The Florida anthropologist

Material Information

The Florida anthropologist
Abbreviated Title:
Fla. anthropol.
Florida Anthropological Society
Place of Publication:
Florida Anthropological Society.
Publication Date:
Quarterly[<Mar. 1975- >]
Two no. a year[ FORMER 1948-]
Physical Description:
v. : ill. ; 24 cm.


Subjects / Keywords:
Indians of North America -- Antiquities -- Periodicals -- Florida ( lcsh )
Antiquities -- Periodicals -- Florida ( lcsh )
serial ( sobekcm )
periodical ( marcgt )


Contains papers of the Annual Conference on Historic Site Archeology.
Dates or Sequential Designation:
v. 1- May 1948-

Record Information

Source Institution:
University of Florida
Holding Location:
Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
01569447 ( OCLC )
56028409 ( LCCN )
0015-3893 ( ISSN )


This item has the following downloads:

Full Text


2000 Florida Anthropological Society Inc.

The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
these publications.

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.



W" / p ^--^j

/t a





THE FLORIDA ANTHROPOLOGIST is published quarterly in March,
June, September, and December by the Florida Anthropological
Society, Inc., c/o Room 130, The Florida State Museum, The
University of Florida, Gainesville, FL 32611. Subscription
is by membership in the Society for individuals and institu-
tions interested in the aims of the Society. Annual dues are
$6.00; student members $4.00. Requests for memberships and
general inquiries should be addressed to the Secretary; dues,
subscriptions, changes of address, and orders for back issues
to the Treasurer; manuscripts for publication to the Editor;
and newsletter items to the President. Second class postage
paid at Gainesville, Florida.


President: Raymond Williams
Dept. of Anthropology, Univ.
ofSouth Florida, Tampa, FL 33620

1st Vice President: George W. Percy
Div. of Archives, History and Records
Management, The Capitol,
Tallahassee, FL 32304

2nd Vice President: Paula J. Fields
1160 King Street
Merritt Island, FL 32952

Secretary: Nedra Lexow
1124 Harrison St.,
Hollywood, FL 33020

Treasurer: Norcott Henriquez
1510 Dewey St.
Hollywood, FL 33020

Directors at Large

Three years: Thomas C. Watson
203 Carolyn Avenue, Panama
City, FL 32401

Two years: Robt. C. Johnson
4250 Melrose Avenue,
Jacksonville, FL 32210

One Year: Ray. C. Robinson
1020 4th Street North,
St. Petersburg, FL 33701


Editor: Jerald T. Milanich
Department of Social Sciences
Florida State Museum
Gainesville, FL 32611

Editorial Board:
Kathleen A. Deagan
Dept. of Anthropology
Florida State University

Albert C. Goodyear
Institute of Archeology
and Anthropology
University of South Carolina

Roger T. Grange, Jr.
Dept. of Anthropology
University of South Florida

Editorial Assistant:
Ann S. Cordell
University of Florida

John W. Griffin
Historic Key West Preservation
Board, Key West

George W. Percy
Division of Archives, History
and Records Management
Florida Dept. of State




Contents Page

Editor's Page ....................................... 88
An Analysis of Shark Tooth Tools from the Boca
Weir Site in South Florida,
by John F. Furey .............................. 89
An Anglo Sucking Cure from Rural North Florida,
by Joseph N. Henderson ....................... 103
A Second Spanish Period Log Water Pump, Duval
County, Florida,
by William M. Jones .......................... 110
The Roberts Bay Site, Sarasota, Florida,
by George M. Luer ............................ 121
A Chronology for the Aboriginal Cultures of
Northern St. Simon's Island, Georgia,
by Jerald T. Milanich ........................ 134


Two conferences are scheduled for the fall which may be of
interest to members of the Florida Anthropological Society. The
Southeastern Archaeological Conference Annual Meeting is being held
jointly with the Conference on Historic Site Archaeology at the
College Inn, University of Southwestern Louisiana, Lafayette,
Louisiana. For information on the conference contact Jon L. Gibson,
Box 4-0198, USL Station, Lafayette, LA 70504.

Treasure hunters, archeologists, and environmentalists will
meet at a conference in Key West November 4 and 5 to discuss conflicts
that develop between commercial shipwreck salvors and those who
demand governmental regulations to preserve the wreck sites for
cultural study. The conference is free and open to all interested
individuals. Registration begins at 12:00 noon on Friday, November 4,
at Florida Keys Community College in Key West. Contact Mr. Baxter
Wood, Treasure Salvors, Inc., Key West, Florida, area code 305-294-
3336, for accommodation information.

The conference is funded by the Florida Endowment for the
Humanities and is being co-sponsored by the Committee for the
Conference on Florida Historic Wreck Archaeology and Florida Atlantic
University. R. Duncan Mathewson is organizing the conference.

Your editor is happy to announce the awarding of two grants by
the National Science Foundation to the University of Florida for
archeological research. The first, received by myself, will be used
to study the Weeden Island peoples who inhabited Columbia and Suwannee
counties prior to the occupation of that region by the historically-
known Utina Indians. Excavations will be centered on the McKeithen
site, a mound-village complex in western Columbia County thought to
have been the major civic-ceremonial center for the Weeden Island
culture. Surveys and test excavations will be carried out at other
sites to determine their political and economic relationships to the
McKeithen site.

The initial grant of $41,900 will be used to begin the excava-
tions and surveys and to establish a semi-permanent research station
at the McKeithen site. A second increment of $32,200 will be used to
continue the work the following year.

Morgan R. Crook, a doctoral student in anthropology, has
received $5,440 for work on Sapelo Island, Georgia. The research will
be written up as Mr. Crook's dissertation which will examine community
patterning of Mississippian populations living on the Georgia coast
after A.D. 1200. The project is being supervised by Dr. Lewis Larson
and myself. Other funds have been awarded the project by the State of



John F. Furey

The Boca Weir Site (PB-56), located just north of Boca Raton,
Florida,between the Atlantic Ocean and the intracoastal waterway,
was excavated between 1971 and 1972 under the auspices of the
Department of Anthropology, Florida Atlantic University. A complete
site report is on file at the University (Furey 1972) or available
through University Microfilms. The habitation sequence dates from
approximately A.D. 500 to 1600 and the dating is based upon ceramic

An examination of the many archaeological reports on the
Glades subarea, or all of Florida for that matter, indicates that
large numbers of shark tooth tools have been archaeologically re-
covered from only three sites: Key Marco (Cushing 1896), Fort
Center (Steinen 1971), and Boca Weir (Furey 1971). Small numbers
of teeth have been recovered from many Florida sites with a number of
sites reporting only one or two shark tooth tools. At the Belle
Glade site, Willey (1949) found evidence of their use as woodwork-
ing implements but recovered only a- few teeth.

The sheer number of shark tooth tools recovered makes Boca
Weir an anomally and, since the shark teeth from Key Marco and
Fort Center were not subjected to extensive statistical analysis, it is
hopedthis analysis will provide a foundation for further research
concerning this tool type.

A total of 175 shark teeth were recovered at Boca Weir. The
method of analysis utilized was partially based on the work done by
Steinen (1971). The specimens were divided into six categories on
the basis of species. Seven fragments of specimens were examined
and in all cases were assignable to a species. Each tooth was then
carefully examined under a variable power reflecting binocular
microscope (5x to 30x) with a light source of three intensities,
adjustable focus, height and angle. Each specimen could be
twisted and rotated in relation to any of the variables for a com-
plete examination of all surfaces. Notes on each tooth were taken
with wear patterns, striations, blunting, notching, cracks, chips,
and gum modification being the major observational categories. A
chart of each species was then compiled which enabled the statis-
tical treatment of each category. The charts' categories were:
Tip: worn, not worn; Edge wear: mesial, distal, tip; Striations:
mesial, distal, tip; Tooth cracked: yes, no; Blunting: mesial,
distal, bilateral; Notching: mesial, distal, bilateral; Tip frac-
tured or chipped and gum modified: yes, no (see Fig. 1). This
system was devised to be flexible in recording all forms of wear/
use and to enable combinations of variables to be computed. For

The Florida Anthropologist, vol. 30, no. 3, September 1977


example, a tooth with mesial blunting could have distal notching,
distal blunting or no modification on the distal edge. This system
allows for single and multiple attributes to be recorded in one
operation, and any combination of attributes can be computed from
this. It also allows teeth to be rejected as tools by multiple
criteria in relation to other teeth of the same species. Natural
wear in shark teeth has not been an object of study and,as such,all
teeth are considered artifacts until a number of areas of wear/use
confirm their status as a tool. Patterns of wear/use within and
between categories became quickly apparent through use of this

Description of the Teeth

Tiger Shark Galeocerdo cuvieri (N = 24, where N is sample size)
This category consists of five fragments, five unperforated
and fourteen perforated specimens. A total of 20 (83.3%) specimens
in this sample show wear/use patterns, while the remaining 4 (16.7%)
show no signs of wear/use. On this basis, the 4 unmodified teeth
are viewed as not having had enough use to show their status as
tools or were not used as tools. The latter view is accepted in
this analysis and unmodified teeth are classified as non-tools.
Fourteen (58.3%) of the specimens had edge/tip wear and, of these
5 (35.7%) were worn only on the mesial edge, 4 (28.6%) only on the
tip and 2 (14.3%) on the distal edge only. All other wear is in-
significant as 7.1% in three other combinations of wear.

Striations were examined on 8 (33.3%) of the specimens and
they were confined to the mesial edge in 50.0% of the cases, tip
37.5% and 12.5% on the distal edge only. No other striations were
observed, and no teeth were chipped or cracked.

Hafting was accomplished in 58.3% (14) of the specimens by
perforations as well as a combination of methods. Blunting was
utilized as a method of attachment in 11 (55.0%) of the specimens,
with 4 (36.4%) exhibiting bilateral blunting. Notching occurred
on only 2 (10.0%) of the specimens with none exhibiting bilateral
notching. Hafting was also accomplished by combinations (n = 9)
with 7 (77.8%) having mesial blunting only and 2 (22.2%) with
mesial notching only. Perforated teeth with bilateral blunting is
the preferred method of hafting this species.

Lemon Shark Negaprion brevirostris (N = 74)
This species is the largest of the six, with all specimens
being unperforated. Seventy (94.6%) of the teeth exhibit wear/
use patterns while the remaining 4 (5.4%) are not worn and are
classified as non-tools because of their lack of cultural modifi-
cation. Of the 70 teeth with wear/use patterns, 30 (42.9%) are
worn only on the tip, 17 (24.3%) are worn on the tip and mesial
edge only. Wear on all three edges occurs on 15 (21.4%) of the


Striations are found on 35 (50.0%) teeth in the tool sample
with 33 (94.3%) occurring only on the tip and 2 (5.7%) occurring
on the mesial edge only. The extreme utilization of the tooth tip
in this species is further indicated by 25 (35.7%) of the specimens
being cracked and 39 (55.7%) having chips removed from the tip.

This species, with a complete lack of perforations, was
hafted in many ways. Blunting occurs on 25 (35.7%) of the speci-
mens and of these, 16 (64.0%) were blunted bilaterally. Notching
is found on 31 (44.3%) of the specimens with 20 (64.5%) having bi-
laterial notching. Three teeth (20.0%) fall in each other category
with distal blunting excluded. Bilateral notching or blunting is
the preferred method of hafting. Gum modification is insignifi-
cant,consisting of one tooth (1.4%).

(no common name) Hemipristis serra (N = 55)
This is the second most numerous species of the total sample
and consists of 19 unperforated and 36 perforated teeth. Tip wear
is found on 45 (84.9%) of the sample with 6 (10.9%) not exhibiting
tip wear. Of this number, 2 specimens do not exhibit any other
sign of wear/use and are not considered tools.

Tip and edge wear on the specimens considered to be tools is
found on 45 (84.9%) of the teeth (n = 53). Wear is basically con-
fined to the tip and mesial edge, 23 (52.8%); the tip alone, 8 (15.1%);
and on all three edges, 7 (13.2%). Wear on the tip can be further
seen as 11 (20.8%) of the specimens are cracked and 9 (17.0%) of
the teeth have chips removed from the tip region. Perforated
teeth comprise 88.9% of the fractured/chipped specimens and indicates
that greater pressure is involved with the use of perforated teeth
or that the perforation allows greater force to be applied.

Striations occur on 34 (64.2%) of the teeth with 17 (50.0%)
occurring on the tip only, 12 (35.3%) occurring on the tip and
mesial edge only, 3 (8.8%) occurring on the mesial edge only and
2 (5.9%) in all three areas.

