2000 Florida Anthropological Society Inc.
The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
all rights to the
and shall be
and images of
The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.
The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.
Contact the Florida Anthropological Society for additional
information and permissions.
A VALIENTE GUAYMI CAYUHO HAULING JUNTA
Louis Daniel Tesar
Among the Valiente Guaymi of the Republic of Panama the junta (Span-
ish, pronounced hoon. tah) represents an ethnographic survival which has
been reinforced in the face of 4 centuries of acculturation while other native
institutions have tended to deteriorate. As such, the junta can be considered
an important aspect in the maintenance of the structural continuity of Guaymi
society. The cayuco (Spanish, pronounced cah.ik.l k) hauling junta has been
selected to demonstrate this proposition. A cayuco is a dug-out canoe made
from a single tree, and a junta is a get-together similar to the house and barn
raising or harvesting activities which occurred in rural areas of the United
States until recent times. (Data for this paper was secured while the author
was a member of the U. S. Peace Corps in the Republic of Panama from De-
cember, 1968 to April, 1971, at which time he participated in 8 cayuca haul-
The Guaymi, who call themselves Ngawbe (pronounced Ngaw. b), are
one of the major Indian tribal groups in the Republic of Panama. They pre-
sently occupy portions of the 3 western provinces of Bocas del Toro, Chiriqui
and Veraguas (Fig. 1); although, they are believed to have also occupied the
Azuero region in the 16th and 17th centuries (Johnson 1948). Their entrance
into the province of Bocas del Toro seems to have been the result of their dis-
placement from the plains area by Spanish colonists (Johnson 1948), or by
other indigenous groups who had themselves been displaced by Spanish col-
Concerning their related migration legends, Tesar (1971:4-11) reports
that the Guaymi (Ngawbe) came from the Pacific side of the Talamanca Moun-
tain Range where they had been displaced by Spanish colonists (or, by other
indigenous groups). The area of Bocas del Toro (Fig. 1) presently occupied
by the Guaymi was formerly occupied by other indigenous groups: Woin,
Deco, Buguta, Wide and Muski (Mosquitos). The wars between these groups
as well as against the invading Guaymi lasted for generations. The Guaymi
eventually killed (or assimilated) what was left after the others had decimated
each other. Both the Deco, who occupied the coastal areas of present day
Bocas del Toro, and the Mosquitos, who came from Honduras on raiding
parties for slaves, were seafaring peoples possessing cayucos.
It is claimed that the Guaymi did not have cayucos or know how to make
them when they came to the Cricamola River area of Bocas del Toro (Fig. 1,
inset). There was, however, a small Mosquito (Muski) colony at the mouth
of the Cricamola River who remained after the others returned to Honduras
at the end of their wars with the Deco (for reasons of a contagious disease?).
Florida Anthropologist, vol. 27, no. 4, December 1974
A GUAYMI JUNTA
Azul C A CARRIBEAN SEA
Laguno de Chiriqui
P ''., i" anam a
SFig. 1. Map of Panam
Slocating Comarca de
S ,: Bocos del Toro (India:
reservation) and Peni:
PACIhFIC OCEAN sula Valiente.
,tied here /
.R Fig. 2. Map of route
Route taken I
taken during cayuco
hauling junta of
o, n Mauricio.
The Guaymi relate that they got their cayucos from those Mosquitos. There
was a small war in which the Guaymi killed all of the Mosquitos in the colony.
The Guaymi used these cayucos to occupy the Peninsula Valiente area (Fig. 1,
insert) from which the Deco had been driven by the Mosquitos.
Today, according to preliminary releases of the Censos Nacionales de
1970 of the Republic of Panama, there are about 50, 000 Guaymi, most of
whom live within the area designated as the Comarca de Bocas del Toro.
Their subsistence centers around slash-and-burn (milpa) agriculture with
corn as the dominant crop on the Chiriqui side of the Talamanca Mountain
Range (Tyson 1972; Young 1968, 1970); whereas, slash-and-fell agriculture
with root crops and bananas dominates on the humid Bocas del Toro and Vera-
guas side of the mountains. Hunting, fishing, gathering and, more recently,
the raising of domesticates such as cattle, pigs, goats, and chickens are of
secondary importance (Tesar 1971).
The Guaymi in the Province of Bocas del Toro divide themselves into 2
principal subgroups: the Cricamolans who occupy the Miranda or Cricamola
River Valley, and the Valiente who occupy the Peninsula Valiente (Fig. 1).
The latter, however, recognize themselves as having originally come from
the region of the Cricamolans. Both immigration and intermarriage within
the 2 areas is relatively common at the present time. The main differentiating
feature between the 2 groups is that the Cricamolan dialect is slower and
more nasal than that of the Valiente. In addition, the vocabulary of the Crica-
molans is more traditional and closer to that of the Sabanero Guaymi in
Chiriqui than to that of the Valiente. The subsistence patterns of the Crica-
molans are also similar to those of other tropical forest peoples of Central
America, while the Valiente depend more heavily on marine resources.
The area of Bocas del Toro is a tropical rain forest which, according
to the National Census of Panama, receives more than 150 inches of rainfall
annually. The land is covered with dense trees and undergrowth (Fig. 5). Ex-
cept for swampy lowlands, the area consists of innumerable steep hills inter-
laced with many creeks and short rivers which constantly flow off the humid
land. Although there are a few trails which go from the individual family
hamlets of the Valiente Guaymi to their farms, the cayuco, or dugout canoe,
is their principal means of transportation.
The average cayuco is from 8 to 14 feet long and will last from 5 to 15
or more years, depending on the type of wood and the care which it receives.
