
particular magic about the equinox as far as ease of observation and poten-
tial accuracy is concerned. Any marker position between about 260 and 280
degrees from the observation point and between the last of February and the
last of April (August and October in the Fall) in time will be essentially as
good as in any other. This is the period of most rapid apparent movement
of the sun and a rather easy observation. The apparent movement is about
25 minutes of an arc on the horizon per day near the equinoxes.
There is another alignment based upon a Stele 1 observation point
used to establish the summer solstice. If, at sunset summer solstice, the
observer standing on Rock D as described above does about a half right-face
he will be looking at the sun hanging in the tops of the trees over Stele 2
(Fig. 21). The camera position for this photograph was on an approximate
Stele 1-Stele 2 line.
Vegetation now prevents a direct line-of-sight from Stele 1 to Stele 2.
The ring around the Burial Complex would have also blocked Stele 2 from an
observer at Stele 1. However, this ring may have been constructed some
hundreds of years after the stele and the central part of the Burial Complex.
It is reasonable to assume that vegetation would have existed to the west in
primitive times and the primitive observer would also have observed the sun
setting in the tree tops. The height and the distance to the trees are the un-
Stele 2 (Fig. 22-25) is a weather-beaten slab 28 inches wide near the
top, 52 inches wide at the original surface level, extending 41 inches above
and 18 inches below the surface (Bullen 1966). Detailed measurements are
given in Figure 25. The front or western end is slightly concave and is not
in a vertical position. The top is not level. There are two sloping sections.
A long concave section descends from the slanting top to nine inches above the
S Edges of *
I xcz v tione
- __ r
k I Nortl st
Fig. 19. Diagram of excavation at Stele 1 show-
ing cobblestone area (Bullen 1966). (A, A' Stele
1; C, E, wedge rocks; D-F, line-of-sight to the
Center Shell Mound.)
Fig. 20. The Crystal River stele system for
determining solar positions. (E, equinox; S.S.,
summer solstice; W.S. winter solstice.)
SOLAR CRYSTAL RIVER
S* :. ",2.', '- .4 V ,.'-
Fig. 21. Stele 2 at summer solstice looking from direction of Stele 1.
Figs. 22-24. Southern side, east end, and west end of Stele 2.
original ground level. This concave section is rounded and thinned. Most of
the material removed in thinning was from the northwestern side. The stele
leans 19 degrees in a northwesterly direction. At the present, an asphalt walk
and support surrounds the stele leaving about 32 inches of the upper part ex-
posed. There is about nine inches of asphalt and fill around the base.
Fig. 25. Dimensions of Stele 2: A, 10 degrees from vertical; B, 10 degrees
from horizontal; C, 20 degrees from horizontal; D, section shown in photo-
graph (Fig. 22); E, fill and support; F, original ground level; G, original un-
derground portion; tilt to NWm 19 degrees; thickness 6-13 inches; not to scale.
The first problem with Stele 2 is the back-sight to Stele 1. This is a
winter solstice line-of-sight and over the eastern end of the village and river
with less interference from vegetation a reasonable assumption in primitive
times. Figure 26 illustrates the best estimate of this alignment. This is a
very impressive sunrise. The photographic technique was to establish a
point on a line between the stele, shift position until the sun was observed
glinting through the trees, then make the exposure.
Stele 2 is located directly west of the cone-shaped burial mound of the
burial complex. The Stele 2-Burial Mound line thus becomes an equinoctial
alignment. A spot at or near Stele 2 is the logical observation point with the
burial mound itself as the marker. Figure 27 illustrates this sunrise.
When the sun rises over the center of the Burial Mound, it is the time of the
equinox. No other assumption is necessary to make this a fairly accurate
SOLAR CRYSTAL RIVER
Figs. 26-28. Sunrise winter solstice over Stele 1 from Stele 2; sunrise
spring equinox over burial mound from Stele 2; sunset winter solstice over
North Shell Mound from near Stele 2 (camera position shifted to left be-
cause of tree in line of sight).
There is also another alignment based upon Stele 2--a sunset winter
solstice line that is about perfect. If the observer stands at or near Stele 2
and watches the sun set at the time of the winter solstice, he will observe the
sun setting directly behind the apex of the North Shell Mound (Fig. 28).
If the summit of the Center Shell Mound seems impractically small as
a base when a structure is considered, then the summit of the North Shell
Mound is ridiculous. It is barely there, but it was obviously built to serve
a special purpose or for the erection of something small on the summit, for
example, a stele to mark the position of sunset winter solstice.
Earlier in this discussion, it was pointed out that a system of three
markers or stelae could be used by interchanging observation points and
markers to determine all of the solstice points. The accuracy of the orienta-
tion, the apparently purposeless small size, the elevation, the alignment of
Stele 1 and Stele 2, and the concept of three stelae needed to make a com-
plete system--all point to the tantalizing probability that the North Shell
Mound was built as a base for a stele. There is only one small problem.
No stele occupies the summit of the North Shell Mound.