Hafting was accomplished on 36 (57.9%) of the specimens by
perforating the gum just below the tooth enamel. Additionally,
blunting occurred in 20 (37.7%) cases with 10 (50.0%) of these
being bilaterally blunted. Of these, 8 (80.0%) occur on perforated
teeth. Thirty-two (60.4%) of the specimens were notched with 14
(43.8%) of these being bilateral; of these 12 (85.7%) occur on per-
forated teeth. This selective factor in hafting is further seen by
7 (31.8%) having mesial notching only, 5 (22.7%) having distal
notching only and 4 (18.2%) with distal blunting-mesial notching.
The other three possible combinations range downward in value
from 13.6%. Gum modification is represented by 1 (1.9%) tooth.

Sand Shark Odontaspis taurus (N = 6)
All examples of this species are unperforated, and it com-
prises 3.5% of the total sample size. All teeth in this category
exhibit wear with 3 (50.0%) worn on the tip only, 2 (33.3%) on all
three edges and 1 (16.7%) on the tip and mesial edge only. These


are the only areas in which wear is found. Striations are found
only on the tip in 1 (16.7%) of the specimens and on the mesial
edge only in another. One tooth (16.7%) is cracked, while 4
(66.7%) have chips and fractures in the tip region.

Hafting by blunting was found on 3 (50.0%) of the specimens
with all exhibiting bilateral blunting. Notching was found on 1
(16.6%) of the specimens with this being bilaterally notched.
No combinations of blunting and notching were utilized and no gum
modification was found in this category.

White shark Carcharodon carcharias (N = 15)
The specimens in this species consist of 2 fragments, 3 un-
perforated and 10 perforated teeth. The species comprises 9.82%
of the total sample size and 14 (93.3%) of these teeth exhibit
wear. The remaining tooth (6.7%) in this sample does not show any
signs of wear/use and is eliminated as a tool.

Tip and edge wear occur on 14 (93.3%) teeth and of this, 10
(71.4%) are worn on the tip and mesial edge only with 2 (14.3%)
exhibiting wear on the tip only. Wear on all three areas, and the
tip and distal edge each are represented by 2 (14.3%) specimens.
No other types of wear combinations are found.

Striations are found on 10 (71.4%) specimens and of these,
6 (60.0%) have striations confined to the tip only and 4 (40.0%)
have these on the tip and mesial edge only. The effects of tip
wear and striations are evident with 4 (28.6%) being cracked,
chipped or both.

Perforation is the major form of hafting, combined with a
preference for notching. Seven (50.0%) are notched with 5 (71.4%)
of these having bilateral notching. Blunting is found on 5 (35.7%)
teeth in the category with 3 (60.0%) having bilateral blunting.
Mesial blunting-distal notching, comprising 2 teeth is the only
other form of hafting noted. Blunting and notching occur on per-
forated teeth only and no gum modifications were found.

Mako Shark Isurus oxyrhincus (N = 1)
The small sample size does not allow for statistical treat-
ment, as this is a description of one tooth. The tip of the tooth
is broken with cracks in the enamel. Chips are removed from the
lingual and mesial edges of the tip as well as a number of pit
marks on the buccal surface. There is no blunting, notching or
gum modification for hafting and the specimen is not perforated.
The tooth is morphologically unsound for perforation.


All species have high percentages of worn teeth with only a
small number being rejected as tools (Table 1). Most wear occurs

Table 1. Distribution of Wear on Shark Tooth Tools, Boca Weir (PB-56).
Total Sample % Classified % Not Classified
Species Size as Tools as Tools

Tiger Shark
Lemon Shark
Hemipristis serra
Sand Shark
White Shark
Mako Shark

83.3 (20)
94.6 (70)
89.1 (53)
100.0 (6)
93.3 (14)
100.0 (1)

16.7 (4)
5.4 (4)
10.9 (2)
6.7 (1)

Table 2,
No. With
Tip & Edge

Distribution of Tip and Edge Wear by Species,
% of Total Mesial Distal Tip & Tip & All Mesial &
With Tip & Tip Edge Edge Mesial Edge Distal Edge Three Distal
Edge Wear Only Only Only Only Only Edqes Only

'iger Shark
anon Shark
;and Shark
bhite Shark







35.71(5) 14.28(2)
2,85(2) 0

3.77(2) 0
0 0
0 0


71.42 (10)



7.14(1) 0
21.42(15) 1.42(1)

13.20(7) 0
33.33(2) 0
7.14(1) 0

% Calculated as a % of total sample size




on the tip and the mesial edge with a culturally selected use of
these tools that does not wear the distal edge alone or only the
distal and mesial edges (Table 2). Striations are found primarily
on the tip and mesial edge (Table 3). With the exception of Tiger
Shark, all other species have teeth that are cracked and chipped,
with a range of 16.7% to 35.7% (Table 4). An examination of the
various forms of hafting shows that Tiger Shark teeth were hafted
in a different manner than the others (Table 5). Tool use required
mesial blunting or mesial notching only in addition to perforation
as method of hafting. No other method was utilized in hafting
Tiger Shark teeth. Such a hafting system would be advantageous
only if great pressure were to be applied to the mesial edge.
Wear percentages show the tip and mesial edge to be the most worked
with striations found exclusively in these regions. Martin, Quimby
and Collier (1947:Fig. 94: F, p. 396) illustrate a sword or slasher
from Key Marco in which the teeth used are Tiger Shark. Wear com-
mon to knives would be tip and mesial edge wear with the greater
pressure exerted in these areas causing teeth to crack and enamel
to chip.

All species show basically the same type of wear and can be
viewed as having served as knives. Tooth morphology is the variable
when perforations are considered and it is wear pattern percentages
which demonstrate similar use between categories. Since hafting and
perforations are determined by tooth morphology, only wear indicates
use and functional similarities between the various species. Stri-
ations are indicators of direction of use, amount of force used and
the hardness of the material being worked. Striations are not an
adequate index of similarity between species since they do not
appear on a large enough percentage of all samples to be meaning-
ful. For this reason, this analysis has used wear percentages to
compare species since it appears to be the best indicator of use.
Comparisons between species would then aid in the delineation of
complexes that were utilized in the same manner and transcend the
natural classification imposed by species with only slight morpho-
logical difference between them. In such a manner a functional
interpretation can be derived. This will confirm or deny other
archaeological and ethnological data and other forms of use can
be postulated. Regional specialization in one type of tool or
another, cultural selection of tooth and tool types and differences
in type and utilization through time and space are other areas
toward which such analysis can lead.

On the basis of wear, three basic analytical complexes arise.
These complexes exhibit internal similarity and are clearly separated
from each other on the basis of wear and morphology. The morpho-
logical differences account for the culturally selected uses of the
complexes and are a function of the wear/use patterns observed.

Complex A, consisting of Lemon Shark and Sand Shark, shows
much wear on the tip region: 42.9% and 50.0% respectively. Complex
B, consisting of Hemipritis serra and White Shark, shows minimal
wearcen the tip; 15.1% and 14.3% respectively. The greater tip wear

Table 3. Distribution of
Total % of Total Mesial
Sample Number Sample With Tip Edge
Species Size Striated Striations Only Only

Striations by Species.
Distal Tip & Tip &
Edge Mesial Edge Distal Edge
Only Only Only

Tiger Shark
Lenon Shark
Sand Shark
White Shark



37.50(3) 50.00(4)
94.28(33) 6.06(2)

50.00(17) 8.82(3)
16.66(1) 16.66(1)
60.00(6) 0

% Calculated as a % of Total Sample Size

Table 4. Percentages of Teeth Cracked and Chipped, Boca Weir (PB-56).
Total Sample % of Teeth % Chipped and With
Size Cracked Fractured Tips

Tiger Shark
Lemon Shark
Hemipritis serra
Sand Shark
White Shark

35.7 (25)
20.8 (11)
16.7 (1)
28.6 (4)

55.7 (39)
17.0 (9)
66.7 (4)
28.6 (4)


Mesial &






on Complex A is due to morphology since the tooth has a long,
narrow profile, a very sharp tip and small serrations. Complex
B is triangular in outline with large serrations on both edges.
B would best be utilized as a cutting tool using the naturally
sharp cusps on each edge while the keen tip found on Complex A
would have been an excellent tool for fine woodworking but could
not have been as effective a cutting tool as Complex B. Complex
C, or Tiger Shark, shows extensive tip and mesial edge wear with
wear confined to these areas almost exclusively.

The above has been inferred from tip wear alone. Further
confirmation must rest in the other classes of wear and hafting
method employed. When these are consulted, the parallel wear/use
patterns between the species in each complex become even more evi-
dent. Wear on both edges and the tip is more common (21.4% and
33.3%) in Complex A than B (13.2% and 7.1%). This would result
from the tip being pushed through some material. The lesser
figures for Complex B would indicate a reliance on the cutting
edgeof the tooth. This is confirmed when wear on the tip and
mesial edge only is computed. Complex B has much more wear in
this region (52.8% and 71.4%) while Complex A has little wear
(24.3% and 16.7%) confined to these areas. Such utilization is
further confirmed by the percentages of chips and fractures re-
moved from the tip region on each complex: A, 55.7% and 66.7%;
B, 17.0% and 28.6%.

Further patterning can be seen in the regions where there
is a lack of wear found on all specimens. Four species show a
complete lack of wear on the distal edge alone with only 2 (14.3%)
examples of this being seen on Tiger Shark teeth. When wear on
the tip and distal edge only is viewed, all five species show mini-
mal wear with a median of 7.1%. Wear on the mesial and distal
edges only is, like the wear on the distal edge only, almost
nonexistent. Only 1 (1.4%) Lemon Shark tooth has this form of

Conclusions resulting from this analysis of shark tooth tools
are many. Of all the variables examined, the most meaningful and
sensitive is that of wear. A complete analysis of wear gives the
primary indication of whether a tooth was used as a tool. The
other areas of analysis will confirm or deny this primary indi-
cator and lead to rejection of specimens. Areas of tooth wear
enable formulation of complexes that exhibit similar wear patterns
and are probable cultural types. The similar morphology of the
species within the complexes is shown to be a valid method of
analysis since wear patterns are similar between certain species.
Cultural selection of these complexes for certain usage is demon-
strated through wear patterns and hafting methods.

This type of analysis gives a more complete examination of
cultural complexes within shark teeth as well as a method for the
formulation of these complexes. The areas of similarity and
difference within and between the species demonstrates the exis-
tence of functional tool complexes within this culture. The
method used enables this same type of analysis to be performed on
other data since the recording aspects of the analysis are the
simple presence or absence of a trait. With this data, all

Table 5.

Distribution of Methods of Blunting and Notching for
Hafting Shark Tooth Tools at Boca Weir (PB-56).


Tiger Shark

Lemon Shark


Sand Shark

White Shark

" ) dP dP
20 55.0 10.0 36.4
(11) (2) (4)
70 35.7 44.3 64.0
(25) (31) (16)

(7) (2)
(7) (2)

64.5 20.0 20.0
(20) (3) (3)

53 37.7 60.4 50.0 43.8 4.6 31.8
(20) (32) (10) (14) (1) (7)

6 50.0 16.7 100.0
(3) (1) (3)

20.0 20.0
(3) (3)

0 20.0

9.1 18.2 13.6 22.7
(2) (4) (3) (5)


14 35.7 50.0 60.0 71.4
(5) (7) (3) (5)

0 100.0

0 0

0 0

* Calculated as a % of the number blunted and notched
** Sub-classification calculated from total sub-classifications


possible combinations of wear and hafting can be computed and the
percentages calculated. Differences in complex and species per-
centages between sites, as well as the different composition of
complexes, will enable hypotheses concerning tool use and haft-
ing methods to be formulated and tested as well as the possibility
of reconstructing tools that are unpreserved in the archaeological


It is impossible to compare the findings of this analysis with
any other site except Fort Center (Steinen 1971) where a large
enough sample of shark teeth has been recovered to warrant exten-
sive analysis.

The many shark tooth tools recovered at Key Marco by
F. H. Cushing (1896) have not been subjected to wear analysis but
were found hafted. Excavations in South Florida by Willey (1949)
and (1949) have recovered what was reported as fossil shark
teeth. Rouse (1951) found few shark teeth in his survey of the
Indian River region. Other excavations in Dade and Borward counties
have not produced large numbers of shark teeth. By itself, Boca
Weir is an anomaly.

The cause of this anomaly may be easily discovered. The sieve
size of the shaker utilized appears to be a major factor in the re-
covery of this artifact type. A one-half inch square sieve size
was used in the shaker for excavating Test Pit I at Boca Weir and
ten shark teeth were recovered in twelve levels of midden. In all
other excavation units a mechanical shaker with a bed of rib lath
was used to sift all the material. With the mechanical shaker,
the second pit produced a total of 20 specimens in five midden
levels while a third produced 93 shark teeth in sixteen midden
levels. The disparity between the three units could be a function
of work areas; however, no teeth are found below level five in the
midden in Test Pit I and all teeth recovered are comparatively
large. A much better explanation is that the difference in sieve
size is the major recovery variable when the material is screened.
This is further supported by the fact that the mechanical shaker
was used at both Fort Center and Boca Weir. No reference to sieve
size utilized is made in any of the publications in the Glades
region and future work should keep sieve size in mind in the plan-
ning stage of excavation for recovery of this artifact type.