These smaller cayucos are for daily use and carry from one to five persons;
however, larger cayucos up to about 35 feet in length are also constructed for
use as coastal work boats. Most cayucos are paddled; although, the larger
ones are frequently equipped with a latim-like sail. Less frequently, but of
increasing importance, is the alteration of the stern to accommodate an out-
A GUAYMI JUNTA
board motor. This is accomplished by cutting off the back end and nailing on
a water-tight fitted stern board.
Making of a cayuco is a skill which every boy learns as he grows up;
however, under the pressure of acculturation this skill is slowly disappear-
ing and being assumed by specialists. Cayuco makers are contracted by in-
dividuals to make cayucos of specified wood and size. The specialist is paid
for his labor by the contracting individual, who usually earns the money by
working for the United Fruit Company, by raising and selling surplus farm
produce or domesticates, or by having an exceptionally good season hunting
green and hawksbill sea turtles and/or diving for spiny lobsters to sell to the
merchants in Bocas del Toro.
Each family, and individuals within each family, have lands designated
for their own use. The trees from which cayucos are made usually come
from these lands; although, trades do occasionally occur between individuals.
Trees from unclaimed lands may be used also. The process of making a
cayuco begins by selecting a suitable tree and marking it with machete or ax
cuts so that others will know that it has been claimed. Later the individual
fells the tree and lets it dry for 1-2 months. A cayuco made from a freshly
cut, undried tree is apt to crack from drying unevenly.
One to 4 or more months are spent in roughing out the cayuco from the
felled tree. If the person can afford it, small juntas are often organized to
finish this work more rapidly. If the maker takes too long in finishing his
cayuco, termites and wood rot are likely to set in and cause all of his labor
to be wasted. When the roughed-out form of the cayuco is completed, a ca-
yuco hauling junta will be organized to bring it to the cayuco maker' s or
owner' s house where work on it can be completed more easily.
As Erasmus (1956) and Young (1968) point out, the junta is the Latin
American word for a form of reciprocal aid in which an individual invites
various members of the community to assist him. Any adult male can or-
ganize a junta when he feels that it is necessary; however, ideally he is ex-
pected to repay each non-family participant with a day' s labor in exchange.
Juntas are usually announced about a week in advance to avoid conflicts.
The organizer provides food and refreshments for workers while they are
helping him. If the organizer wants a good turn-out, meat is ideally included
in the meal; however, he often tries to make up for its lack by providing
quantities of other foodstuffs. In addition, workers each receive a food bundle
to carry home to their families. This latter is not pay, but rather compensa-
tion for time lost in subsistence activities. Although rum or huarapo (pro-
nounced wah. rah. pi, home-made fermented sugar can juice) ensures a greater
turnout, I did not find it to be as indispensable among the Valiente Guaymi as
Young (1968:259) found chicha de maiz (fermented corn drink) to be among the
Western Guaymi of the San Felix district. In fact, the drinking of rum and
huarapo was observed on only 2 occasions -- both of these in celebration after
completing land clearing juntas.
Erasmus (1956) sees a relationship between the gradual disappearance
of exchange labor and increased involvement in a cash economy, especially
that of wage labor. Young (1968:264) writes that the Ngawbe appear to rep-
resent the beginning stages of this general process of individualization. He
notes that, although the Ngawbe participate as wage laborers, exchange labor
is still general. The Guaymi in Bocas del Toro also appear to be in this tran-
sitional process; however, the junta, especially the cayuco hauling junta
among the Valiente Guaymi, may be a good example of an ethnographic surviv-
al which has been reinforced in the face of modernization.
The junta is still a very important and necessary part of Valiente Guay-
mi culture as it now exists. It is a way of working which engenders group
spirit and accomplishes work which would be very time consuming orwhich
the individual could not possibly do alone. Among the Valiente Guaymi the
junta is mainly used in house building (especially for thatching a roof), land
clearance, and cayuco hauling. Since nearly all of the larger trees near the
coast and rivers have been felled in order to make cayucos or sell to the saw
mills, individuals now have to go a mile or more inland to find suitable trees
for making large cayucos.
Mauricio, who was making his own cayuco, invited me to participate
in his cayuco hauling junta, which was scheduled to begin at sunrise on a
Saturday. Since I had already participated in 2 cayuco hauling juntas, I ac-
cepted his invitation in order to have a photographic record of what takes place
on such a junta. The events which occurred are typical. They are presented
here as an example of a cayuco hauling junta.
It was a clear, sunny morning on Saturday. Three of my neighbors and
myself paddled our cayuco to Mauricio' s house where we joined about 31 other
men, ages ranging from about 15 to 40 years old. Most of the participants were
either affinal or consanguinial relatives of Mauricio. We were each given 2
home-made bread rolls known as "Jimmy-Johnny cakes" and some black,
sugary coffee. Most of the coffee was served in old tin cans. After we had
eaten this breakfast and waited the customary time (about 30 minutes) for late
arrivals, Mauricio led the way to our cayucos.
Mauricio led our small caravan-like group as we paddled to the mouth
of a creek and entered it through a gap in the mangrove swamp. Soon, we
reached a point where it was too shallow to paddle any farther (Fig. 2). We
tied our cayucos to the bank and prepared to walk inland.
A GUAYMI JUNTA
Mauricio led the way inland. Sometimes we waded in a shallow creek
bed and at other times we walked along its bank. The land gradually became
more rugged and the creek bed altered between swift flowing rapids and deep
pools at the bases of water falls. Several of the men at the head of our group
began widening the trail with their machetes and axes. After almost an hour
of walking, we reached the cayuco (Fig. Z) on which Mauricio had been work-
ing during the last 4 months.