Stele 3 refers to a large roundish boulder dug up during the excavation
for the foundations of the museum building (conversation with Ripley P. Bullen).
This boulder (Fig. 29) is now located east of the walk on the south side of the
museum building. Like Stelae 1 and 2, it is made of limestone rock, but the
material is more cherty and appears harder. Stele 3 is 30 inches high and
30 inches in diameter near the base. It is well-rounded and near the base
on the front side are incised or pecked and incised figures.
Starting at the left (Fig. 30) is an irregular or wiggly line suggesting
a snake and ending at the base with a hole. Immediately to the right of this
line, and perhaps part of the same figure is the "print" of a cupped palm
possibly holding something or with something tied around the wrist (Figs. 31,
32). The thumb is large and parallel to the ground. Four indentations re-
present the fingertips. Centrally located is a vertically placed oval-like
figure with a large hole in the base. To the right is the final figure--
another oval--smaller than the central one with, possibly, two eyes. The
markings remind us of the following incidents in the wanderings of Quetzal-
coatl after he left Tollan (Tula) taken from Father Sahagun (Burland 1967):
"Quetzalcoatl beat his hands in sorrow on the great rock where he sat, and
the shape of the palms appeared deep and clearly marked in the rock as if
his hands had been pushed into it. This quotation is an Aztec legend, but
the legends of Quetzalcoatl and Kulkulcan are apparently quite old in Meso-
Note two items: (1) Stele 3 was not visible on the surface. It came
from a low area, one which sometimes contained water, and may have been
buried underground. (2) In this area, that of the Museum Building, there
is no reasonable extension of a solar alignment that it could mark nor can
we suggest any other reason that would indicate it was originally erected in
this area. How did it get where it was found? It was, possibly, a hold
object buried or hidden during a time of upheaval and destruction.
This rock is probably the needed Stele 3 from the top of the North
Shell Mound. In its original location, the incised figures would have faced
the sunset winter solstice.
Other Astronomical Attributes
The South Temple Mound
This wounded giant is still impressive. A primary original purpose
would have been to awe and impress the uninitiated. Sheer size alone (over
330,000 cubic feet of shell) is enough to accomplish this purpose. However,
:- I ,
P* ~iC bS
r Y.- Y~I
B 35 ~
'^ .)*. ' -
Figs. 29-32. Stele 3; details of symbols on face; hand on clay to form imprint;
imprint like symbol on left part of face of Stele 3.
the primitive architect had still another trick up his sleeve. The entire lay-
out of South Plaza-South Temple Mound is such that the original ramp des-
cended toward the plaza with a well-defined causeway (Bullen 1953) con-
necting the base of the ramp and the plaza. From this South Plaza at the
winter solstice, the sun sets (Fig. 34) behind the top of the ramp or the
center of the South Temple Mound. The sun will set some place behind the
mound for a period of two to three months in winter, depending upon the ob-
servation point, when viewed from the village area.
When watching this sunset, it requires very little imagination to almost
hear the chants fading across the marshes or to see figures in elaborate
costumes silhouetted against a sinking solisticial sun.
The Crystal River culture was a highly conservative one that lasted for
hundreds of years. Its sphere of direct "political" influence extended south
to Tampa Bay and north to Horseshoe Beach based upon the distribution of
Their trademark--limestone-tempered pottery. The durability and areal
spread seems based upon two factors: (1) The ability of the priests to awe,
impress, and control the population, and (2) the fact they were not dealing
entirely with myth.
The various alignments at this site make it reasonable to assume that
sun worship in some form was a dominate element in their ceremonial
practices. These alignments also imply an element of accuracy and obser-
vation of the positions of the sun. Such sites as Crystal River are largely
primitive solar observatories. The priests had an accurate basis for advice
and prediction and therefore performed a service for the population.
The Mound System
The North Temple Mound is a very unusual structure, very long and
narrow, any building on its summit could be expected to range from small to
unbelievable. The North Temple Mound is not convincing as primarily a
platform for a structure or structures. In flat country, such as the Crystal
River area, any increase in elevation gets an observer above some vegetation
and increases his ability to observe something on the horizon--in this case,
the sun. As a platform for observing the sun, the North Temple Mound be-
comes believable. The orientation is approximately a 290-110 degree line.
Sunset summer solstice-sunrise winter solstice is a 297 1/2-117 1/2 degree
line with no allowances for vegetation interfering with the line-of-sight.
The North Temple Mound is about 30 feet wide and nearly 200 feet long.
There is room enough on this summit for ten degrees of maneuvering with
diagonal lines. Figures 33 and 35 are photographs of winter solstice sun-
rise and summer solstice sunset from positions on the North Temple Mound.
The winter solstice sunrise is at a slight diagonal across the mound (Fig. 36).
In the summer solstice photograph, the sun, as at Stele 2, is in the tops of the
SOLAR CRYSTAL RIVER
trees and seems to line up with the edge of the mound. The camera location
for this photograph was on the north edge of the mound.