To determine if sieve size could account for the large dis-
crepancies noted in the recovery of shark teeth in excavations in
the Glades area, an experiment was conducted. The 175 shark teeth
from Boca Weir and the 197 teeth from Fort Center were poured into
the one-half inch sieve by species. The screen was shaken no
longer than five seconds for each species; however, the majority
of teeth feel through the screen immediately. The results of this
experiment follow:


Boca Weir

Number Remaining in Shaker

Tiger Shark n = 24

Lemon Shark n = 74

Hemipristis serra n = 55

Sand Shark n = 6

White Shark n = 15

Mako Shark n = 1


Fort Center

Number Remaining

in Shaker

Tiger Shark n = 79 17

Lemon Shark n = 46 2

Hemipristis serra n = 20 10

Sand Shark n = 7 2

White Shark n = 13 4

Mako Shark n = 6 2

Carchardon megalodon n = 5 0

Unknown n = 6 3

Fragments n = 15 1

197* 41

*3 large unclassified fossil teeth not included.

The 41 teeth that remained in the shaker represents 20.5% of
the Fort Center Sample and is four times the figure for Boca Weir.
Reasons for this difference can be considered twofold: The
difference in sample sizes and perhaps a selection for larger teeth
in trade relationships since Fort Center was located inland. It is
interesting to note that the three species remaining in the sieve
at Boca Weir are the same as the three largest species that remained
from the Fort Center sample. The results of this experiment are
considered to be reflective of sieve size utilized rather than the
lack of occurrence of this artifact form.




A comparison of the Boca Weir shark teeth with those from
Fort Center can only be made in a general way since the findings
(Steinen 1971) report only general characteristics within the
categories and not percentages. Additional information on each
species is lacking in many areas and combinations of wear were
not computed for the Fort Center sample. While general trends
are important, the areas in which wear is not found are equally
important for inter-category comparison. Percentages appear to
be a sensitive method of comparing these categories.

The Tiger Shark teeth are very similar at both Boca Weir and
Fort Center. Perforations are present in just over 50% of each
sample with heavy wear on the tip and mesial edge. Beside the
fact that both samples have blunting and notching confined to the
mesial edge, no further comparisons can be made for other forms
of wear or preference for notching or blunting since the Fort
Center data is not given in this manner.

A comparison of the Lemon Shark teeth reveals a different
pattern. The Fort Center specimens demonstrate predominantly
mesial and distal wear while the Boca Weir specimens have a high
percentage of tip wear with only 1.4% of the teeth having only
mesial and distal wear. The preferred methods of hafting at both
sites are blunting and notching; however, bilateral blunting and
notching is the major pattern at Boca Weir with only one example
of bilateral attachment found at Fort Center. Bilateral blunting
or notching makes sense when the tip of the tooth or both edges
are the areas being used, since direct pressure would require a
well held tooth. There appears to be a different function between
this species at each site.

The Hemipristis serra sample differs between the two sites
under comparison with only the percent of perforated teeth remain-
ing approximately the same. The Tiger Shark and Hemipristis serra
teeth from Fort Center are similar. Wear on the tip and mesial
edge are prevalent at both sites; however, the Boca Weir sample
exhibits many diverse methods of hafting. Both samples have some
chipping of the occlusal region but the Boca Weir specimens exhibit
20.0 percent cracking of the tooth.

Both collections of Sand Shark teeth are quite small and show
the same wear characteristics. A major area of difference is that
all teeth show bilateral blunting or notching at Boca Weir with
only slight blunting found at Fort Center (the number of examples
with blunting at Fort Center were not given).

The White Shark teeth at both sites parallel one another;
however, both samples are relatively small. The Mako Shark teeth
cannot be compared with any accuracy since the Boca Weir sample is
composed of a single tooth and the Fort Center sample is composed
of 6. The other species utilized at Fort Center, while not unique
in Florida, were absent at Boca Weir.



1k- ~LC i

C b c d
CentimetersI I II I Il I 1 '''5


Fig. 1. Shark tooth tools from the Boca I.
Weir site. Note the worn hole on a, Centi
the gum modification and tip wear on
b, the blunting on c, and the extensive
tip wear on d; e is a Lemon Shark tooth
drill. Scales for a-d and e are different.

1 e ?


One last point of differences between the sites is the exis-
tence of two specimens that were evidently used as drills at Boca
Weir: one Lemon Shark tooth (see Fig l,e) and a Hemipristis serra
tooth. Wear of this sort was not found at Fort Center and these
are the only known specimens in the Glades area with this form of


Conclusions resulting from this analysis and comparison
indicate a number of differing aspects between the samples at
both sites. The same species were utilized and wear patterns
indicate different uses for some teeth between the sites. The
complexes delineated at the Boca Weir site do not appear at
Fort Center. A probable cause for the different wear between the
sites is the specialized nature of the analyzed sample from
Fort Center. Woodcarving was a major specialization in
the A-B/Pond area of Fort Center and could be a major cause of the
different wear patterns found. The time differences between the
sites could also be a cause for these differences or even the
environmental situations toward which each site is oriented.
In future analysis, wear and not striations should be the major
factor observed, since striations are found in fewer examples
and are generally associated with advanced wear.



The results of this analysis support the use of one tool
type at Boca Weir: This is the shark tooth knife. The existence
of shark tooth tool-use in the Glades area from Hopewellian times
to perhaps A.D. 1600 represents a lengthy continuum. Possibly
composite shark tooth tools are a late phenomenon or geographically
restricted and were not present at Fort Center and Boca Weir. They
are positively known only from Key Marco. The distribution of these
tools through space and time indicates that they were confined to
the Glades region. The finding of drilled shark teeth throughout
the Glades indicates a knowledge and some use of the tool over a
wide area, but heavy utilization may not have occurred north of
the Glades. Using present data this hypothesis may be true;
however, when other sites are excavated using a smaller sieve
size, it would not be surprising if shark tooth tools are found to
be a very widespread tool type throughout the Glades region and
much of coastal Florida.

References Cited

Cushing, F. H.
1896 Exploration of Ancient Key Dwellers Remains on the
Gulf Coast of Florida. Proceedings of the American
Philosophical Society, XXXV, No. 150, pp. 329-432.

Furey, John F.
1972 The Spanish River Complex: Archaeological Settlement
Patterning in the Eastern Okeechobee Subarea, Florida.
Unpublished masters thesis, Florida Atlantic University.

Goggin, John M. and F. Sommer III
1949 Excavations on Upper Matecombe Key, Florida. Yale
University Publications in Anthropology, No. 41.
New Haven.

Martin, P. S., G. I. Quimby and D. Collier
1947 Indians Before Columbus. University of Chicago Press,

Rouse, Irving
1951 A Survey of Indian River Archaeology, Florida. Yale
University Publications in Anthropology, No. 44,
New Haven.

Steinen, K. T.
1971 Analysis of the Non-ceramic Artifacts from a Hopewellian
Affiliated Site in Glades County Florida. Unpublished
masters thesis, Florida Atlantic University.

Willey, G. R.
1949 Excavations in Southeast Florida. Yale University
Publications in Anthropology, No. 42.New Haven.

Spencer, Massachusetts
December, 1976



Joseph N. Henderson

Self-treatment of disease is common in all cultures.
In our culture, pharmacies sell numerous non-prescription medicines
which find their way to the home medicine "shrine," that special
cabinet set aside for shelving medications for treatment of disease
(see Miner 1956). From here, chemicals and apparatus for adminis-
tration are taken for use according to some previously made self-
diagnosis. As the stockpile of medicines builds, the "patient" can
pick and choose from recently acquired medicines, left-overs from
some previous illness, or even share with friends and neighbors.
This systematic process often characterizes the early stages of
health care seeking behavior (Suchman 1965:151-152).

If folk medicine can be defined as the self-treatment of
disease based on informally gathered information which is applied
according to some systematic paradigm, then all people, urban and
rural, educated and not, engage in folk healing processes. Two
alternative medical treatment protocols are available: professional
and folk. The relative use of one over the other is difficult to
numerically describe. For example, if health care behavior patterns
are described by data obtained only from physicians and hospitals,
the data are likely to be unrepresentative of the whole (Mechanic
1972:446). The unreliability results from those persons whose ill-
ness behavior paradigm dictates some alternative health care treat-
ment to the professional scientific medical practitioner. In one
household survey, for instance, only one-third of those persons re-
porting illness sought a physicians care (White et al. 1961:891).
The remaining two-thirds of sick people who did not seek the
physician's care may not, however, have been idly suffering from
their complaints. They were most likely engaged in self-diagnosis
and treatment. Corroborative evidence for the presence of this
alternate health care behavior is found in many parts of the
country including Florida.

In rural north Florida information concerning self-diagnosis
and treatment was gathered by anthropologist Alice Murphree (1965;
1968; 1970). Although an abundance of remedies of which 56 in-
volved plants of 40 species), the local physicians knew little of
the incidence of these treatment methods (Murphree 1970:405).
Nevertheless, they exist and are a lively facet of health care

In the Fall, 1975, the author began a series of interviews
with a native Floridian knowledgeable and active in self-diagnosis
and treatment. The informant is 65 years old, white and female.
She was born in Taylor County near the town of Perry. Most of her
knowledge of self-diagnosis and treatment modes was learned after
moving to rural Alachua County in about 1937.

The Florida Anthropologist, vol. 30, no. 3, September 1977



Many of her treatments are characteristic of those reported
from rural north Florida (Murphree 1965; 1970). However, at least
one disease, diagnosis, and treatment method was collected that is
not recorded in the literature by Murphree and is, perhaps,
sufficiently unusual among rural whites to warrant attention.

This treatment procedure involves sucking blood from a series
of incisions made in the skin between the shoulder blades. Medical
history records various procedures for removing blood from the body
to effect a cure (Ackerknecht 1973:51, 111-112). Still, the treat-
ment described by the informant does not neatly fit into one of the
usual "bleeding" categories of blood-letting, leeching, or cupping.
For example, blood-letting and leeching were simply methods to
drain blood via arterial pressure and parasitic feeding. Cupping
involves the use of an intermediary instrument with which to form
a vacuum and collect blood. Furthermore, to characterize the
treatment as surgical scarification is too incomplete. Also, it
is certainly not therapeutic or decorative cicatrization. In the
medical sense, cicatrization would indicate an aid to healing by
intentional scar formation. The anthropologist considers cicatri-
zation a method to produce keloid tissue for beautification of the
body. However, the informant's therapeutic procedure fits into
none of these precisely.

The need for the therapeutic procedure is indicated by a
set of clinically visible symptoms. The diagnostic and therapeutic
procedure to be presented has been cross-checked with an older
female relative of the informant and was found to be accurate.
According to the informant, this affliction is specific to infants
and is simply termed "hives."

Symptom onset of "hives" often appears within one week after
birth and may reappear intermittently during the latter part of in-
fancy. The informant states that "hives" are visually similar to
measles or heat rash and characterized by elevated "bumps." At
this point in the disease process, the therapeutic procedure is
quite simple: administer a tea of catnip leaves (if necessary,
sassafrass tea can be substituted). However, the situation may
be such that the appropriate plants are not available. If the
disease is allowed to progress without treatment, then the "hives"
are reported to turn black, peel, and disappear. They are then
said to have "gone in." This represents a worsening of the con-
dition. Within hours, the fingernails and patches of skin will
become bluish-black thereby signalling the impending death of the

To prevent death, a final measure is taken. In the words of
the informant, the infant is scarifiedd." This process involves
making three small incisions about one quarter inch long between
the infant's shoulder blades. The formation of the incisions is
similar to that used to make a Roman numeral three (i.e., three
vertical incisions, side-by-side). The incising instrument in


the reported case was a straight razor. Thus far, this operation
describes simple scarification. But, this is only half of the

After someone makes the incisions, that person then puts his
or her mouth over the cuts (contacting the skin) and sucks blood
from them three successive times. The blood is spat into a towel
after each extraction. Note that no intermediary instrument is
used to separate the mouth from the skin or to collect the blood.
Therefore, the therapeutic procedure described by the informant is
a blend of surgical scarification and mouth-to-skin sucking to re-
move blood for curative purposes.

If this procedure is appropriately performed, the infant
should survive. Such was the case of the informant's son who was
scarifiedd" as an infant in 1929. She further mentions that she
personally knew of another infant who developed "hives" about the
same time as her son but had neither the tea treatment nor the
scarificationn" performed because of parental disdain for "old time
remedies." This infant died with the bluish-black skin and nails.
The informant's son survived.