In its roughed out form Mauricio' s cayuco was about 30 feet long, 5
feet wide and 3-4 feet high. Its sides were still 5-6 inches thick and its bot-
tom was from 9-12 inches thick. It appeared sufficiently sturdy to withstand
the 1.5 mile or so that we would have to drag it to reach our cayucos.
While most of us stood around waiting, Mauricio directed 3 of our group
to begin widening the trail in preparation for the passage of the cayuco. In
order to provide hand-holds, he then requested several others to cut small
trees into prepared lengths which were nailed across the hull at 2-foot inter-
vals for the length of the cayuco (Fig. 3). We turned the cayuco on its side
while Mauricio rounded off the bottom of the front end with his ax so that it
would not snag while being dragged. Then short sticks were nailed across the
front end to which tow ropes were tied. In addition, a rope was fastened to
several of the poles in the back in order to stop the cayuco from falling over
some of the steep inclines and water falls along the trail. After this prepara-
tion, Mauricio called a rest break and passed around cigarettes and pipe to-
bacco. During this time the manner in which the cayuco was to be hauled and
the position for each person was decided. The strongest individuals were as-
signed to the front in order to lift the cayuco over low obstacles. Mauricio,
calling directions to the others, led the way.
We took our places and hauled forward, and after some maneuvering
by-passed our first obstacle, a large rock. Then the cayuco started sliding
down the slope until it came to a sudden stop against the opposite bank of the
creek. Using small trees which we cut for levers, we struggled a few minutes
before we were finally able to lift the front of the cayuco onto the bank (Fig. 4).
Exhausted, we rested for a few minutes to regain our strength.
The sky slowly became cloudy as we started hauling the cayuco forward
again. Since the initial delay had given our advance cutters time to clear a
wide trail down the muddy mountain slope, our progress continued unimpeded
for about 10 minutes as we entered the now wider creek bed. Suddenly, amid
frantic shouts from those in the front, everybody strained with all of their
might to stop the cayuco' s forward motion. Simultaneously, 3 of the men in
the back quickly played out the rope which had been prepared for such emer-
gencies and made several turns of the rope around a nearby tree. At the last
possible moment the cayuco came grudgingly to a stop with nearly a third of
Fig. 3. Preparation
of the cayuco for
Fig. 4. Straining to
lift the cayuco onto
the opposite side
of the creek.
Fig. 5. Lowering
the cayuco slowly
down the second
incline into a
deep pool of water.
A GUAYMI JUNTA
its length jutting out precariously over a steep rocky incline. With
streams of perspiration running down our bodies, partly as a result of the
tenseness of the moment, we relaxed in a state of complete exhaustion.
While we rested, everybody discussed the near mishap. After passing out
cigarettes and tobacco, Mauricio entered into a discussion with some of the
more experienced men. Together they surveyed the situation and decided
upon the best way to lower the cayuco safely down the steep incline.
Reaching a decision, we cautiously began to lower the cayuco down the
steep rocky slope. The safety ropes tied to the back were wrapped around
tree trunks and slowly played out. We reached a level area, hauled the cayuco
forward another 60 feet, and then began to descend another incline. We slowly
lowered the cayuco down this second incline (Fig. 4) into a deep pool of cold
crystal clear water. Pulling it forward, again we had to strain to lift its
front onto the surrounding bank. The steep inclines at this point smoothed into
impassable rocky rapids. We decided to go down the less hazardous sloping
hill side which bordered it.
It began to rain as we struggled to pull the cayuco onto the opposite bank.
The taut rope in the front yielded to the strain as the nails suddenly pulled
free. Most of the men involved let go of the rope; however, Chogoli, who was
the anchor man, was caught off balance and fell backwards into a rain filled
mud hole. There was a great deal of laughter as he was helped to his feet
again. We again rested while more levers were cut to aid our efforts, and
while the wood was renailed to the front of the cayuco. Chogoli went to the
deep pool to wash the mud off and soon returned as good natured as ever.
We succeeded in pulling the cayuco onto the bank and started forward
again. We progressed uneventfully to the bottom of the hill and then entered
the now wider, shallow creek bed, which we followed for some distance until
we came to a place where it fell away through a narrow gap between several
large boulders. At this point we were again confronted with difficulties. It
was physically impossible to follow the course of the stream bed any further.
The only route left was to cautiously proceed along the side of a steep hill
which sloped off into the ravine through which the creek ran. We had to lift
the front of the cayuco onto a three-foot high ledge in order to begin our as-
cent. This we did with much effort and then rested. Shortly, we started for-
ward again and struggled to move the cayuco another 35 feet up the hillside
until we reached the level of a well-worn trail. This trail, which led to some-
one' s farm, went along the edge of the slope. There was much discussion con-
cerning the possibility of the cayuco sliding off the trail into the ravine; how-
ever, we cautiously started to move forward again until we reached a narrow
part of the trail on the steeply sloping hillside.
At this point we were about half way to our destination. It had taken us
more than 2 hours to drag the cayuco slightly more than .75 miles. It was
here that Mauricio called a long rest period and passed around cigarettes
and tobacco. We were soon joined by Mauricio' s wife, daughter, and father.
They brought refreshments of cooked Quaker' s Oats (havena) which was
served as a thickened drink, passed around in used tin cans. As soon
as everyone had finished, the empty cans were collected and our 3 visitors
returned to Mauricio' s house to prepare for our arrival.
During the refreshment break many of us looked over the next 60 feet
of the trail which lay before us. We would be extremely lucky if we were to
pass safely along the trail without sliding down the hillside onto the rocks
below. While Mauricio gave instructions to be very careful, we cautiously
inched forward. After 20 feet one of our group, Victoriano, slipped on the
rain-slicked, red clay hillside; however, he safely regained his balance and
we inched forward again. After nearly 50 feet the front of the cayuco began
to slide down the hillside. Frantic efforts were made to stop it, but they
were useless. Fortunately, the cayuco was stopped by a rotten log or it
might have hit the boulders. It was undamaged, but the slope was too steep
to drag it up again. Even if we could, we would still have to pass the same
spot over which the cayuco had just fallen. We decided that the only thing to
do was to try to continue forward by placing logs across the rocks and then
lowering the cayuco into a deep pool which lay below them.