There is no available evidence for markers or observation points on the
North Temple Mound. There is only the alignment of the mound. Perishable
materials such as posts could have been used for markers or with a slight de-
crease in accuracy a procedure such as a line-of-sight down the center of the
mound would establish the summer solstice position. An observation point
near the northwest corner with the southeast corner as the marker could give
a very close approximation of the winter solstice. Sighting through or from
small structures would also be a possibility.
Another line-of-sight is necessary to complete the mound solstitial
system. The probable use of the ends of the North Temple Mound provides
the clue. If an observer stands on the eastern end of the North Temple Mound
(Fig. 36) at sunset winter solstice, he will observe the sun setting over a
point near the center of the Northwest Mound. The back-sight for this align-
ment--observation point on the Northwest Mound with the marker the eastern
end of the North Temple Mound--becomes convincing when the observer stands
at this approximate point on the Northwest Mound and realizes that the point
could be adjusted with about as much accuracy as desired.
Further, a north-south line passing through Stele 2 and the top of the
ramp of the South Temple Mound intersects the solstice line at the observa-
tion-marker point on the Northwest Mound. In addition, there is a mean-
ingless alignment, as far as solar relationships are concerned, of Stele 1,
the summit of the Burial Complex mound and this observation-marker point
on the Northwest Mound. This is obviously an important point. All that is
needed is one lonesome post hole or a small bed of oyster shell. The chances
of finding either look rather dim. During foresting operations in 1958, a
bulldozer cut a wide trench, presently refilled, through this mound. Much of
the rest was excavated in 1960 and 1964 (Bullen personal communication).
Safety Harbor period pottery and a group of flexed burials were found.
There is another relationship between the North Temple Mound and
other features of this site. The eastern end of the North Temple Mound is
located due north of Stele 1; the western end is located due north of the sum-
mit of the Burial Complex mound. These linear and solar relationships
strongly suggest that the North Temple Mound and the Northwest Mound were
constructed by people of the same culture that constructed the southern fea-
tures of this site, but that the dates of construction were separated by an un-
known period of time.
Stele 2 Revisited
Stele 2 is the central feature of this complex. It is also the pivot point
or central point of a system for observing the solstitial positions of the sun.
Figs. 33-36. Sunrise winter solstice North Temple Mound; sunset
winter solstice over South Temple Mound from South Plaza area; sun-
set summer solstice along line of North Temple Mound; Crystal River
observatory system for solstitial points.
Is there some other use or some missing observation? The centuries have
not been kind to this rock slab. The surfaces are pitted from weathering.
If there were ever any figures or symbols on any of the surfaces, they cannot
be detected now.
158 SOLAR CRYSTAL RIVER
The stele is set on end--almost certainly--it was in a vertical position
when first erected. Could Stele 2 be a gnomon? A shaft or a squared slab
could be used to study a shadow on the ground. The top of Stele 2 slopes
slightly downward. The eastern end slopes downward in a concave curve to
nine inches above ground level. (Figs. 22-25). Could Stele 2 be so constructed
that it does not cast a shadow at the summer solstice? It does NOT and here
is a bonus. The shadow creeps down the concave section and reaches the
bottom at the time of the equinox. The shadow at this point will be an ac-
curate time marker. Stele 2 is a calendar stone!
In order to reach these conclusions and to demonstrate the results, a
way of studying the shadows cast had to be devised. Observing the stele dir-
ectly would not work. In the spring, summer, and autumn months, Stele 2
is in dense shade except briefly in the late afternoon. Preliminary infor-
mation that these ideas were rather accurate was obtained using a rough
scale model and artificial light. Photographic troubles developed with this
set-up. The shadow edge was not sharp enough. Another completely ad-
justable and more precise instrument (Fig. 37) was constructed to use the
sun as the source of light.
The model of Stele 2 was carefully constructed to scale. A photogra-
phic enlargement of the correct length was made of the section of the stele
that shows above the asphalt now. The photograph of the stele was cut out
and glued to a piece of board. The correct angle of slope for the critical
back slope at the top of the concave section was maintained in mounting the
photograph. The stele was then sawed around and sanded into the finished
shape. The portion of the model not covered by the photograph represents
the fill and support added above the original ground level. The base is slightly
less than one-sixteenth of an inch longer than accurate. A vertical line near
the end at the base of the concave section indicates the end of the stele ac-
cording to available measurements (Fig. 38).
The shadow cast by an object (a model) at any time of the year can be
set on the instrument and photographed or the unit can be set up and the sha-
dows cast observed at any particular time of the year and/or orientation,
for example, the equinoxes or solstices. The circular azimuth scale at the
base is used only when a particular orientation is desired. In use, simply
to set a shadow, the entire unit is rotated around the vertical axis until the
shadow falls directly behind the model. There are two horizontal axes. On
the bottom axis, the unit is tilted until the rectangular model does not cast a
shadow. This is the position of the sun when it is directly overhead. Next,
the altitude of the sun above the horizon at the desired time and orientation
is set on the top axis and protractor. The result is a shadow on the model at
the desired time of the year and orientation.