The anthropological literature is replete with worldwide
instances of "sucking cures." These may involve sucking the skin
directly through any bloodletting of intermediary instrument
(Jones 1972:67, 77-78), or sucking the skin with a cupping instru-
ment but again without bloodletting (Dentan 1968:88). It is
interesting to note that one example of mouth-to-skin blood suck-
ing is reported from the Tuscaroras when they still resided in
North Carolina (Vogel 1970:173). The informant acknowledges the
interchange of medical knowledge between native Americans and
Europeans but claims no knowledge of the origin of the cure she

In disease causation theory, sucking cures are often associated
with beliefs concerning disease-object intrusion (Clements 1932;
Fejos 1963). When the informant was asked why blood was sucked
from the infant she again reported that she did not know. If this
cure was borrowed from southeastern native Americans at some point
in time, it seems that the "disease-object" aspect was lost. For
example, the towel into which blood is spat receives no special
treatment (e.g., hiding, burning, examination for objects, etc.) and
therefore it seems unlikely that the cure was directed toward re-
moval of specific disease object. Perhaps it should be mentioned
that although healing ceremonies to remove disease causing objects
often involved the display of the removed object (see Opler 1969:
72-73), this is not always necessary (see Wagley 1943:73-74).

Lastly, what disease might the informant have observed and
treated? The informant and her sister stated that this affliction
and treatment was common during their younger years (ca. 1920-1930)
but is not so today. They attribute the apparent decrease in the
incidence of "hives" to a "generally healthier youngster."




Although accurate determination is unlikely, one possible
disease category is the meningococcal infection. This type of
infection afflicts primarily persons younger than 15 years of age
with 50 percent of the victims less than 5 years of age.
Generally, the meningococcal infections are characterized by a
morbilliform skin rash, fever, and subcutaneous bleeding resulting
in petechial lesions (Mathies 1975:1-2). These symptoms are
similar to those described by the informant: measle-like rash,
fever, and black patches of skin.

The history of meningococcal infections is best recorded in
discourses concerning epidemic meningitis. The earliest medical
report of epidemic meningitis in the Unites States was in 1805
(Ackerknecht 1965:84). By 1900, epidemic meningitis was tenth on
the list of leading causes of death in the United States but is
presently not included (Underwood 1975:59). In Florida, the time
period during which the informant's son was ill and subsequently
scarifiedd" (i.e., 1929) was marked by an increase in epidemic
meningitis (Fig. 1). In general, however, the disease was de-
creased from 1900 to the present. This trend is consonant with
the informant's impression that "hives" has decreased in incidence.

Unfortunately, epidemic meningitis is but one manifestation
of meningoccal infections. Numerous other symptoms are indicators
of specific types of meningococcal infections. Three types may
occur. Meningitis is the most common and among children involves
fever, headache, failure to feed, vomiting, and a high pitched
cry. Two-thirds of meningitis patients have a skin rash.
Ultimately, lethargy, delirious behavior, and a comatose state
proceeds death (Mathies 1975:1-2).

Meningococcemia is a septicemic form that may have a ful-
minate course with extensive rash, high fever, and shock. No
other meningococcal symptoms need be present (Mathies 1975:1).

Lastly, chronic menigococcemis is the "mildest" form. It
is characterized by a slight fever, fleeting rash, and joint swell-
ing. These symptoms are self-limiting and resolve without treat-
ment (Mathies 1975:1-2).

In reviewing these three forms of meningococcal infections,
it is obvious that the informant's symptomatic description is not
sufficiently detailed to permit exact determination of the specific
infection she observed. It seems likely that all three menin-
gococcal forms may have existed in the community and because they
all have similar symptoms, observers may have confounded them.
Specific symptom patterns may have been superimposed on one
another, the fatal course not separated from the self-limiting

The treatment, whether it be administration of catnip tea
or the scarificationn" operation, would be the same for any of the
forms. In the self-limiting chronic meningococcemia form, the



treatment would be held responsible for the subsidence of symptoms.
If death occurred as a result of meningitis or meningococcemia,
failure to observe proper treatment protocol may be blamed or the
fault may be transferred to the supernatural (e.g., it was God's

In summary, the existence of bleeding to effect a cure has
a long history and wide geographic distribution. An informant from
rural Florida reports a disease of infants manifesting fever, skin
rash, and, finally, bluish-black patches under the skin. The treat-
ment involves administration of catnip tea or, if the tea is un-
available, scarificationn." Scarification involves incising an
area of skin between the shoulder blades and sucking blood from the
incision by mouth-to-skin contact.

The disease may have been in the category of meningococcal
infections. These infections, depending on form, may be self-
limiting or they may rapidly result in death unless adequate inter-
vention is employed. For the informant, the forms of meningococcal
infections were not distinct. Therefore, the self-limiting forms
would reinforce the treatment paradigm while a form with a fatal
outcome is traced to a failure to follow the proper therapeutic
procedure or to supernatural intervention.


1923 1924 1925 1926

1927 1928



References Cited

Ackerknecht, Erwin H.
1965 History and Geography of the Most Important Diseases.
Hafner, New York.

1973 Therapeutics: From the Primitives to the 20th Century.
Hafner, New York.

Clements, Forrest E.
1932 Primitive Concepts of Disease. University of California
Press, Berkely.

Dentan, Robert K.
1968 The Semai. Holt, Rhinehart, and Winston, New York.

Fejos, Paul
1963 Magic, Witchcraft and Medical Theory in Primitive
Cultures. In Man's Image in Medicine and Anthropology,
ed. by lago Galdston. International Universities Press,
New York.

Jones, David E.
1972 Sanapia: Comanche Medicine Woman. Holt, Rhinehart,
and Winston, New York.

Mathies, Allen W.
1975 Meningococcal Infections. In Practice of Pediatrics,
ed. by Vincent Kelley. Harper and Row, New York.

Mechanic, David
1972 Some Implications of Illness Behavior for Medical
Sampling. In Medical Men and Their Work, ed. by Eliot
Friedson and Judith Lorber. Aldine-Atherton, Chicago.

Miner, Horace
1956 Body Ritual Among the Nacirema. American Anthropologist

Murphree, Alice H.
1965 Folk Medicine in Florida: Remedies Using Plants.
Florida Anthropologist 18:175-185.

1968 A Functional Analysis of Southern Folk Beliefs
Concerning Birth. American Journal of Obstetrics and
Gynecology 102:125-134.

1970 Physician Dependence, Self-treatment Practices, and
Folk Remedies in a Rural Area. Southern Medical
Journal 63:403-408.

Opler, Morris E.
1969 Apache Odyssey. Holt, Rinehart, and Winston, New York.



State Board of Health of Florida
1933 Florida State Board of Health Annual Report, 1923-
1932. State Board of Health of Florida, Jacksonville.

Suchman, Edward A.
1965 Stages of Illness and Medical Care. In Patients,
Physicians and Illness, ed. by E. Gartly Jaco. Free
Press, New York.

Underwood, Jane H.
1975 Biocultural Interactions and Human Variation. Dubuque,
Iowa, Brown.

Vogel, Virgil J.
1970 American Indian Medicine. Ballantine, New York.

Wagley, Charles
1943 Tapirape Shamanism. Antropologia 3:73-74.

White, Kerr L., T. Franklin Williams, and Bernard G. Greenberg
1961 The Ecology of Medical Care. New England Journal of
Medicine 265:885-892.

Gainesville, Florida
June, 1976



William M. Jones

According to Webster, the pump is defined as "any of
various machines that force a liquid or gas into, or draw
it out of something, as by suction or pressure." There are
two basic types, the "lift" or "suction", and the "force"
pumps, both of which are reciprocal in motion.

The lift or suction pump consists of a cylinder in which
a piston moves up and down. Two valves are usually employed
which open upward, allowing water to flow in that direction.
The "foot valve" is located in the lower section of the
cylinder, while the other valve is mounted directly on the
moving piston. Since we are only concerned with the suction
or lift type, it will not be necessary to explain the mechanics
of the force pump.

From an historical standpoint, the suction and force pumps
both date as early as the second century B.C., when they were
developed by the Greek physicist and inventor, Ctesibius of
Alexandria. In the middle of the sixteenth century, Georg
Bauer Agricola of Saxony, said to be the founder of the sciences
of mineralogy and geology, employed suction pumps in mining
operations in Bohemia. In his work published in 1556,
De re Mettallica, there is a woodcut showing a suction pump
that had been made from a tree trunk (Agricola 1556: Book 6,
p. 135). We are especially interested in this method
because the pump we are going to discuss was made in a
similar fashion (see Fig. 1).

The Log Pump and other Artifacts

The log pump featured in this report was found in May,
1960, adjacent to historical site 8DU78X during the time it
was being excavated (Fig. 2). At this same time, a mosquito
control unit was digging ditches along the marshy shoreline in
the area. After these people completed their project, we
surveyed the nearby ditches and found a number of logs, tree
limbs and other debris that had been removed from the mud along
with the usual spoil. In the midst of this debris at a point
northeast of the main excavations, the remains of the pump
and other artifacts were found (Fig. 3).

The Florida Anthropologist, vol. 30, no. 3, September 1977



L. -4-



Fig. 1. Woodcut of Agricola's 1556 suction pump.


47 1 r-: r4


- -Bridle path--
(/ Bridle path



A /_ _%' W -


To Atlantic Blv4.,
three miles.

wooded area

I Old fence line

wooded area



\ r&Bridle path

old fields

Fig. 2. Plan showing main excava- I;
tions and the place where the
wooden pump was recovered; 1 inch
equals 80 feet.





___ I ___


The pump cylinder appeared in the form of a hollow log
broken into two sections by the machine that was operating here.
One of these sections had been placed on the side of the ditch
while the other section was still buried nearby (Fig. 3, A-B).
After removal, both sections of the pump cylinder and other
wooden artifacts were immersed in a nearby pool of water so
as to maintain them in their present waterlogged condition.
The other artifacts that were found in contact with the pump
cylinder are as follows:

A total of fifty pine wedges;

rectangular pine blocks;

oak and pine plugs, all of which varied in size
(Fig. 3, I); it is doubtful these wooden objects
were related to the log pump and the reason for
their presence at this point is unknown;

another wooden object that appeared was a section
of plank that had been cut in the "two forked"
configuration, with one fork missing. It was
probably made from a 5.1 cm. by 15.1 cm. pine
plank, and was .91 m. in length. It may have
been used as the piston rod in this pumping
system (Fig. 4);

a pine plank with rectangular slots near each
end, and measuring 5.1 cm. by 15.1 cm. by
1.81 m. in length. It has not been identified
(Fig. 3, E);

a complete wooden trough made of pine measuring
1.4 m. in length, and 28 cm. wide inside. It
was constructed of 5.2 cm. thick planks with
the mark "XIIII" showing on one side. This is
an assembly mark used by early builders (personal
communication, Charles H. Fairbanks, University
of Florida). This trough was likely placed
under the pouring spout of the pump to carry
excess water from the immediate area (Fig. 4);

an oak "collar", measuring 8.21 cm. in length
by 12. cm. wide on one end, and slightly smaller
on the other. Nail or tack holes were in evidence
on the larger end, and a 1.255 cm. wide groove
appeared around the middle. It was found inside
the bore of cylinder section "A", and could have
been either the piston, or part of the foot
valve assembly;







-\ PLUGS--

G F / \ D


5/14/60- 5/IS/60

Fig. 3. Distribution of wooden pump sections.


two odd shaped oak blocks (this and other woods
were identified by Frank H. Hill, District
Forester Florida Department of Agriculture)
measuring roughly 7.6 cm. wide by 4.53 cm.
thick. Nail or tack holes show on one flat
side of each block, with facets cut around
the edges (Fig. 4). We assume these blocks
were used as check valves in a manner that will
be discussed;

one complete 25 cm. diameter pewter plate with
an obscure hall mark on the bottom that resem-
bles a crown over the Roman X;

one large and one small fragment of a creamware
vessel. The presence of this ware is important
because it relates to nearby site DU78X which
produced the same type ware in quantities.

Description of the Log Pump

As we stated earlier, the log pump cylinder had been
broken into two sections by the dragline that exposed it.
Placing the two jagged ends together as best we could, the
cylinder totalled a length of 4.56 m. with a diameter of
30.53 cm. at the top and 20.51 cm. at the bottom, and was
made from a cypress tree trunk. In early England the elm
tree trunk was selected for the purpose (Eubanks 1971: 45).

The log cylinder had been drilled throughout its length,
with a 12.55 cm. bore in the top section, a 12 cm. bore in
the middle section, and a 8.2 cm. bore in the bottom section.
A pine plug had been driven into the bore at this point
(Fig. 4).

The water intake holes were 5.1 cm. in diameter and were
bored through the sides near the bottom of the cylinder in
the fashion shown in Figure 4. The outlet was a 10 cm. hole
bored through one side near the top of the cylinder and in
which a pouring spout was likely placed (Fig. 4).

Directly opposite the outlet hole a flat place had been
mortised in the outside wall and from which protruded two
"treenails", and several rusty wrought iron nails (Fig. 4).
The pump handle could have been anchored at this point, or
the cylinder may have been secured to a supporting surface
with the treenails

As mentioned earlier, the log pump here was the suction
type that employed a moving piston and check valve in the



Two-forked -
piston rod.