Men were sent nearby to cut some trees. They soon returned and we
put the logs in place on top of the boulders. We then pulled the cayuco back-
wards a little in order to lift the front onto the fallen log which had saved it
from hitting the boulders. This was accomplished with considerable effort.
We slid the cayuco slowly forward over the emplaced logs. After a near ac-
cident when a log slipped off the rocks, we reached the edge of the falls and
lowered the cayuco into the deep pool which lay below. We easily pulled it
forward to the opposite side of the pool where the stream flowed out. While
some of our group bailed out the cayuco, others cooled off in the pool. We
were out of the rough terrain and were now following the slightly descending
stream bed. Finally, we reached our cayucos and started for Mauricio' s
house, while he and 5 others paddled the new cayuco behind us.
At Mauricio's house we pulled the new cayuco onto some logs which had
been placed near the shore for that purpose. We then covered it with palm
fronds to protect it from the sun, which could dry it out too quickly and cause
cracks. This task completed, we went up to Mauricio' s house where we each
received food and a leaf-wrapped food bundle to carry home. It contained
boiled Manatee (sea cow) meat and fish along with boiled, unripened bananas,
root tubers, and rice. After receiving our gift of food, we paddled to our
homes. It was after 3 P.M., and so ended Mauricio' s cayuco hauling junta.
A GUAYMI JUNTA
The manatee meat served during Mauricio' s junta was unique, since it
is a rare animal in the Province of Bocas del Toro. On other juntas we re-
ceived boiled pig meat and/or that of the green turtle; however, boiled and
smoke-dried fish is the usual meat served. The starchy root tubers (cassava,
yame, yampi, dasheen, and otoi), rice and bananas are nearly always served
as the bulk of the meal.
Of the cayuco hauling juntas in which I participated, only Santiago' s was
larger than Mauricio' s. It took nearly 10 hours for 48 of us to drag Santiago's
cayuco more than 2 miles -- up and over a hill and through the swampy low-
lands. During Santiago' s junta, there were 2 refreshment breaks in which we
were served mitchula, a thick drink of boiled ripe bananas which have been
mashed to an applesauce-like consistency and mixed with milk extracted from
grated coconut meat. In addition, we were served boiled pig meat, root tubers,
bananas, and rice during our mid-day lunch break. Our evening meal and food
bundle was a larger version of the lunch.
The events involved in a Valiente Guaymi cayuco hauling junta are pre-
sented above. In summary, the organizer invites various individuals to par-
ticipate in his junta from 3 days to a week inadvance. He either sends some
male member of his family or goes himself to the houses of those he wants to
invite to see if they can participate. He can generally expect only a 50 to 75
percent turnout of those individuals who indicate they will attend, since many
individuals do not like to say no and to indicate why. The percentage of invi-
tees who participate is in direct relation to the quantity and meat content of the
food provided. The organizer spends 3-7 days before the junta to provide the
surplus food used during the junta; in this he is helped by members of his fam-
ily (usually members of his immediate household) whom he in turn helps when
they have a junta. Female members of his household, in addition to other fe-
male family members, spend all day in preparing the food consumed during
The morning of the junta the participants usually gather at the house of
the organizer where they are given coffee, or chocolate, and "Jimmy-Johnny
cakes. This time is spent in general comradery and preparation for the
junta, axes and machetes are sharpened, and a festival-like atmosphere is
maintained. When everybody is present and has eaten, the organizer leads
the group to the cayuco. At this point final preparations take place and the
decision is made as to what each participant will do during the junta. The
organizer is recognized as being in command of the junta; although, his com-
mands are not obligatory. During the junta, when problems occur, members
of the group help decide how to solve them. Rest breaks are called when needed.
If the junta lasts for any length of time, refreshments are brought out to the
workers. This usually occurs about mid-way through the junta. After the
cayuco has been dragged to its destination, the participants return to the
organizer's home where they are fed, and afterwards given a food bundle to
carry home to their families when they leave.
In this society where there are no roads and the cayuco is the principle
means of transportation, an increasing Valiente Guaymi population will have
to decide whether their trees are to be sold to the saw mills or saved to make
cayucos when needed. The beginning of this conflict, instant money versus
long term needs, was observed while I was working among the Valiente Guay-
mi. Related to this is the fact that having to go farther and farther inland to
find suitable trees for the making of large cayucos may be reinforcing the
need for large, well-organized cayuco hauling juntas. If so, this would be
an interesting case of acculturation stimulating greater co-operation in a
Reverte (1967), in reviewing historical documents concerning the 4th
voyage of Columbus in 1502 and forays of later explorers and missionaries in
the area, reports on the occurrence of very, large cayucos on several occa-
sions; however, no mention is made concerning cayuco hauling juntas. Rev.
Ephraim Alphonse (1956) who lived in the area from 1917 to 1938, does not
even describe the junta in any of its forms. The proposition that large, well-
organized juntas are relatively new, or at least more frequent in recent times,
remains unsubstantiated in the literature. However, I recorded ethnographic
information which indicates that both the capacity to construct cayucos and
the native ability to manually transport them to the water during juntas has
been greatly facilitated by the selected acceptance of western technology in,
at least, the last three or four generations.