If the high end of Stele 2 faced 180 degrees, data would not be much of
a problem. The perverse slab does not face south. The orientation ob-
Figs. 37-38. Equipment used in determining shadow positions on Stele 2,
rectangle to left is for positioning the instrument and comparison; duplicate
shadow conditions on Stele 2 at summer solstice with no shadow on concave
section, altitude of sun 83 degrees above horizon, azimuth 225 degrees.
SOLAR CRYSTAL RIVER
trained at the level of the surface of the asphalt is 225 degrees. This 225
degrees is the exact MIDPOINT between due south and due west and for this
reason strongly suggests that this orientation is the original orientation. The
stele does not have to face south, as we first thought. The only consideration
favoring a north-south orientation is the shadow will be shortest at noon at the
summer solstice. Except for the solstices, the end of any shadow cast by an
object will be at any one precise spot only twice during the year. This is the
principle upon which Stele 2 is based.
It is highly improbable that Stele 2 was erected in the present leaning
position. Reasoning from the radiocarbon date of Stele I and that the stelae
were part of a system for observing the sun, it can be assumed that the erec-
tion date for Stele 2 is about the same as Stele 1 or around 440 A. D. Around
4Q0 A. D. the mean sea level on the Florida Gulf coast (Lazarus 1965) was
some five feet lower than at present. Since these times, the mean sea level
has risen and today abnormally high tides can (and have) put water around
Stele 2. There are plenty of reasons for Stele 2 to lean. It is amazing that
this stele set only eighteen inches in the ground is still standing.
If any degree of accuracy is to be attributed to Stele 2, a plausible me-
thod of erection and/or calibration must be determined. The erection of the
complete stele precisely orientated and with a precise long axis tilt would be
a rather formidable task. With the model illustrated below, it is assumed
there has been no significant up or down movement on the long axis.
However, if the stele were erected and shaped or partly shaped in place,
the only calibration problem would be to determine by some other means,
(for example, line-of-sight), a desired point (for example, the equinox), and
then work the slab so the shadow fell at the desired point. Bullen (1966)
thinks Stele 2 was shaped in place, "the limestone flakes suggest that Stele 2
may have been shaped or trimmed on location. Stele Z could have been
simply and easily calibrated in position.
At the stage in this work when it was determined that the movement of
a shadow on the concave section of Stele 2 had an approximation of the range
from summer solstice to the equinoxes all orientations possible had to be
considered. The data needed to set a shadow on the model is the altitude of
the sun above the horizon for the solstices and equinoxes and the possible
azimuths. This data was supplied by the Nautical Almanac Office, U. S.
Naval Observatory for azimuths from 180 degrees to 240 degrees at ten de-
gree intervals. Stele 2 is "read" when the shadow falls directly behind it.
It is possible that marks or paint could have been used on the concave section.
However, no marks are necessary. The fiducial point most likely was the
joining of the shadow of the top position with the shadow of the nine inch base
upon the ground. Figure 38 illustrates the shadow for the summer solstice
or no shadow. Figure 39 shows the shadow about half way down the concave
portion of the stele. Figure 40 indicates the shadow conditions at the time
Figs. 39-41. Duplication of shadow conditions on Stele 2 for early August,
for the equinoxes (altitude of sun 52 degrees for azimuth of 225 degrees),
for near winter solstice (long shadow).
SOLAR CRYSTAL RIVER
of the equinoxes and, as a matter of interest, the long shadows of the ap-
proaching winter solstice are shown in Figure 41. On the model in Figure 40,
the shadow does not reach the bottom of the concave section at the equinox.
This gap amounts to about one inch on the actual stele. The exact shape of
the base of the concave section is covered by the asphalt walk. If a very
slight rounding at this point exists, this discrepancy is corrected to perfect.
Other corners on the stele are rounded and a slight rounding of this corner is
the most likely condition.
What accuracy can be expected with Stele Z? At the summer solstice,
the accuracy on the model is not impressive. One degree of variation in de-
clination produces only a just noticeable difference in the lighting and shadow
on the top section. At the time of the equinoxes, the declination of the sun
is changing rapidly and a half degree of change in declination is easily obser-
ved in the position of the shadow on the base of the concave section of the
Stele 2 model. The change in the position of the end of the shadow would be
easier to observe on the stele itself because the stele is larger. The change
in declination of the sun between March 20 and March 21, 1967 (1967 Ephem-
eris) was 23 minutes of an arc. This easily observable nearly one-half de-
gree change in altitude amounts to a potential accuracy for determining the
equinox of plus or minus a day--not bad at all for a rock slab instrument.