Collar and oak
block used as -

12.55 Cm.


Collar and oak
block used as
foot valve

12 Cm. ,

8.2 Cm.--


Plug '

30.53 Cm.

Water outlet.

Cypress log.

Two oak block check valves.
Tack or nail holes can be
seen on one.

Fig. 4. Possible placement
of parts in wooden water
pump; not to scale.

020.51 Cm.--



upper bore of the cylinder, and a foot valve in the lower
section. It is quite likely the oak collar and oak blocks
with tack holes appearing on one surface were used as one
or the other, as is explained below.

It was concluded that one of these blocks, with leather
attached, was fastened on top of the oak collar so the
leather would serve as both hinge and gasket in the style
of the "swing valve", allowing water to flow upward only.
In addition, the groove situated around the middle of the
collar would have held a gasket so water would be forced
through the open collar and check valve.

As stated above, the oak collar could have been used
as either the piston or as part of the foot valve assembly.
It is quite likely, however, that it served as the piston
because it was found resting in the upper 12.55 cm. bore,
and has a width of 12 cm. which is just enough tolerance
to allow unimpeded movement during the pumping cycle,
(Fig. 4). This would not be possible in the lower sections
of the cylinder because the bore diameters are smaller and
would not allow freedom of movement.

Just how the piston rod could have been connected to
the top of the collar in a manner that would not restrict
movement of the check valve, is not known. In this type of
pump, there are two accepted methods of doing this. In one,
the piston rod connects to the center of the piston with two
or more valves mounted on either side. In the other method,
the rod is equipped with a yoke that extends out to the edges
of the piston, allowing the check valve to be mounted in
the center. The two-forked section of wood found on the
site may have served as the rod in this instance (Fig. 4).

On the other hand, if the collar had been used as the
foot valve assembly, it would have been a simple matter to
push the collar down in the bore to its resting place, which
in this case would have been where the 12.55 cm. bore reduced
to 12 cm. (Fig. 4). Once the wooden parts and leather
gasket became waterlogged and expanded, it is doubtful the
collar would move during the pumping cycle. There is an
early sketch that shows a foot valve very much like the one
above (Eubanks 1971: 42).

At this point, we are unable to explain how the remaining
oak block was used. It is obvious it was equipped with a
leather gasket similar to the other block, but how it could
have been installed in the cylinder bore to act as a valve,
escapes us. Possibly some moving parts are missing from the
site inventory.



Suction pumps such as the one in this report must be
equipped with a strainer (Eubanks 1971: 42) on the bottom of
the cylinder, or the cylinder must be mounted in a "sump"
so as to prevent sand and organic materials from being drawn
into the bore with the incoming water. It is likely the
log pump was mounted in a sump on this site.

In early times, log pumps and log pipes were bored by
hand in the manner shown in Agricola's drawing in Figure 1,
or they were bored using water power. The first boring was
usually done with a small pod auger which established a
pilot hole throughout the log. Larger pod augers were then
used to increase the size of the bore to the desired diameter,
which was about 12.55 cm. These augers may also be seen in
Agricola's drawing.

The log pump from 8DU78X had three diameter bores stait-
ing with 12.55 cm. at the top and graduating down to 8.2 cm.
at the bottom. This was probably the results of the boring
procedure. It was noted that the wall of the 12.55 cm. bore
was exceptionally smooth with no decernable imperfections.
It must have been bored with a very sharp and efficient tool.

When the pump was found along the shoreline of the salt
marshes we were skeptical of the idea of pumping fresh water
under such conditions, and dug a test hole that soon filled
with water. An assay was done later by flame photometry,
and it was found that the water from the site contained less
sodium chloride than the domestic water of nearby Jacksonville,
Florida (personal communication, Paul C. Hudgins, Research
Chemist, St. Vincent's Medical Center, Jacksonville). The soils
here were also checked, and were found to be typical of the
"Ridgeland Series" (personal communication, L. T. Stem, Soil
Scientist, U.S. Soil Conservation Service, Jacksonville).

Site 8DU78X

Site DU78X was found during the year 1957 by William M.
Jones and a group of volunteers (Fig. 3) from the Jacksonville-
area. Later, it was excavated by Jones and the same group
under the sponsorship of the late Dr. John M. Goggin of the
University of Florida.

Among the many artifacts that were found on the site
were 3044 fragments of ceramics that included creamware,
delftware, white salt glazed stoneware, scratch blue salt-
glazed stoneware, and other types that all belong to the
eighteenth century. The only nineteenth century wares were
several bits of edged pearlware and banded ware. The



remainder of the recovered artifacts will not be discussed
since this report deals mainly with the log pump found nearby.

As the excavations progressed and materials were brought
to light, Goggin theorized the site belonged to the British
period, 1763-1783. As time passed, however, it became evident
the site was established during the second Spanish occupation,
1783-1820. There were several reasons for this last conclusion.
Site 8DU78X is located on a peninsular that is bounded on the
west by Greenfield Creek and marshes, on the east by Pablo
Creek and marshes, and on the north by the St. Johns River
and its marshes. During Spanish times this area was known
as "Punta del Cano de San Pablo", and approximated one
thousand acres.

According to the Spanish land grant records (1941: 109),
Robert Clarke Maxey of South Carolina, settled in this area
in 1791 with his family and slaves, a total of nineteen persons.
Later, he received this one thousand acre tract as a grant
from the Spanish government.

According to the same records, Maxey lived on and improved
the Punta de San Pablo Creek lands for the following thirteen
years or until 1804, when he sold the tract to John McQueen.
McQueen, in turn, conveyed the tract to George Taylor the
following year.

In addition to the above evidence, an early map of this
region shows Maxey's name on the north end of the peninsular,
near site 8DU78X. Also, a Spanish one real milled silver coin
dated "1780" was found. The appearance of this coin might
place this site in a later period than the English occupation,
since the coin would likely circulate some time before it
reached this remote area. The recovery of several fragments
of edged pearlware and banded ware would suggest some
activity here during the very early nineteenth century.

In view of the above evidence we may reason that site
DU78X was occupied during the second Spanish period and re-
presents the dwelling house of the Punta del Cano San Pablo
Plantation of Robert C. Maxey. We may also conclude that
the log water pump, found nearby on the marsh, was part of
that complex.

According to a deposition by Jan Eduardo Fate on May 17,
1821, Robert C. Maxey, a lieutenant in the Spanish militia,
died at "Small Hope Plantation", the home of his son, Peter
Maxey (Spanish Land Grants 1941: 85). Today, the site of
his San Pablo dwelling and the log pump are covered with a



thick growth of trees and brush, and other than a few bits
of tabby scattered about, there are no signs of an early

In conclusion, we regret our failure to find any early
sketches or information that might have explained the function
of the many unidentified objects found with the log pump
cylinder. No doubt some of the objects were thrown into the
marsh to dispose of them, while others were likely part of
the log pump system.

References Cites

Agricola, Georg Bauer
1556 De re Mettallica. Bern. (Original edition in
the Library of Congress.)

Eubanks, Bernard M.
1971 The Story of the Pump and Its Relatives. Salem,

Spanish Land Grants in Florida
1941 Briefed Translations, Confirmed Claims: K-R,Vol. 4,
Prepared by the Historical Records Survey
Division of Community Service Programs, Works
Progress Administration, L. B. Hall, Manuscripts
Editor. State Library Board, Tallahassee.



George M. Luer

The author first learned of the Roberts Bay site upon
seeing a large hammock-covered shell midden during a visit to
the Palmer Ranch in March, 1972. The Palmer Ranch was an
approximately 250-acre cattle ranch established by the Palmer
family during the 1910's and operated by the family into the
1960's. Discussion in 1972 of the development of the area into
a shopping center and subdivisions for homesites prompted the
author to request and secure permission to visit and dig at
the midden in order to perform a cursory investigation before
the possible disturbance or destruction of the site.

The Site

Figure 1 shows the location of the Roberts Bay site. The
site is on a cul-de-sac bay just to the north of the maze of
mangrove isles and oyster beds of the delta of Phillippi Creek.
The waters of the creek flow toward the coast from the north-
east through what was once pineland but are now residential
areas. Siesta Key and the Gulf of Mexico lie a short distance
from the site to the south and west. Downtown Sarasota lies
six kilometers to the north.

.A hammock of live oak (Quercus virginiana), red cedar
(Juniperus silicicola), and cabbage palm (Sabal palmetto) covers
the area of shell deposit. The hammock undergrowth is primarily
composed of marlberry (Ardisia escallonioides), white stopper
(Eugenia axillaris), and wild coffee (Psychotria undata).
Both hammock trees and undergrowth on the midden are character-
istic vegetation of old coastal ridges of sand and shell on
offshore keys. The similarity of certain conditions afforded
by the abandoned man-made shell heap to those of old coastal
ridges has led to the establishment on the midden of the
vegetation typical of old coastal ridges. The shore at the
Roberts Bay site supports red mangrove (Rhizophora mangle),
black mangrove (Avicennia germinans), white mangrove (Laguncularia
racemosa), and buttonwood (Conocarpus erectus).

Figure 2 illustrates the contour of the major shell de-
posits. Flat sandy ground underlies the shell. This pre-midden
ground surface is just off of a natural sand rise, once
covered by palmetto (Serenoa repens), and pine (Pinus elliotii
densa), but now by improved pasture, upon which lies a narrow
asphalt road that skirts the northeastern edge of the site.

The Florida Anthropologist, vol. 30, no. 3, September 1977



~ YIF~i~ Bay
r-I ~--~--, ~~7

7~ ~-CI --..I. J) :
-r- --~- :B::
-- ---t.~ r
r--^ 9

x-e ,
L~ j-_--Q
r-, 4
;';~;:::~iliii ;,;;.
::::i;: L



Fig. 1. Roberts Bay ca. A.D. 1900; stippled area represents
grass or sand shallows; swamp symbols represent wet areas;
tree symbols represent mangrove areas; 1, Roberts Bay site;
2, Martin site; 3, Old Oak site.


1 km
I !


* -

This sand terrace, to which extend low ridges of shell at the
north of the site, meets Roberts Bay just to the north of the
midden where the shore, as a consequence, is sandy and high.
On the southeast the contour map shows two shallow saltwater
"ponds" joined by a man-made ditch and connected with the bay
through pipes. During the 1920's workers at the Palmer Ranch
cleared and deepened these "ponds", which were shallow salt
water-filled depressions (probably invaded by salt water only
in recent centuries due to the rising level of the sea), and
connected both to Roberts Bay. Sometime during the same decade
workers dug a pond, now filled, about 60 m. east of the above-
mentioned depressions to provide a source of fresh water for
cattle. The excavators are said to have found several human
skeletons. Perhaps the area about this now-buried pond is a
cemetery associated with the midden to the west.

The ridge of shell and sand lying along the shore of
Roberts Bay is the main refuse and habitation midden. Addi-
tional midden deposit extends southward, however, about 100 m.
from the southern terminus of this ridge. Leading eastward
from the ridge are two narrow "causeways" of midden material,
one of which leads directly to a distinct mound apparently also
of shell and sand. The second "causeway" meets still another
leading from the aforementioned mound. At the junctureof
these two "causeways" a ramp descends to the southeast roughly
in the direction of the possible cemetery. The land to the
north and south of both eastward-leading causeways is of sand.
Immediately to the east of the mound and associated causeway
is a flat area from which, on the north, rises a ramp to the
midden. This flat area is bound on the east by two narrow
ridges of shell and sand. A five meter gap, centered with the
mound to the west, separates the two ridges.

Artifact Surface Collection

During several afternoons from March, 1972, to the summer
of 1975 the author, with the aid of family and friends, made
a surface collection of artifacts from the Roberts Bay site.
Hammers made from the shells of fighting conchs (Fig. 4) abounded
upon the midden and tens of such hammers were found. Three
left-handed whelk shell implements similar to the type called
by Goggin Busycon pick A (Goggin 1950: 241-2) were discovered.
A shell celt and a shell gouge, both carved from the body whorl
of left-handed whelk shells, were found. A polished and per-
forated piece of body whorl, also cut from the shell of a
left-handed whelk, was found broken and buried in midden debris
in the eroded bayside slope of the midden. Similarly dis-
covered was a small bone barb, perhaps a fishhook barb. Many
left-handed whelk shell columellae and two olive shell "beads"
were found. Many notched valves of quahog clams were found,
some of which possessed a battered external surface, especially
near the umbo.