Metal tools such as axes, adzes and machetes have replaced their stone
predecessors so long ago that the former are not even remembered. Men 70 -
80 years of age do remember, however, felling large trees with the aid of a
fire built around their bases. Augers are presently used to drill holes (which
are later plugged) in various places in the cayuco hull during construction to
facilitate cutting the sides to a uniform thickness. The cross-bars used to
facilitate dragging the cayuco, although formerly tied to the hull with vines
looped through these holes, are presently nailed down and have a greater
carrying capacity. Stronger commercial ropes have nearly completely re-
placed the use of home-made ropes and vines. The partial reliance on a cash
economy in other aspects of Guaymi lifeways, along with improved marine
hunting and fishing capacity by the borrowing of other non-indigenous mate-
rials, has allowed individuals to provide large quantities of surplus food for
use in juntas, and hence support more participants. In view of more than 4
centuries of efforts to westernize the Guaymi, the junta can be considered an
important aspect in the maintenance of the structural continuity of Guaymi
Society, and as an ethnographic survival which has been reinforced in the
face of acculturation.
A GUAYMI JUNTA
1956 Guaymi grammar and dictionary, with some ethnological notes.
Bureau of American Ethnology, Bulletin 162. Washington.
Erasmus, Charles J.
1956 Cultural structure and process: the occurrence and disappear-
ance of reciprocal farm labor. Southwestern Journal of
Anthropology, Vol. 12, pp. 444-469.
1948 Caribbean lowland tribes: the Talamanca divisions. In Hand-
book of South American Indians, Julian H. Steward, ed. Vol. 4,
pp. 231-252. Bureau of American Ethnology, Bulletin 143.
Panama, Republica de
1970 Censos Nacionales de 1970 (Preliminary releases).
Reverte, Jose M.
1967 Los Indios Teribes de Panama. Panama.
Tesar, Louis Daniel
1968-71 Unpublished field notes, tapes and photos.
1971 Two and one half years with the Valiente Guaymi: an ethno-
graphy (Part One). Copy of manuscript on file Department of
Anthropology, Florida State University, Tallahassee.
Tyson, Edwin C.
1972 Personal communication.
Young, Philip D.
1968 The Ngawbe: an analysis of the economy and social structure
of the western Guaymi of Panama. University of Illinois, Ph.D.
1970 Notes on the ethnohistorical evidence for structural continuity
in Guaymi society. Ethnohistory, Vol. 17, Nos. 1-2, pp. 11-29.
March 1, 1973
LA ALVINA DE PARITA: A PALEO-INDIAN CAMP IN PANAMA
D.L. Crusoe and J.H. Felton
In 1967 we undertook an archeological survey of scattered localities
throughout the Republic of Panama. In the course of this survey, new sites
were found and a number of older and well known sites were revisited. One
such site was the La Alvina de Parita (HeF 108) site which may correspond
to Willey and McGimsey's (1954:110-114 La Mula site (He 30). The 1967
collection was composed of well over a thousand artifacts (Crusoe 1970).
The La Alvina de Parita site was revisited in 1968 because of the se-
nior author' s interest in the site's "horizontal" rather than "vertical"
stratigraphy. That is, there are small apparently short-term or seasonal
living floors which are isolated in space with sometimes 20-30 meters of
sterile ground separating each one (Fig. 1). There are perhaps several hun-
dred of these floors since the site stretches some 400 to 500 meters along
and at least 100 meters behind and in front of an undulating 10 meter contour
line. The 10-meter line, referred to in the caption of Figure 1, is an old
Pleistocene beach which is situated in front of a silted-in lagoon. Today, only
a few trees stand on the site which is heavily erosioned. The amount of water
run-off which occurs with a normal dry-season shower is phenomenal--the
place is literally flooded within fifteen minutes of the onset of the shower.
Midway across the site, and in the center of the second undulation, the
junior author discovered an "hour-glass" shaped hearth (Fig. 2). Within and
without the hearth were stone artifacts and several ceramic fragments appar-
ently of the Sarigua phase (ca. 1500 B. C. 500 B. C.). There was no ev-
idence of any other cultural deposit in the area of the hearth. We set out a
5-by 5- foot square about the hearth. Then we collected materials from within
and without the hearth which were sacked accordingly.
The "kicked-up" walls of the hearth (Fig. 2) stood 26 cm. above the ex-
terior surface of the ground. The interior depth of the hearth, before excava-
tion, ranged from 14 cm. in the northeastern compartment to 12 cm. along the
north wall of the main compartment below a level line which approximated the
upper edges of the structure. The southern wall of the main compartment had
been almost completely eroded by water action; in the hearth' s southern area,
the surface of the surrounding ground approximated that of the hearth.
The area within the test square, but external to the hearth was first ex-
cavated. No artifacts were unearthed. It was found that the hearth rested upon
sterile red clay--the same material used in its construction. The hearth' s in-
Florida Anthropologist, vol. 27, no. 4, December 1974
LA ALVINA, PANAMA
^ 4 *' -' -i, -
Fig. 1. Small living floor in La Alvina de Parita area
with 10-meter contour line in background.
4' .~ -
Fig. 2. Paleo-
ment in fore-
CRUSOE AND FELTON
terior was next cleaned. There were numerous egg-sized chunks of charcoal
excavated from the hearth's floor. These were sealed by 15 cm. of ashy, but
otherwise sterile, fill. It is possible to attribute no more than 2 cm. of the
fill to wash. Upon completion of the excavation of the hearth' s interior com-
partments, the floor of the hearth was found to lie between 30-33 cm. below
the level line.
The carbon sample (FSU 300) yielded a date of 9350250 B. C. (the av-
erage of two runs on the same sample. This date was both exciting as well as
alarming. We were excited by the early date but alarmed because it had been
thought that our date was associated with the Sarigua materials recovered in
the collections. If the carbon was associated with the artifacts then the date
should have fallen in the mid portion of the first millennium B. C.; it did not.