A slab set in the ground and shaped so that it casts no shadow is an
unusual idea. This would also be a rather difficult idea for primitive people
to come up with, that is, unless the person or persons originating the idea had
lived in an area south of 23 1/2 degrees north latitude where the sun is di-
rectly overhead twice a year and at these times a vertically erected object
does not cast a shadow. No part of Florida, including the Keys, is south of
the 23 1/2 degree north latitude. The Tropic of Cancer runs just north of
the northernmost section of Cuba and a little over a hundred miles north of
the northern tip of Yucatan. The shape of Stele 2 is a good piece of evi-
dence suggesting trans-Gulf contact with Yucatan.
The currents in the Gulf of Mexico have a direct bearing on this trans-
Gulf problem. The usual idealized illustration of these currents shows a
current hugging the coast and encircling the entire Gulf. This is not correct.
The currents of the Caribbean funnel into the Yucatan Channel (Fig. 42) and
at the northern end spread into the Gulf like water from the head of a fountain.
The strongest current goes directly across the Gulf and splits off the mouth
of the Mississippi River with the stronger current (Collier 1964) going east.
The westward flowing portion of this current extends well down the Texas
coast and the currents in the west Gulf in general are more variable.
If there were accidental trans-Gulf contact, the most likely place to look
for evidence would be north of Tampa Bay in Florida. The next most likely
area would be the Louisiana coast west of the Mississippi Delta.
Fig. 42. A composite of the currents of the Gulf of Mexico for all
months of the year. Modified from Collier (1964).
To this point we have been concerned with the details of observing the
sun or keeping track of time at a single site--Crystal River. There are no
other known stele in the Southeastern United States. There is, however, a
high probability that mounds were also used as observation points and mark-
ers at Crystal River. Hence, other temple mound sites in the Southeast
were checked for possible solar relationships. Fantastically enough they
Maps of two large well-preserved sites--Moundville, Alabama and Win-
terville, Mississippi--are used here to illustrate this fact (Figs. 43-44).
These maps are essentially self-explanatory. The diagonal lines are solsti-
tial alignments. On the western side of both sites, there are midpoint mar-
kers between the equinox and solstices. This is a midpoint in space, not in
SOLAR CRYSTAL RIVER
Many other sites show at least a single solstitial line. Single solsti-
tial lines exist at Etowah and Ocmulgee in Georgia. It is highly probable
that major features have been destroyed at the latter.
In Florida, the later (Ft. Walton period) mound group near Lake Jack-
son, north of Tallahassee, form a probable observation site. A bluff and
the vegetation to the west could account for the slight error in alignment.
Also, in Florida the Big Mound City site contains plausible gigantic "protrac-
tors of sand" embankments that could be accounted for if they were used in
Kolomoki, in the southwest Georgia, is a well-preserved site with some
similarity to Moundville and Winterville. Mound B, a marker point, is des-
cribed by Sears (1956), "This is the most unusual mound it has ever been
the writer' s misfortune to encounter. . It rapidly became obvious that
Mound B consisted of a collection of postholes. Very large posts, 24-30
inches in diameter erected successively in this small area. . Later posts
often cut through the remains of earlier ones, so that in only a few instances
were complete outlines produced. Were the posts periodically changed, or
was there trouble "calibrating" the mound?
In addition to Winterville, other Mississippi sites with solstitial align-
ments include: Jaketown, Menard, Shell Bluff, and Alligator. The Prairie
Jefferson Works, Morehouse Parish, Louisiana, according to a Squier and
Davis map, is an elaborate layout of solstitial and equinoctial alignments.
Based upon Moorehead' s 1928 map and an assumed true north orientation,
most of the eight-plus mounds at Cahokia, Illinois, show solar or north-
It is not possible to work out solstitial alignments with the features of
all temple mound sites. Examples of these sites include: Oliver and Nedlett
Landing in Mississippi, Kincaid in Illinois, and Hiwassee Island in Tennessee.
These sites vary. Some are small. Hiwassee Island is in hilly country where
a landmark might have been used. Kincaid might have been a system for some
particular point like Mound A at Moundville and these are hard to determine.
The Hopewellian earthworks in the north--200 B. C. to 450 A. D (Brennan
1970); maximum florescence 100 B.C. to 200 A. D. (Willey 1966) were ex-
amined. There is a possibility these huge earthworks were solar observa-
tories but it doesn't look promising. It is possible to find some alignments
that are close to solstitial alignments, for example, the axis of the octagon-
circle complex at the Newark Works is close to a winter solstice line when
an allowance is made for terrain. Another octagon-circle complex, the
High Bank Works, is not even close. There are differences in the construc-
tion of the two complexes.
The North Fork Works contain a solstice alignment with no allowance
W S I
Figs. 43-44. Maps of Moundville, Alabama, and Winterville, Mississippi,
sites showing solar alignments. "1/2" is half-way point between an equinox
SOLAR CRYSTAL RIVER
made for terrain. It is possible to build a theoretical system for the squares
or circles using the mounds as observation points and postulating markers on
the embankments. This requires at least one piece of evidence--a marker
on an embankment--or it remains unconvincing.