Potsherds gathered from the surface of the Roberts Bay
site were either of sand temper or of mixed sand and "limestone"
temper. Plain sand-tempered sherds were most abundant
although plain sand- and "limestone"-tempered sherds were not
uncommon. The latter sherds contained variable amounts of
sand and variable amounts of different-sized pieces of a gray-
or tan-colored material in which was sand. Dr. Bullen has
called this material "limestone" (Bullen, pers. comm.) Dr.
J. Milanich suspects that some of the material is a temper
of bits of fired clay, perhaps of crushed potsherds or bits
of Fuller's earth. Similar tempering material from sherds in
Sarasota and Manatee counties has been shown to be Fuller's earth
(J.T. Milanich, pers. comm.) The exact composition of the
material is unknown. Rim sherds that contained this unknown
temper were incurved and identical in curvature to the in-
curved rim sherds found on bowls along the Gulf coast and in
southern Florida. Willey's Perico Plain pottery (1949: 364-5)
appears to be the same as this sand- and "crushed sherd"
-tempered pottery. Many potsherds of sand temper also
exhibited an incurved rim and some exhibited an incurved
chamfered rim. Only one decorated fragment, a gray sand-
tempered check-stamped sherd, was discovered. Rim profiles
are shown in Figure 3.

The presence of chamfered-rimmed potsherds suggests that
the Roberts Bay site was inhabited during the Perico Island arch-
eological period, or approximately 2000 years ago (Bullen, pers.
comm.). Sand- and "crushed potsherd"-tempered sherds, called sand-
and limestone-tempered sherds by Bullen, were apparently associated
with early Weeden Island pottery at the Terra Ceia site (Bullen
1951:14) and are perhaps from a contemporaneous period of habita-
tion at the Roberts Bay site. The abundance of fighting conch
shell hammers is indicative of the Perico Island period and, in the
Sarasota area, of deposits in which early Weeden Island potsherds
occur (Bullen 1951:14-15).


The author, with the generous help of a friend, Stanley
T. Stokes, dug a square meter pit into the midden (Fig. 2) during
late August and early September, 1974. The pit was excavated
in arbitrary ten centimeter levels and specimens uncovered
were boxed according to level of origin. The excavators
first cleared the excavation area and then removed a layer of
humus, five centimeters in depth, from the surface of the pit.
The first ten centimeter level began at the surface of shell
beneath this humic layer.


LUER 125

Fig. 2. Contour map of midden at Roberts Bay site in feet above
mean high tide level. Stippled area represents area of shell
deposition. Arrow indicates location of excavated pit.


Fig. 3. Rim profiles of potsherds; a, sand-tempered plain with
chamfered rim; b-d, sand-tempered plain; e,f, sand and "sherd"-
tempered plain.

Four layers were encountered in the pit, three of which
were composed primarily of shell refuse and the fourth of
sand and occupational debris. The upper two layers encountered
(5-60 cm) were trash strata, the shell of which was clean
along the bases of the layers but broken and mixed with sand
along the tops where each layer had at some time been exposed
and had formed the surface of the midden. Oyster and cross-
barred venus valves and the shells of king's crowns abounded
in the two strata. Bones and sherds were common. Beneath
these refuse strata was dark gray sand (60-160 cm) that con-
tained potsherds, charcoal, burned and unburned bone and
shell, and areas of gray ash-like material. Another shell refuse
stratum was under the sand and was mainly composed of oyster
shell just beneath the sand and scallop valves, many of which
were "stacked", beneath the oyster shell. Much of this shell
layer was below the present-day level of high tide and the
presence of water and the narrowing of pit walls forced
excavation in the pit to terminate. The base of midden deposit
was not reached.


As is shown in the table of the vertical distribution of

S5ccm I


LUER 127

collections from the pit, sherds that contained "crushed pot-
sherd" temper were absent below a depth of 85 cm. Such
sherds were present, however, in every ten centimeter level
above 85 cm. A similar absence of like sherds (called lime-
stone-tempered sherds) in lower depths of midden deposit was
found 32 km to the north at Terra Ceia Island (Bullen 1951:15)
and 64 km to the south at the Turtle Bay 3 site and Cash Mound
(Bullen and Bullen 1956:17-22, 28-29). Several thick sand-
tempered sherds with an incurved chamfered rim were uncovered,
all of which were from a depth below 125 cm. Although thick
sand-tempered sherds (around 1.2 cm in thickness) were found
in shallow levels of the pit and several thin sand-tempered
sherds (around 0.6 cm in thickness) in deep levels, thin
sand-tempered sherds were most abundant in shallow levels
while thick sand-tempered sherds were most abundant in deep
levels. This phenomenon was also duplicated at Terra Ceia
Island and the Cash Mound as well as at the Sarasota County
Mound (Bullen 1971:7).

In order to obtain a better idea of when sherds of mixed
sand and "crushed sherd" temper and thick sand-tempered
chamfered-rimmed sherds were produced, two worked shells of
Busycon contrariumfound in the pit, one from 45-55 cm and the
other from just below 185 cm, were sent to the University of
Georgia Geochronology Laboratory for radiocarbon dating.

The shell from the 45-55 cm level (UGa-1133), a hammer similar
to one illustrated by Bullen and Bullen in "Excavations on
Cape Haze Peninsula, Florida" (Plate IV,F), was dated at
123075 B.P. or (Present=A.D. 1950) approximately A.D. 720.

The shell from just below 185 cm (UGa-1134), perhaps a pounder,
was dated at 2215165 B.P. or (present=A.D. 1950) approximately
265 B.C. The first date lies within the early Weeden Island
period and the second within the Perico Island period.

Thus sand- and "crushed sherd"-tempered pottery appears
to. have been made in the Sarasota area at a time that was con-
temporaneous with the early Weeden Island period. This, as
was mentioned earlier, was also indicated by the findings of
Bullen at Terra Ceia Island. The second shell radiocarbon
date re-enforces the belief that thick sand-tempered pottery
displaying an incurved chamfered rim is of the Perico Island

Three left-handed whelk shell columellae were found at a
depth of 135-145 cm and were of original apex-to-siphonal canal
length, the thick body whorl completely removed from each.

Another Busycon shell columella uncovered at a depth of 155-165 cm
was a columella “hammer", the basal end of which was worn.
Worked bone that was discovered included a cut piece of the


plastron of a turtle, Chrysemys sp., (35-45 cm), the end of
a cylindrical bone pin broken at a girdling cut and a burned
piece of a second bone pin or awl (85-95cm),and a cylindrical
bone awl and a small worked piece of unidentified bone
(115-125 cm). Identified charcoal was of the burned wood of
slash pine and buttonwood. Inhabitants of the Roberts Bay
site gathered the former wood inland, the latter along the

Table 1. Vertical Distribution of Collections

Collections Depths in centimeters Totals
5-45 45-85 85-125 125-165 165+
sand-tempered 88 70 78 49 13 298
"potsherd"-tempered 13 26 39
limestone 6 11 3 4 1 24
sandstone 1 1 2 4
bog iron 2 1 1 4
mineralized bone 1 1 1 3
Strombus hammer 1 4 5 2 2 14
Busycon hammer 1 1
Busycon columella 5 5
notched quahog valve 1 1
olive shell "bead" 1 1
perforated ark shell 1 1
Worked bone 4 4
Charcoal X X X X

Food Remains

Mollusk. The shells of 30 species were found. The letters
on the left of the list that follows indicate the relative abun-
dance of shells in the pit: (a), abundant; (c), common: (1),
less common; (r) rare. The list of species is arranged in
descending order according to the relative abundance of the


Crassostrea virginica (common oyster)
Ostea equestris (crested oyster)
Melogena corona (king's crown)
Choine cancellata (cross-barred venus)
Argopecten irradians concentricus (bay scallop)
Euglandina rosea
Busycon contratium (left-handed whelk)
Strombus alatus (fighting conch)
Fasciolaria tulipa (brown tulip)
Mercenaria campechiensis quahogg)


LUER 129

11. (1) Pleuroploca gigantea (horse conch)
12. (1) Busycon spiratum pyruliodes (pear whelk)
13. (1) Fasciolaria lilium hunteria (banded tulip)
14. (1) Crepidula fornicata (slipper shell)
15. (1) Crepidula maculosa (spotted slipper shell)
16. (1) Crepidula aculeata (spiny slipper shell)
17. (1) Polinices duplicatus (cat's eye)
18. (1) Trachycardium egmontianum (rose cockle)
19. (1) Macrocallista nimbrosa (sun ray)
20. (1) Carditamera floridana (broad-ribbed cardita)
21. (1) Genkensia demissa granosissima (ribbed mussel)
22. (1) Atrina sp. (pen shell)
23. (1) Spisula solidissima similis (surf clam)
24. (1) Neotia ponderosa (ponderous ark)
25. (1) Dinocardium robustum vanhyningi (giant cockle)
26. (r) Murex pomum (apple murex)
27. (r) Conus spurius altanticus (alphabet cone)
28. (r) Oliva sayana (olive)
Shell (not food refuse): 29. (r) Vermicularia sp. (worm shell)
30. (r) Spiroglyphus sp. (worm shell)

The Phillippi Creek delta supplied the inhabitants of
the Roberts Bay site with an easily obtainable abundance of
oysters and king's crowns. The nearby shallows of Roberts
Bay and Little Sarasota Bay produced most of the other
shellfish species consumed including bay scallops and the
cross-barred venus, a small bivalve that lives below the
level of low tide commonly in beds of manatee grass
(Diplanthera wrightii). High grass about the margins of
hammocks and swamps yielded Euglandina rosea. Surf clams
and ponderous arks and some pen shells and giant cockles
were probably obtained on gulf beaches.

Crab. Two claws of the stone crab (Menippe mercenaria)
were uncovered.

Shark and Ray. The remains of seven species of shark
and two species of ray were discovered. All the sharks
and rays are shallow gulf and estuarine species (P. W.
Gilbert, pers. comm.).

1. Ginglymostoma cirratum (nurse shark)
2. Carcharhinus leucas (bull shark)
3. Carcharhinus maculipinnis (spinner shark)
4. Galeocerdo cuvieri (tiger shark)
5. Negaprion brevirostris (lemon shark)
6. Rhizoprionodon terraenovae (sharpnose shark)
7. Sphyrna sp. bonnetheadd shark)
8. Pristis sp. (sawfish)
9. Dasyatis sp. (stringray)


Boney Fish. Fish of almost all the species of which
remains were unearthed live, or can be caught, in bays. The
diversity of the species represented indicates that fishing
methods had to have been employed that could secure bottom
fish such as flounder and black drum, grassbed fish such as
seatrout, fast-swimming species of open bay such as mackeral
and crevalle jack, and surface fish such as mullet and
snook. The presence of the remains, in six levels of the
pit, of little tuna or bonita, a fast-swimming fish of the
gulf, and of the remains of large grouper individuals
indicates that some fishing was done in the Gulf of Mexico.
Fishhooks (Fig. 4, f) and probably bows and arrows and fish
spears were used in fishing,and nets and weirs could also
have been utilized.

1. Arius felis (bay catfish)
2. Centropomus sp. (snook)
3. Epinephelus sp. (grouper)
4. Caranx hippos crevallee jack)
5. Haemulon sp. (grunt)
6. Archosargus probatocephalus sheepsheadd)
7. Lagodon rhomboides pinfishh)
8. Cynoscion sp. (seatrout)
9. Pogonias cromis (black drum)
10. Sciaenops ocellata (redfish)
11. Chaetodipterus faber (spadefish)
12. Mugil sp. (mullet)
13. Euthynnus alleteratus (little tuna, bonita)
14. Scomberomorus maculatus (Spanish mackeral)
15. Paralichthys albigutta (flounder)
16. Chilomycterus schoepfi (striped burrfish)
17. Diodon hystrix (porcupinefish)

Amphibian and Reptile. The amphibian and reptilian
remains found are of two species common to pinelands and of
six species that live in or around water in swamps, ponds,
and streams (Conant 1975).

1. Siren lacertina (mud eel)
2. Natrix taxispilota (brown water snake)
3. Pituophis melanoleucus mugitus (Florida pine snake)
4. Farancia abacura (mud snake)
5. Crotalus adamanteus (eastern diamondback rattlesnake)
6. Chelydra serpentina osceola (Florida snapping turtle)
7. Chrysemys floridana peninsularis (peninsula scooter)
8. Deirochelys reticularia (chicken turtle)

Bird. Most bird remains could not be attributed to any
species. Only two bird bones were identified and were of
red-breasted merganser (Mergus serrator) and bald eagle
(Haliaeetus leucocephalus).


4 cm


Fig. 4. Artifacts; a-c, g, from pit; others from surface; a-c, Busycon
columellae tools; d, Busycon celt; e, notched quahog valve; f, bone barb (for
fish?);g, cylindrical bone awl; h, olive shell bead; i,1, views of Strombus hammer.




Mammal. In addition to remains of five terrestrial
species, a tooth and a piece of a sternal segment of marine
mammal (cetacean) were found.