The runs of the sample were rechecked by the carbon laboratory and nothing
was found to be abnormal. A third visit to the site was necessary to recheck
Little remained of the hearth in mid-March of 1972. The remaining por-
tion of the wall of the northeastern compartment was removed in hopes that
some charcoal could be extracted by floatation to obtain additional carbon
dates. Floatation of this material produced negative results.
Close examination of similar areas along the undulating 10-meter con-
tour revealed the existence of at least two other burned places. These, like
the one reported above, were almost completely destroyed by erosion. No
cultural materials were found to be in association with these fire-hardened
clay features. Also there was no charcoal. In short, these other "hearths"
could date from almost any period, but their position relative to that of HeF
108.3 would have made them good candidates for a Paleo-Indian camp site.
Our interpretation of the Sarigua materials associated with the hearth
is simply that they are the result of migration due to the velocity of the water
run-off. That is, the hearth is located about 5 meters below the 10 meter con-
tour, and it is situated more or less centrally in a "U" shaped section of the
undulating contour. It would appear that with each rain, artifacts from the
Sarigua occupation located upon the top of the 10 meter contour were washed
down the slope, being captured by the fire-hardened walls of the hearth. It is
our opinion that the carbon date actually dates the usage of the hearth. As far
as the two-compartment structure of the hearth is concerned, this might rep-
resent usage on two separate occasions. But, in our investigation we were un-
able to find any factor which would hint at the separate construction of the com-
While it is difficult to give an estimate of the number of days that the
hearth was used, we feel that judging from the degree of fire discoloration and
CRUSOE AND FELTON
the hardness of the hearth' s clay walls, it would appear that it was probably
repeatedly used over a short period of time, perhaps less than one week.
Our carbon date has one further implication. Often it has been stated
that the 10-meter contour line represents an old shoreline whose fore-shore
has been silted in. If the contour does represent such a coastline, then the
position of the hearth with regard to the date and the silting-in hypothesis
would necessitate a late Pleistocene time (ca. 20,000 B. C. ) when the waters
of the bay washed this old beach.
Although no artifacts were recovered from the La Alvina de Parita
hearth, the structure of the hearth, and the possibility that other early hearths
will be uncovered by erosion makes the site important in Central American
Archeology. Our carbon date is the earliest recorded in Isthmian Central
America and provides evidence of Man' s presence in the area about eleven
thousand years before the present.
1970 Manuscript submitted to the National Museum of Panama.
Willey, G.R.. and C.R. McGimsey III.
1954 Peabody Museum of American Archeology and Ethnology,
vol. 49, no. 2. Cambridge.
WERE THERE PRE-COLUMBIAN CULTURAL CONTACTS BETWEEN
FLORIDA AND THE WEST INDIES: THE ARCHAEOLOGICAL EVIDENCE
Ripley P. Bullen
Students from Fewkes (1914) to Sturtevant (1960), including Love'n,
Gower, Stone, Willey, Rouse, and John W. Griffin, have discussed possible
prehistoric cultural contacts between Florida and the Antilles. Until recently
enough detailed chronological and typological information has not been avail-
able in both areas for a satisfactory resolution of this problem. The most re-
cent and thorough study, that made by Sturtevant, was based on ethnological
evidence. He (1960:43) found no traits requiring transportation by human
The excavation in Cuba of ceramic sites earlier than those previously
known (Tabio and Guarch 1966) and Cruxent' s rather early (ca. 2600 B. C.)
C-14 dates for preceramic Hispaniola (Cruxent and Rouse 1969)--as well as
the uncovering of a dugout canoe in Florida, C-14 date to 1000 B.C. (Bullen
and Brooks 1967)--have reopened the question of Floridian-Antillean com-
munication. I will discuss the preceramic picture for both areas and sketch--
with appropriate C-14 dates--the two ceramic sequences.
Examination of appropriate maps will indicate that the southern part of
Florida is near the western end of Cuba, while central and eastern Cuba, His-
paniola and Puerto Rico, and the Virgin Islands extend easterly a substantial
distance. The comparative student should realize that Paleo-Indian or Meso-
Indian (preceramic Archaic) artifacts are not known for extreme southern
Florida, the area nearest Cuba. They are not found south of Charlotte Harbor
on the west coast or very much south of Vero Beach on the east coast of Flor-
ida. I should also mention at this time that the earliest pottery on the west
coast--Orange Plain and Incised (fiber-tempered) and C-14 dated at 1450 B. C.,
perhaps much earlier (E. Ross Morrell, personal communication)--has been
found at Marco Island, a little south of Naples, while on the east coast the
earliest pottery comes from near Hollywood, a little north of Miami. The lat-
ter is semi-fiber-tempered and called Norwood Plain (Wilma Williams, per-
sonal communication) and carries with it a date of 1000 B.C. Otherwise, in
extreme south Florida and the Keys, which extend southwesterly towards
western Cuba, the earliest present evidence of man consists of sand-tempered
Glades Series pottery dated perhaps as early as A.D. 300-400. The earliest
Glades Series radiocarbon date refers to the introduction or development of the
first decorated types, Fort Drum Punctated and Incised, at A.D. 500 (John W.
Griffin, personal communication).
Florida Anthropologist, vol. 27, no. 4, December 1974
FLORIDA AND THE WEST INDIES
A somewhat analogous situation exists in Cuba. While preceramic
remains are known, excepting Jamaica, for all the Greater Antilles in-
cluding St. Thomas in the Virgin Islands and Antigua to the east, pre-
historic pottery of any kind is not found in the western half of Cuba, the
region nearest Florida. These demographic situations do not favor com-
munication between Florida and the Antilles during the preceramic or
early ceramic periods.