No two Hopewell complexes are alike. Only some features are simi-
lar. They were made up of circles, parallels, squares, etc. or combina-
tions of geometric figures. Geometric figures are common magical con-
structions on a small scale. James B. Griffin as quoted by Brennan (1970)
seems to sum up Hopewell correctly as "the cultural climax of the Middle
Woodland Tradition. It could be that all these elaborate enclosures are
nothing but the shaman' s magic symbols gone a bit wild in size.
Previous publications dealing with temple mounds have reached few
conclusions concerning their functions other than that they were built to
support temples for ceremonial purposes. The obvious similarities in plaza
arrangements with those in Mesoamerica have, of course, been noted. This
conclusion can now be extended. The ceremony included sun worship or, at
least, calendrical determinations.
Peter Farb (1968) reports and quotes a description of one of the func-
tions of the Great Sun (chief) of the Natchez from observations by Father
Maturin Le Petit in 1699: "Le Petit was particularly impressed by the high
temple mounds the Natchez had built in order, he said, that the earthly sun
and the heavenly sun could converse: 'Every morning the great chief honors
by his presence the rising of his elder brother (the sun), and salutes him with
many howlings as soon as he appears above the horizon... afterwards raising
his hand above his head and turning from east to the west, he shows him the
direction he must take in his course. This ethonologically observed cere-
mony probably is the remnant of much more complicated sun-worship cere-
monies which occurred a thousand years earlier.
The writer wishes to express his gratitude to those who assisted in
some way with this paper. Ripley P. Bullen is responsible for this paper
at this time. Without his diplomatic prodding, this material would still be
a maze of notes. Doris Whitesell, secretary, Learning Resources Center
at Lake City Community College, gave much assistance in obtaining books
and papers used for references. Frank Zecher of the Forestry Department
at Lake City Community College assisted with the map of Moundville.
My wife, Marjorie, served as helper in much of the picture-taking and
observing needed for this paper. She also did the first rough draft typing of
it. Ramona M. Rondez, student assistant at Lake City Community College,
did the final deciphering of corrections and typing. Others to whom I am
grateful for assistance are the personnel of Winterville Mounds State Park;
David L. DeJarnette, curator of Mound State Monument, Moundville, Ala-
bama, who supplied an accurate survey of the Moundville site; and the Nau-
tical Almanac Office, U. S. Naval Observatory, who prepared the data used
with the model of Stele 2.
Bullen, Ripley P.
1953 The Famous Crystal River Site. The Florida Anthropologist,
Vol. 6, no. 1, pp. 9-37.
1966 Stela at the Crystal River Site, Florida. American Antiquity,
Vol. 31, no. 6, pp. 861-865. Salt Lake City.
Burland, Cottie A.
1967 The Gods of Mexico. G.P. Putnam. New York.
1964 The American Mediterranean. In: West, Robert C. (ed.),
Handbook of Middle American Indians, Vol. 1, Natural En-
vironment and Early Cultures. University of Texas Press.
1968 Man' s Rise to Civilization as Shown by the Indians of North
America from Primeval Times to the Coming of the Industrial
State. E. P. Dutton. New York.
Lazarus, William C.
1965 Effects of Land Subsidence and Sea Level Changes on Elevation
of Archaeological Sites on the Florida Gulf Coast. The Florida
Anthropologist, Vol. 18, No. 1, pp. 49-58. Gainesville.
Lockyer, J. Norman
1894 The Dawn of Astronomy. The Massachusetts Institute of
Technology Press. Cambridge.
SOLAR CRYSTAL RIVER
Moore, Clarence B.
1903 Certain Aboriginal Mounds of the Florida Central West Coast.
Journal of the Academy of Natural Sciences of Philadelphia,
Vol. 12, Pt. 2, pp. 362-438. Philadelphia.
1907 Crystal River Revisited. Journal of the Academy of Natural
Sciences of Philadelphia, Vol. 13, Pt. 3, pp. 406-425.
1918 The Northwestern Florida Coast Revisited. Journal of the
Academy of Natural Sciences of Philadelphia. Vol. 16, Pt. 4,
pp. 514-518. Philadelphia.
Phillips, Philip, J.A. Ford, and J.B. Griffin
1951 Archaeological Survey in The Lower Mississippi Alluvial Valley,
1940-47. Peabody Museum Papers, Vol. 25, Harvard Univ-
Sears, William H.
1965 Excavations at Kolomoki Final Report. The University of
Georgia Press. Athens.
Smith, Hale G.
1951 Crystal River, Revisited, Revisited, Revisited.
American Antiquity, Vol. 17, No. 2, pp. 143-144.
Salt Lake City.
1960 Physical Geography. John Wiley and Sons. New York.
Willey, Gordon R.
1950 Crystal River, Florida: A 1949 Visit. The Florida Anthro-
pologist, Vol. 2, Nos. 3-4, pp. 41-46. Gainesville.