1. Didelphis marsupialis (opossum)
2. Procyon lotor (raccoon)
3. Odocoileus virginianus (deer)
4. Neotoma floridana (eastern woodrat)
5. Sylvilagus floridanus (eastern cottontail)

Concluding Remarks

Indians lived at the Roberts Bay site during the Perico
Island archeological period and into a period that was
coeval with the early Weeden Island archeological period
farther north. It is possible that evidence of earlier
occupation at the site, such as fiber-tempered sherds, might
eventually be found. Midden remains indicate that the
Indians obtained much of their food by hunting, fishing,
and gathering shellfish. The Indians fished in both bay
and Gulf waters and hunted in both pinelands and swamps.
Remains of 5 species of mammal, 2 of bird, 7 of reptile,
1 of amphibian, 26 of fish, 1 of crab, and 30 of mollusk
were identified in midden refuse. As with other Florida
coastal sites, it is not known if the Roberts Bay site was
inhabited seasonally, at irregular intervals, or contin-
uously. Also unknown is whether or not the Indians
practiced horticulture. The Perico Island period is believed
to have preceded the practice of agriculture in the region
although horticulture (the term signifying a less intense
form of cultivation than the former) could have been
practiced (Bullen, pers. comm.). Our knowledge concerning
plant cultivation in the Sarasota area during the period
that was coeval with the early Weeden Island period is
equally uncertain.


The author would like to thank his friend Stanley T.
Stokes for his aid in excavating the pit. The aid of the
late Ripley P. Bullen and of Jerald T. Milanich, of the
Florida State Museum, was greatly appreciated. The author
would also like to think Dr. Elizabeth S. Wing, also of
the Museum for allowing me to compare excavated bones with
bones in the collections of faunal remains at the Florida
State Museum. While at the museum, Drs. Steven Cumbaa and
Rochelle Marinan helped me identify many bones. The aid of
Dr. Perry W. Gilbert of Mote Marine Laboratory was also
appreciated. The author would also like to thank his parents
for making financially possible the obtainment of two
radiocarbon dating.



References Cited

Bullen, Ripley P.
1951 The Terra Ceia Site, Manatee County, Florida.
Florida Anthropological Society Publication
No. 3. Gainesville.
1965 Florida's Prehistory. Florida From Indian Trail
to Space Age, Chap. 23 (Tebeau, Carson; Chauvin,
Bullen and Bullen). Southern Publishing Co.,
Delray Beach.
1971 The Sarasota County Mound, Englewood, Florida
Florida Anthropologist 24:1-30.

Bullen, Ripley P. and Adelaide K.
1956 Excavations on Cape Haze Peninsula, Florida.

Conant, Roger

Contributions of the Florida State Museum,
Social Sciences, No. 1. Gainesville.

A Field Guide to Reptiles and Amphibians of
Eastern and Central North America. Houghton
Mifflin Company, Boston.

Goggin, John M.
1950 Stratigraphic Tests in the Everglades National
Park. American Antiquity 15:228-246.

Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast. Smithsonian
Miscellaneous Collections, Vol. 113. Washington.

Sarasota, Florida
May, 1976



Jerald T. Milanich

In 1970 at the Twenty-Seventh Southeastern Archaeological
Conference held in Columbia, South Carolina, Joseph R. Caldwell pre-
sented a paper summarizing the cultural chronology for the Georgia
coast, citing twelve new radiocarbon dates from St. Catherine's
Island. Since the publication of Caldwell's data in 1971 the Univer-
sity of Florida has completed a three-year (1973-75) study of the
aboriginal cultures on Cannon's Point, located on the northern end of
St. Simon's Island. The purpose of this paper is to present the
Cannon's Point chronology (as derived from a ceramic seriation and
supported by fourteen radiocarbon dates) and correlate it with Cald-
well's chronology.

The University of Florida excavations on Cannon's Point in
Glynn County investigated thirteen aboriginal sites ranging in time
from ca. 2500 B.C. to A.D. 1650. One site was a ceremonial mound;
the remaining twelve were habitation areas ranging in size from small,
circular, shell piles 3.5 meters in diameter to a village 200 meters
on a side. All thirteen sites were located within a 448-acre peninsula
of land bordered on all but the south side by salt marsh and tidal
streams. Little St. Simon's Island and Sea Island, both east of St.
Simon's, protected the point from the ocean and allowed the formation
of a salt marsh system all along the eastern edge of the island.

Table 1 shows the site names (left side), the ceramic frequen-
cies (center), and the culture periods and dates (right side).
Table 2 lists the provenience data for the sites and the culture
periods. Included in the tables are three sites on St. Simon's Island
excavated by Preston Holder in the late 1930's (the Airport site and
Tests VB and VD) and two sites located on Cumberland Island, Georgia,
excavated by myself in 1971.

The pottery in Table 1 is classified by surface decoration (or
lack of it). In addition, the fiber-tempered ware (the two columns
on left in Table 1) is distinguished from the non-fiber wares (which
include pottery tempered with sand, grit, clay, and all combinations
of the three and which are shown as the ten columns on right in
Table 1). Type names are not specified. Analysis of the ceramics
from the Cannon's Point sites has demonstrated that while the pub-
lished descriptions of the earlier ceramic series--St. Simon's (a
variety of fiber-tempered pottery related to Stalling's Island Fiber-
tempered), Deptford, and Wilmington--work quite well for classifica-
tion purposes, the descriptions of the later (post-A.D. 1000) pottery
types do not.

Quite likely, this is due to the fact that the descriptions
for the Savannah and Irene series were based largely on pottery speci-
mens recovered from ceremonial contexts at the Irene site near the

The Florida Anthropologist, vol. 30, no. 3, September 1977


tid rliasdi lcar l
years 1700
AD 1700,

North Ceper Field -
Indian Field-
Seeot Coaper Field -

Test C -

Test B-C_
Test I-
Test 0-
Test A-

Upper Test I-

Stafford North -

Table Point -

Marsh -


AD S100

AD 1000

-.~ ictresicil -

AD 500 -


500 I.C -

Lewer Test E- 1000 BC

1500 BC -

Upper 9SC76-

Upper 9IC57-
Lower 9n76 C.
2000 IC

Lewer 9l 57-

2500 BC.

0 20 40 00 80 oT

Camberland Island, a.

Table 1. Ceramic seriation.

A1I690 -
A. 1625 -

A 1540 -
- .0 1400 -

A .. 1250 -

ADI 0 0 -

A. 700 --


Sutherland Bluff
Pine Harbor


St. Catherines


A.D. 400 -

II Deptford

AD. I -

500 B.C -


1000 BC. -


S2500 B.C. -



Type of site, location
reference for information


North Couper Field

Indian Field

South Couper Field


Test G


Test B-C

Test F Small shell midden (6 m by 7 m and ca.
15 cm thick at center) located back from
marsh, probably associated with an unlo-
cated house structure; southern Cannon's
Point, St. Simon's Island; Martinez 1975:66
Test D Small, shell heap (4 m in diameter and
18 cm deep) on bluff next to marsh;
east-central Cannon's Point, St. Simon's
Island; Martinez 1975:60-63

Airport Large village in central portion of
St. Simon's Island; Holder 1936

Test A Extensive shell heaps with little other
refuse extending along salt marsh; often
reaches 3 m in height; this test in edge
where midden was only 50 cm thick;
northern end of Cannon's Point, St.
Simon's Island; Martinez 1975:49-50

Upper Test E Extensive shell heap adjacent to marsh;
east-central Cannon's Point, St. Simon's
Island; Martinez 1975:60, 63-66

Stafford North Refuse associated with house structure
adjacent to marsh; west-central Cumber-
land Island, Georgia; Milanich 1971:39-44

Table Point Refuse and living floor associated with
house structure adjacent to marsh; north-
west Cumberland Island, Georgia; Milanich

Marsh Midden underlying marsh muck on old land
surface adjacent to 9-Gn-57; northeast
Cannon's Point, St. Simon's Island;
Marrinan 1975

Lower Test E Non-shell living floor under extensive
shell heap adjacent to marsh; east-
central St. Simon's Island, Georgia;
Martinez 1975:64-66
Upper 9-Gn-76 Shell ring located on dry land; north-
east Cannon's Point, St. Simon's Island
(just southwest of 9-Gn-57); Marrinan

Upper 9-Gn-57 Shell ring located in marsh (just
northeast of 9-Gn-76); northeast
Cannon's Point, St. Simon's Island
Marrinan 1975

Lower 9-Gn-76 Shell ring (see above)

Lower 9-Gn-57

Shell ring (see above)

Oyster shell and refuse dump adjoining
living area on north end of Guale
village (ca. 200 m in diameter); north-
eastern Cannon's Point, St. Simon's
Island, Georgia; Wallace 1975

Refuse scattered within and around
large, pavillion-like structure,
central portion of same village
area as above; Wallace 1975

Refuse scattered within and adjacent to
mortuary structure; central portion of
same village area as above; Wallace 1975

Low shell and refuse heap on eastern
edge of same village; Holder 1937:7

Small, low, circular shell midden
(ca. 3.5 m in diameter; probably associ-
ated with unlocated house structure;
southern Cannon's Point, St. Simon's
Island; Martinez 1975:66-68

Low, oval shell midden back from marsh
(ca. 20 m by 10 m); northwestern
Cannon's Point, St. Simon's Island;
Holder 1937:6

Same as V-B; Martinez 1975:50-60

Table 2. Provenience data for seriation.


Sherd sample size and typology
used in seriation; reference

545; Wallace 1975:253-254

574; Wallace 1975:253-254

1375; Wallace 1975:253-254

57; Holder 1937:11, 12

210; Martinez 1975:79-82

695; Holder 1937:11, 12

315; Martinez 1975:79-82

122; Martinez 1975:79-82

204: Martinez 1975:79-82

21,077; Holder 1936:7

90; Martinez 1975:77-78

277; Martinez 1975:79-82

105; Milanich 1971:42

306; Milanich 1973:113

216; Marrinan 1975:57

197; Martinez 1975:79-82

33; Marrinan 1975:56

61; Marrinan 1975:56

II; Marrinan 1975:56

114; Marrinan 1975:56

Date in radiocarbon
years; reference

none, dated by

none, dated by

none, dated by
seriation; European
materials present

none, dated by

710+75: A.D. 1240
(UM-670); 510+75:
A.D. 1440 (UM-701);
Machover 1975

none, dated by

990+75: A.D. 960
(UM-667); Machover

1060+70: A.D. 890
(UM- 73); Machover

1130+70: A.D. 820
(UM-r68); Machover

none, dated by

1190+70: A.D.760
(UM-r72); 1240+75:
A.D. 710 (UM-671);
Matchover 1975

1300+80: A.D. 650
(UM-669); Machover

none, dated by

1895+95: A.D. 55
(UGA-129); Milanich

2770+95: 820 B.C.
(uM-519); 2785+80:
835 B.C. (UM-5T8);
Marrinan 1975:49

none, dated by

3605+110: 1655 B.C.
(UM-523); Marrinan

3765+90: 1815 B.C.
(UM-R21); Marrinan

3860+90: 1910 B.C.
(UM-522); Harrinan

4190+90: 2240 B.C.
(UM-520); Marrinan


mouth of the Savannah River. Also, as Larson (1958) and Caldwell
(1971) have both noted, there is considerable variation in ceramic
inventories along the Georgia coast. The St. Simon's specimens
used in the seriation--all from village contexts--exhibit greater
varieties of attributes than those specified in the type descrip-
tions. Savannah and later period sherds, for example, display a
wider range of tempering combinations than those noted in the type
descriptions. Also, the degree of workmanship executed in the
application of surface decorations generally is less fine than
that described in the type definitions. In other words, the later
St.Simon's specimens cannot be sorted according to the present defi-
nitions. The variety of tempers, rim forms, surface decorations,
etc., is such that some check stamped and complicated stamped sherds
fromDeptford contexts cannot be distinguished from similar sherds
from late Savannah contexts. Single sherds are generally not diag-
nostic as to cultural association. In order to try and produce
pottery definitions which are more usable and more diagnostic of cul-
tural periods, a detailed attribute analysis of the ceramics from
the St. Simon's excavations is being carried out and will be published
at a later date. Eventually, due to the variety of types and wares
within any one site (even those sites restricted temporally), it
will probably be necessary to chart the changing relative frequencies
of types and ware groups through time in order to produce the most
useful ceramic seriation. As Stanley South (1973) has shown for the
South Carolina coast, there is no one-to-one correlation between ware
group and culture period. And, as the coastal cultures evolved
toward the historic period, their greater cultural complexity (and
more rapid cultural evolution) is reflected in greater ceramic diver-
sity in village middens.

The Ceramic Sequence

The sequence in Table 1 follows Caldwell's sequences with two
changes. As noted by Larson (1958) and reaffirmed by the University
of Florida's research, the Irene ceramic complex found on the northern
Georgia coast is not present on St. Simon's. The local archeological
complex, which is related to Irene, is Pine Harbor. The Pine Harbor
period is followed by the Sutherland Bluff complex and period, the
local equivalent of Altamaha. Pine Harbor and Sutherland Bluff are
the archeological manifestations of the Guale Indians. Larson has
demonstrated that Spanish materials occur at Pine Harbor sites in
relatively large quantities, indicating that Pine Harbor temporally
extends into the mission period.