Preceramic traits of the Greater Antilles include prismatic blades
and percussion flaked tanged points from Hispaniola and Cuba (Rouse 1941:
Pls. 1-3), edge grinders and hammerstones from Puerto Rico (Alegria,
Nicholson, and Willey 1955), and, for Puerto Rico and St. Thomas, work-
shops where stone axes were made by chipping, pecking, and grinding
(Bullen and Sleight 1964). The last have no counterpart in Florida, prob-
ably because volcanic rock is not available there. Both the edge grinders
of Puerto Rico and the tanged points of Hispaniola can be traced back to
south Central America (Alegria et al 1955, Coe 1957, Bullen 1964a) .
Specimens like the large tanged points and blades of Cuba and Hispan-
iola(Figs.l,A;4,A) are not typical of Florida where a "blade industry" as
such is not found. A few prismatic blades are known but rarely do they have
the symmetric rectangular outlines illustrated for Hispaniola. Much rarer yet
are prismatic blades exhibiting evidence of use. On the other hand, Florida
has thousands and thousands of amorphous shaped flakes of which a great
many have a scraper-like retouch or exhibit evidence of use. While some
projectile points, particularly during the late preceramic Archaic and last
prehistoric ceramic periods, are made from concavo-convex flakes with a
minimum amount of chipping on the flatish concave side, Florida points and
knives from the 10,000 year old Clovis Fluted to the small Pinellas triangular
points found by the early explorers are bifaces (Fig. 2, left side). In brief,
the prismatic flakes and tanged knives found in Cuba and Hispaniola are not
found in Florida while Floridian points and knives--so common on the penin-
sula--are unknown in the Antilles.
The amorphous knives and scrapers--or fragments of them--illustrated
by Cruxant and Rouse (1969:50-51, a-c, f-h) for the Rancho Casimira site in
the Dominican Republic can be found at any sizeable Archaic workshop in
Florida. They are also found in Mexico and Central America, and to my mind,
are too generalized to demonstrate any cultural communication with Florida.
I should probably also mention here that the ten bifacial projectile points re-
putedly found near Old Harbor, Jamaica, by a sailor (Lov4n 1935*P1.13), innocently or
otherwise, represent a plant. Judging from the illustrations these are good
typical Florida specimens belonging to three known archaeological periods,
but they are certainly out of context in Jamaica where an extensive site survey
and excavation program during the last five to seven years has failed to pro-
duce any similar objects (Ronald L. Vanderwal, personal communication;
James W. Lee, Mandeville: Archaeology Jamaica). Equally anomalous is
a stemmed point that recently turned up on the disturbed surface of a pre-
ceramic site on Antigua (Desmond Nicholson, personal communication). This
point is unquestionably made of Florida chert and exhibits Florida workman-
ship but resembles nothing else found on that island in either material or
One more stone object needs mention. It is a typical Antillean t-shaped
ground stone ax (FSM-3535), reported plowed up at the Prairie Creek midden
near Gainesville, Florida, many years ago and given to the Florida State Mu-
seum (Goggin and Rouse 1948). Documentation of this specimen, like that of
the projectile points from Jamaica, leaves something to be desired. Since its
"discovery" a substantial amount of archaeological work has been done in
Florida while vast industrial and real estate developments have been made with-
out the discovery of another example. Even if it came to Florida in prehistoric
times--by trade, raid, or in a storm transported canoe--it certainly had no
cultural influence on Florida aborigines. At most, it might have been an item
Turning to the ceramic periods, I have mentioned the absence of any
pottery in western Cuba. For diffusion to have occurred in either direction,
with such a substantial gap in between, is extremely unlikely. At least it
must have been very infrequent. However, it is necessary for us to example
the ceramic typologies and sequences of the two areas.
For the purposes of this paper, the ceramic history of the Antilles may
be divided into three major periods. The first covers the diffusion of Saladoid
ceramic types from Venezuela northward through the Lesser Antilles into
Puerto Rico. This dispersal started before the time of Christ and reached
Puerto Rico before A. D. 120 (Rouse 1964).
This pottery is characterized by geometric white-on-red painting, fine
crosshatched incision bounded by wider incised lines, simple or complicated
button-like rim lugs or adornos, a generalized inverted bell-like shape; and
good-sized strap handles frequently "peg-topped" (Fig. 3, D). None of these
traits--nor any pottery--are found west of Paerto Rico and extreme eastern
Hispaniola until the next ceramic period. It is probably unneccessary to state
that nothing remotely resembling Saladoid ceramics has ever been found in
Florida (not even in the Keys) while pottery making was fifteen hundred to two
thousand years old before the art was introduced into Puerto Rico.
The second major period is referred to by Rouse (1952) as "Ostiones."
Pottery was fairly thin and well made with, in general, a smooth overall red-
painted surface. Rim horns or lugs, simple side applique, and boat-shaped
FLORIDA AND THE WEST INDIES
Fig. 1. Artifacts from Hispaniola. A, preceramic stone daggers and flake knives;
B, sherds of Meillac vessels; C, punctated container and Carrier type rim lugs.
A r c h a i c
8 I a n
Glodesm GlodesII Gldes I
St Johns Fine Cord
Check Stomoed Morked
Weeden Island Swift Creek Deptfor d
Cleor Fork gouge Arredondo
S Transitional Period
Doton Swonnge Perid
D Ion Suwonne
Fig. 2. Chipped stone tools and typical sherds from Florida.
^ inches b
,<'A/^ t^^^ A
4% r t 'k
S-ee 99 ii
bb CC dd I f h
Fig. 3. Selected artifacts from Florida and the Antilles. A, Florida
Keys; B, southeast Florida (Miami); C, Cayo Ocampo, eastern Cuba;
D, Pearls site, Grenada, represents earliest pottery in Puerto Rico.