1966 An Introduction to American Archaeology, Volume 1,
North and Middle America. Prentice-Hall. Englewood Cliffs,
Lake City Community College
Lake City, Florida
June 2, 1971
EXCAVATION OF A DEPTFORD MIDDEN BURIAL, DESTIN, FLORIDA
Jennings W. Bunn, Jr.
The OK-19 site is located approximately two miles east of Destin,
Florida on the south side of the Choctawhatchee Bay. The burial area is
two tenths of a mile east of Holly Creek, and was actually only part of the
OK-19 site which is some seventy-eight hundred feet in length. A Fort
Walton period cemetery is also located approximately one mile west of the
The burial site was found accidentally by Mr. Allan Svara of the Ben-
dix Corporation and the author while surface collecting along the banks of
Choctawhatchee Bay. The area is quite popular with collectors due to the
amount of erosion of the site. The banks in that location are approximately
twelve feet high, and eroding at the rate of three to five feet each year. A
large number of bones from the burials had washed down the bank, and the
left side of a cranium was plainly visible protruding from it. We did not
attempt removal of the burials as it was realized that they were of signifi-
cant archaeological value, but instead immediately notified Mrs. William C.
Lazarus, Curator of the Ft. Walton Temple Mound Museum.
The land on which the burials were located is owned by Mr. Coleman
L. Kelly of Destin who, after seeing the location and deterioration of the
burials, kindly consented to our removing them. Mr. Kelly has requested
to retain ownership of all material, but agreed to place it on loan to the
Temple Mound Museum until such time as Destin has a museum. I would
like to thank Mr. Kelly again for his kindness and consideration in this matter.
The most pressing reason for excavating these burials was obviously
to save them from being totally destroyed by nature, but also aside from this
is the fact that few midden burials have been found and recovered within a
controlled excavation in this area of Florida.
We began excavation on the morning of the sixth of October, and by
evening had excavated down to the level of the burials by six inch intervals.
One skull and a few long bones were photographed and removed, but due to
lateness and the disarranged state of the remaining bones it was decided to
wait until the following day to continue. Mrs. Lazarus was unable to join
me next morning, so M/SGT James Macrander of Eglin AFB agreed to lend
a helping hand, and the excavation was completed that evening.
The burials lay at a depth of twenty-six inches from ground surface at
Florida Anthropologist, vol. 24, no. 4, December 1971
the bottom of a grey sand pit, and directly on top of the orange sterile sand
zone (Fig. 1). The grave shaft was first encountered between the twelve
and thirteen inch level. The cranium which had been observed protruding
from the bank was lying on its left side on top of the long bones and ribs. The
The second cranium, which was very thin and badly deteriorated, was lying
twelve inches east south east of the first. This cranium was lying in an in-
verted position, and a large number of teeth were found inside the cranial
cap. Also a large number of teeth were encountered lying loose in and
around the remaining bones. From the number of teeth recovered it was
assumed that there were three individuals, but it was later learned from
Dr. David S. Phelps of Florida State University, to whom the material was
sent for study, that there were actually six individuals.
Following is a list by levels of the material recovered. The test area
was approximately three feet by five feet, and designed so as to encompass
the burial area, and also to do as little damage to the banks as possible.
Level 1, 0-6 inches. This level was totally comprised of Ft. Walton
period material. A humic layer, approximately one-half inch thick, overlay
two inches of oyster shell. Underneath the oyster shell was four inches of
small white clam shell. This was a typical midden consisting of sherds,
animal and fish bones, shells, and small particles of charcoal. We found
three shell tempered sherds of Pensacola variety, one Lake Jackson Plain
sand-tempered sherd, and two Ft. Walton Incised sand-tempered sherds, a
fragment of dark green glass of wine or rum bottle type, as well as fish,
bird, and deer bones.
Level 2, 6-12 inches. The top two inches of this level were slightly lensed
from the leaching of the material directly above. This complete level was
sterile indicating a long period without human occupation. Sand coloration
Level 3, 12-18 inches. Re-established signs of human occupation although
in no large concentration. Deptford material was initially encountered in
this level as well as grave shaft. Specimens included two Deptford Linear
Check Stamped sherds, one Deptford Simple Stamped sherd, and one plain
sand-tempered sherd and some fish, bird, and deer bones.
Level 4, 18-24 inches. This seemed the same as Level 3. Twelve Dept-
ford period sand-tempered plain sherds were uncovered. All shells en-
countered at this level were clams, possibly indicating a change in the sa-
linity of the water. Also present were both burned and unburned bird,
fish and deer bones, three fragments of milk quartz, and a piece of slate
white sterile sand
Fig. i. Burial pit and overlying strata.
Fig. 2. Female torso found near Deptford burial.
At twenty-four inches deep and eighteen inches north of the burials
and two inches in higher elevation, we found the clay female torso illus-
trated in Figure 2. The torso has a small ground groove around it just
below the bosom. The reason is problematical. This interesting speci-
men may be seen at the Temple Mound Museum in Ft. Walton Beach.