In addition to these changes, it is noted below that the con-
cepts of Deptford III, Wilmington, and St. Catherine's as arche-
ological complexes and periods (as used by Caldwell for the northern
coast) may also be somewhat different in the Glynn County locality.

Although the succession of cultures and temporal periods listed
in the two chronologies are similar, it is difficult to make detailed
comparisons since no quantified ceramic data accompanied Caldwell's
(1971) data. Consequently, the diagnostic ceramic type-frequencies
for the Savannah I period on St. Simon's may differ from those



of the same period in the Savannah River region. However, Caldwell's
paper which accompanies his chronology does provide some information
regarding ceramic traits, and in the following summariesa few compari-
sons can be made. The reader is referred to Table I for the frequency
changes mentioned in the summaries.

Sapelo Period--2500 B.C. to 1000 B.C.

As elsewhere on the Georgia coast, the ceramics of this period
are fiber-tempered. Marrinan's (1975) excavations on St. Simon's
Island showed that decorated sherds (some incised with Orange-like,
Florida fiber-tempered motifs, some with Tick Island motifs, and
some incised and punctated with Stalling's Island-like motifs) are
as early as plain specimens.

Refuge Period--1000 B.C. to 500 B.C.

This period is characterized by the addition of grit and sand
tempered pottery to the fiber-tempered assemblage. By the end of this
period, grit and sand replaced vegetable fibers as the primary form
of grog. Appearing with the new tempering techniques is the practice
of malleating pot surfaces with carved, wooden paddles. Check and
simple stamped motifs, the latter most popular, appear during this

Deptford Periods--500 B.C. to 700 A.D.

Although the Deptford periods are well-represented elsewhere on
the Georgia coast, few Deptford sites were found on Cannon's Point.
Intrusion of the salt marsh into the eastern side of the point appears
to have eroded away most of the sites. Those that were found were
late Deptford and were located well back from the marsh.

Elsewhere on the coast, however, the Deptford I period, 500
B.C. to A.D. 1, is characterized by an increase in check stamped
pottery along with the other Deptford carved paddle, malleated
motifs. Decorated pottery increased in popularity to more than
fifty percent of the total ceramic inventory. As noted by Caldwell
previously, Deptford II, A.D. 1 to A.D. 400, is characterized by
the addition of complicated stamping to the Deptford inventory.
Deptford III, A.D. 400 to 700, is defined as that time when plain
pottery again increased in frequency to greater than fifty percent
of the total inventory.

A small percentage of the pottery (4%) from the Cannon's Point
Deptford III site (Upper Test E) is tempered with clay particles
(ground sherds?). Caldwell (1971:90) notes clay tempering in the
St. Catherine's Island Deptford III pottery, also. Throughout the
Deptford periods, complicated stamped pottery increased in popu-
larity; various Swift Creek motifs, all poorly executed, predominate.

Wilmington Period--A.D. 700 to 1000

During the Wilmington period there appears to have been an
intrusion to the coast of the Wilmington ceramic complex--clay-


tempered, cord marked pottery--if not an actual movement of people
from the inland Coastal Plain river valleys. New evidence for such
a movement and its temporal placement, recognized in the 1930's by
Caldwell, Waring, and others, has been presented elsewhere (Milanich
et al. 1976). The distribution of Wilmington-like ceramic assem-
blages in south-central Georgia is much greater than previously
suspected (personal communications, Frank Schnell, Columbus Museum
of Art; and Chris Trowell, South Georgia College). The Cannon's
Point excavations suggest that the Wilmington peoples and the in-
digenous population were both utilizing the barrier island during
the same general period. Both sites characterized by Wilmington
ceramics (Test D) and other sites with a late Deptford ceramic
assemblage (Tests A, F, B-C, Holder's VB) were apparently occupied
during this same period.

Interestingly, the St. Simon's Airport site excavated by
Holder in the 1930's also seriates in the chart about A.D. 700,
the time that complicated stamped pottery is highest in popularity
in the late Deptford sites. The Evelyn site, a Swift Creek mound
complex and village located in Glynn County, westward from St.
Simon's Island, possibly was a "colony" established in the coastal
strand by Swift Creek peoples. This introduction, prior to A. D.
700, must account for the increase in complicated stamped ceramics
at Deptford III sites and the recently recognized presence of a
number of Swift Creek ceramic sites (all smaller than the Airport
Site) in Glynn County, both on the mainland and on St. Simon's
Island (Cook 1977). Cook (1977) has proposed the name Kelvin for
this ceramic complex.

If indeed Kelvin (Swift Creek), Wilmington, and late Deptford-
like ceramic assemblages were somewhat contemporary in the Glynn
County area, then there must have been competition between
various peoples for access to horticultural lands on the mainland
and camp and village locations on St. Simon's Island. That this
period was a time of conflict may be reflected in an adult male
burial recovered from the Test B-C excavations which had five
stone projectile points in his body (Martinez 1975:52-54), This
period needs a great deal more study in the Glynn County area
before any definite statements or interpretations can be made re-
garding the effects of new ideas on the conservative, coastal way
of life.

St. Catherine's Period--A.D. 1000 to 1250

This period, the ceramics which are described by Caldwell
as transitional between Wilmington and Savannah, is poorly repre-
sented in the St. Simon's excavations. Tests B-C and F fall close
to this period, but the respective ceramic inventories differ from
those described by Caldwell for St. Catherine's Island. However,
a few sherds of net marked pottery which, according to Caldwell
(1971), characterizes the period, were recovered from Test G. From
the gap in the Table 1 seriation, it seems likely that on Cannon's
Point cord marked pottery continued to increase in popularity




during the St. Catherine's period. Trying to fit the St. Catherine's
period into the St. Simon's culture sequence may, however, be a
futile exercise. The recognition of a Kelvin complex and the
chronological evidence from Cannon's Point together indicate that
cultural developments in the immediate area may differ from those
elsewhere on the coast.

Savannah Periods--A.D. 1250 to 1540

On St. Simon's Island the Savannah I period, A.D. 1250 to
1400, is characterized by an increase in check stamped ceramics,
a gradual decrease in cord marked pottery, and a more rapid decrease
in plain pottery (see Table 1). During the succeeding Savannah II
period, A.D. 1400-1540, cord marked pottery again increased in
popularity as did complicated stamped pottery tempered with
various combinations of sand, grit, and clay. It was in the
Savannah periods that two events occurred in the St. Simon's
Island area profoundly affecting the evolution of the central
coastal cultures. First, Mississippian influences from the
north Georgia coast (e.g.,the general area of the Irene site)
and/or from central Georgia began to influence the central coast.
It was also in this period, in the early 16th century, that
the first Spaniards reached the Georgia coast. One dog burial
from a late Savannah charnel house on St. Simon's Island
contained a dog burial with amusket ball between the ribs.

Pine Harbor Period--A.D. 1540 to 1625

As noted by Larson (1958), the Pine Harbor culture extended
temporally into the early Mission period. A Pine Harbor ceremonial
mound on St. Simon's contained European coins, beads, and wrought
iron spikes with burials. Spikes were also found with a cache of
ceremonial vessels which included Irene Incised and San Marco
Stamped bowls. During this period complicated stamped ceramics con-
tinued to increase in popularity and cob marked potsherds appeared.
Tempering with heavy particles of grit and clay, many two to five
millimeters in diameter, was practiced. Rectilinear stamped
motifs (San Marcos-like) also became popular. Cord marking declined.

Sutherland Bluff Period--A.D. 1625 to 1680

The Sutherland Bluff St. Simon's village probably was aban-
doned early in the period. Perhaps, as was often the practice, the
aboriginal population was moved and consolidated with a mission
village elsewhere--probably one near the south end of the island
where the mission is thought to have been. Throughout the Suther-
land Bluff period San Marcos Stamped pottery increased in popularity
until it became the most popular decorated type. By the time of
the abandonment of the Guale missions San Marcos pottery and
its variant, Altamaha Line Blocked, were found all along the
Georgia coast at Spanish-Indian sites.




The St. Simon's Island project was funded by grants from
Sea Island Foundation and the National Science Foundation awarded
to Dr. Charles H. Fairbanks and the author. An earlier version of
this paper was presented at the Annual Meeting of the Society for
American Archaeology in St. Louis in May, 1976.

References Cited

Caldwell, Joseph R.
1971 Chronology of the Georgia Coast. Southeastern Archae-
ological Conference Bulletin 13, pp. 88-92.

Cook, Fred C.
1977 Kelvin: A Middle 1st Millenium Adaptive System of
Marsh Efficiency. Paper presented at the Annual
Meeting of the Florida Anthropological Society,

Holder, Preston
1936 [Letter written to A. R. Kelley; St. Simon's Island,
Georgia, December 12, 1936.] On file, Southeast
Archeological Center, NPS, Tallahassee, Florida.

1937 [Letter written to A. R. Kelley; St. Simon's Island,
Georgia, July 26, 1937.] On file, Southeast Archeo-
logical Center, NPS, Tallahassee, Florida.

Larson, Lewis H., Jr.
1958 Cultural Relationships between the Northern St. Johns
Area and the Georgia Coast. Florida Anthropologist

Machover, Andrew W.
1975 The Use of Radiocarbon Dating in the Determination of
the Cultural Sequence for St. Simon's Island, Georgia.
Report on file, Dept. of Social Sciences, Florida State
Museum, Gainesville.

Martinez, Carlos A.
1975 Culture Sequence on the Central Georgia Coast, 1000 B.C.-
1650 A.D. MA Thesis in Anthropology, University of
Florida, Gainesville.

Marrinan, Rochelle A.
1975 Ceramics, Molluscs, and Sedentism: The Late Archaic
Period on the Georgia Coast. Ph.D. Dissertation in
Anthropology, University of Florida, Gainesville.

Milanich, Jerald T.
1971 The Deptford Phase: An Archeological Reconstruction.
Ph.D. Dissertation in Anthropology, University of
Florida, Gainesville.



Milanich, Jerald T.
1973 A Deptford Phase House Structure, Cumberland Island,
Georgia. Florida Anthropologist 26:105-118.

Milanich, Jerald T., Carlos A. Martinez, Karl T. Steinen, and
Ronald L. Wallace.
1976 Georgia Origins of the Alachua Tradition. Florida
Department of State, Bureau of Historic Sites and
Properties Bulletin 5, pp. 47-56.

South, Stanley
1973 Indian Pottery Taxonomy for the South Carolina Coast.
Paper presented at the Thirtieth Southeastern
Archaeological Conference, Memphis, Tennessee.

Wallace, Ronald L.
1975 An Archeological, Ethnohistoric, and Biochemical In-
vestigation of the Guale Aborigines of the Georgia
Coastal Strand. Ph.D. Dissertation in Anthropology,
University of Florida, Gainesville.

Gainesville, Florida
July, 1976



Back issues of the FLORIDA ANTHROPOLOGIST, if available,
may be purchased from the Treasurer at $6.00 per volume ($1.50
for single and $3.00 for double numbers). Volumes 25:2, pt. 2
(Publication No. 6) and 28:3, pt. 2 (Publication No. 7) are $2.00
each. Volumes 1; 2:3-4; 5:1-2; 6:1, 4; 7; 8; 9; 10; 11:1,2,4;
13:1; 17:1; 18:1; 20:1-2; 24:1; 25:3-4; and 26:1 are out of print.

a monograph series, may be purchased as follows: Numbers 5, 6,
and 7 at $2.00 each, Number 8 at $3.00. Numbers 1, 2, 3, and 4
are out of print.

Orders with remittances for any of the above should be sent
to the Treasurer whose address is given on the inside of the front
cover. The Society has no resale discount for dealers or resale


THE FLORIDA ANTHROPOLOGIST published original papers in all
subfields of anthropology with an emphasis on archaeology. Contri-
butions from allied disciplines are acceptable when concerned with
anthropological problems. The journal's geographical scope, in
order of importance, includes Florida, Southeastern United States,
Caribbean, and related regions.

Manuscripts should be double-spaced and typed on one side
only of an 8h by 11 inch sheet leaving at least a 1 inch margin
on all sides. Contributors are referred to recent back issues of
the FLORIDA ANTHROPOLOGIST for matters of style and reference.
Textual references should be similar to "(Smith 1964: 78-80; Jones
1969: PI. II, b)." No footnotes. In the "Reference Cited"
section, at the end of an article, do not abbreviate names of
series or journals. Books need full title, publisher's name, and

All illustrative material--line drawings (in India ink) and
photographs--should be included in one numbered series of "Figures"
suitable for a maximum reduction to 6 by 8 inches. Parts of
figures are designated by lower case letters. Figures must be
accompanied by brief descriptive captions. All artwork should be
camera-ready, excepting captions. Figures 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.

Authors should submit the original and one copy of their

Full Text