No additional burial goods were recovered, although the pit was ex-
tended an additional twelve inches below the burials.
Ft. Walton Beach
HAFTED FLAKE KNIVES
Ben I. Waller
The purpose of this brief paper is to describe a peculiar chipped stone
tool which has been found with some regularity in the north-central part of Flor-
ida, particularly in and around the vicinity of the Santa Fe River. I shall refer
to these tools as "Hafted Flake Knives" but some have jocularly referred to
them as "Waller knives. They seem to be found in those areas which I have
previously referred to as "kill sites" (Waller 1970) although I know of two spec-
imens that have been found at land sites.
These stone artifacts appear to be made primarily from spalls or flakes
knocked off a fair sized core during an initial stage of chipping. One side is
concave or convex and smooth as it is when knocked off a core. The other side
is chipped all over. These blade-like tools have side notches, presumedly for
hafting, carefully chipped near one end. Sometimes the edges adjacent to the
notches are carefully retouched. No grinding is evident on most specimens
but a few exhibit a suspiciously worn appearance in the notches. This could
have been caused by heavy "wearing" of thongs in these notches.
Most Hafted Flake Knives range from 31.5 to 58.0 mm. in length but
one specimen exceeds 91 mm. Most of them are less than 36 mm. wide while
the average thickness is 5 mm.
Edges are unifacially retouched, sometimes steeply, on one or both
edges. Commonly the left hand edge exhibits more retouch than the other. Ap-
parently, these tools were used with a knife motion by people who were domi-
nantly Iight handed.
Being thin and having a good cutting edge, I would suspect that these
Hafted Flake Knives could be used quite well for removing hides from animals
and for slicing portions of meat from carcasses. The amount of wear does
not seem to suggest very rough usage.
Although I have no knowledge of one of these tools having been found in
situ in a stratigraphic column or in a demonstrably early site, I suspect that
they are probably very early tools relatively. They have been found at sites
yielding primarily Paleo-Indian and early Archaic projectile points and it is
my opinion that a date for them would be in the vicinity of 6000 to 8000 years
Florida Anthropologist, vol. 24, no. 4, December 1971
HAFTED FLAKE KNIVES
That Indians were selective in their choice of spalls for use as
Hafted Flake Knives is apparent from the fact that only those with very
flat surfaces were used. Some exhibit a remarkable similarity to flute
spalls from Paleo-Indian projectile points.
In summary, I believe that the Hafted Flake Knives appear to be
quite old as they are usually found in the vicinity of Paleo-Indian "kill"
sites. They seem to have served a knife-like purpose and are obviously
not projectile points, drills, or the usual type of scraper. Of the more
than twenty specimens observed, all--including some slightly variant ex-
amples--appear to have been made from chert spalls.
Waller, Ben I.
1970 Some occurrences of Paleo-Indian projectile points in Florida
waters. Florida Anthropologist, vol. 23, no. 4, pp. 129-34.
Copies of Volumes I through XIV may be purchased from Johnson
Reprint Corporation, 111 Fifth Avenue, New York, N. Y. 10003 at $10.00
per volume or $3.50 for single and $6.00 for double numbers if available.
Copies of Volumes XV through XXIII may be purchased from the Treasurer
of the Society at $4.00 per volume or $1.00 for single and $2.00 for double
numbers. The Society has no resale discount for book dealers.
The FLORIDA ANTHROPOLOGICAL SOCIETY PUBLICATIONS
(a monograph series), nos. 1-5, may be purchased from Johnson Re-
print Corporation, 111 Fifth Avenue, New York, N. Y. 10003 at $3. 50
per number. Number 6 has never been published.
INFORMATION FOR AUTHORS
THE FLORIDA ANTHROPOLOGIST publishes original papers in all
subfields of anthropology with an emphasis on archaeology. Contributions
from allied disciplines are acceptable when concerned with anthropological
problems. The journal's geographical scope, in order of importance, in-
cludes Florida, Southeastern United States, Caribbean, and related regions.
Manuscripts should be double spaced and typed on one side only of an
8 1/2 by 11 inch sheet leaving at least a 1 inch margin on all sides. Contri-
butors are referred to American Antiquity and the Chicago Manual of Style
for matters of style and reference. Textual references should be similar to
"Smith 1964: 78-80; Jones 1969: Pl. II, b). No footnotes. In the "Refer-
ence Cited" section, at the end of an article, do not abbreviate names of ser-
ies or journals. Books need full title, publisher's name, and place.
All illustrative material-line drawings ( in India ink) and photographs-
should be included in one numbered series of "Figures" suitable for a maxi-
mum reduction to 6 1/2 by 8 1/2 inches. Parts of figures are designated by
lower case letters. Figures must be accompanied by brief descriptive cap-
tions. All tabular material must be included in separately numbered "Tables."
Twenty-five reprints without covers of each article are furnished free
to each author or group of authors. Additional copies, with or without covers,
may be purchased but the editor must be informed of the number desired when
the article is offered for publication.