T'ig. 4. Specimens from eastern Cuba. A-C,
from Arroyo del Palo; D, additional examples
from the Mayari culture; E, artifacts of the
Taino or final phase.
FLORIDA AND THE WEST INDIES
containers were present but typical Saladoid decoration had disappeared.
However, a few plain rim and side lugs clearly reflect Saladoid antecedents.
Utilitarian vessels, which are not so well finished, sometimes exhibit broad
scraping marks. Chronologically, this period started around A. D. 700 and
lasted until nearly A. D. 900 (Rouse 1964; Bullen 1964a; Hoffman 1970:18).
Ostiones including probably the early part of the succeeding Santa Elena
period was a period of expansion and, I presume, one of population explosion.
Colonies were established in Jamaica (collections of the Institute of Jamaica,
Howard 1965) and the Bahamas (Hoffman 1970:102-5); and pottery making in-
troduced into Hispaniola and eastern Cuba. Ostionoid traits are definitely pre-
sent in the Sub-Taino ceramics of eastern Cuba (Tabio and Rey 1966:250).
Ostiones pottery is a major component in lower levels at Magans Bay on St.
Thomas, on Antigua, and identical smooth, red-painted, pottery with typical
rim lugs and double horns are identifiable as far south as the Westerhall site
on Grenada in the Lesser Antilles (Bullen 1964b:Pl. XVII). In spite of this great
areal distribution, Ostiones ceramics are unknown in Florida.
The last major period includes Carrier and Meillac (Fig. 1, taken from
Rouse 1941) of Hispaniola (Rouse 1941), Capa, and Esperanza of Puerto Rico
(Rouse 1952), and the incised and modeled Taino pottery (Fig. 4, E) of east-
ern Cuba (Tabio and Rey 1966:253). In a modified, less developed, or simpli-
fied form, pottery of this period is also found in the Virgin Islands and traces
are evident in the Bahamas and on Jamaica.
In Hispaniola this ceramic stage is characterized by coarse incised
crosshatching, parallel incised straight lines with terminal dots, punctations
below lips, and simple--usually notched--applique forming Rouse's (1941:
Pls. 7-23) Meillac style and wide incised flowing designs on casuela vessels
above the shoulders and fairly baroque rim adornos many of which exhibit a
hooded or masked face (Fig. 1, C) forming Rouse's (1941:Pls. 27-34) Car-
rier style. Adornos of this type are unknown to Florida archaeology. For-
getting for the moment some of the Meillac incised designs, these styles
have no correspondence in Florida ceramics.
My ordering of the ceramics of eastern Cuba, the area nearest Flor-
ida, would place the Ostiones derived Sub-Taino ceramics first (ca. A. D.
700-900), the predominantly straight line incised Mayan pottery (Fig. 4,
B-D) second (ca. A.D. 850-1200), and the ornate Taino vessels (Fig. 4, E)
last (ca. A.D. 1200-1500) (illustration taken from Tabio and Guarch 1966 and
Tabio and Rey 1966). This is a little different than the sequence proposed by
Tabio and Rey (1966:10) but it correlates well with Rouse's Ostiones-Meillac-
Carrier sequential arrangement on Hispaniola (Rouse 1964). As implied above,
there seems to be no obvious ceramic similarities between Sub-Taino and
Taino pottery and that found in Florida.
For our purposes Mayyri ceramics are best exemplified at Arroyo del
Palo (Fig. 4, B-C) and Mejias (Fig. 4, D) in Mayarf and Aguas Gordas in
Banes, Cuba, where they are radiocarbon dated A.D. 930 to 980 with anupper
date for Arroyo del Palo of A. D. 1190 (Tabio and Rey 1966). Decoration, con-
sisting of incised straight lines in some cases with terminal dots, appears to
be a simplification of the Meillac type of decoration found on nearby Hispaniola.
That it represents a dispersal from there, relatively early in the development
of that style, seems logical. The alternative, that this type of decoration was
introduced into eastern Cuba and then transferred to Hispaniola to there devel-
oping to the Meillac style seems too remote for serious consideration. The
date (A. D. 900), is much too late for diffusion from any American Formative
site (where straight line incision would be found) while a stylistically accept-
able donor culture of the correct time period is unknown.
It should be mentioned in this connection, however, that bands of par-
allel incised straight lines, incised crosshatching, and bands of punctations
do occur as Carrabelle Incised and Punctated and Keith Incised pottery types
in Florida during the Weeden Island period (Willey 1949), approximately
A. D. 500 to 1300 (Bullen t965:306), on the northwest coast of Florida.Also, ter-
minal dots are found at the ends of curvilinear lines--sometimes short straight
lines--as parts of larger designs in other contemporaneous types, Weeden
Island Zoned Red (Bullen 1965:Fig. 4, lower left), Weeden Island Incised,
and Papys Bayou Incised (Willey 1949), but never--as far as my memory
goes--at the ends of rows or bands of parallel straight lines. None of these
types are ever found on the west coast of Florida south of Charlotte Harbor,
on the southeast coast, or in extreme south Florida and the Keys--the re-
gions nearest Cuba and Hispaniola.
In Florida the earliest pottery, belonging to the Orange period of the
American Formative and radiocarbon dated between 1750 and 1000 B.C. is
tempered with fibrous material and decorated with incised straight lines
forming reasonably simple designs (Fig. 2, bottom). After 1000 B. C.,
these incised designs were carried over onto limestone-tempered, sand-
tempered, and chalky (temperless) pastes and diffused over all of peninsular
Florida except extreme south Florida and the Keys. The temporal span of
this decoration, in spite of its resemblence to that found on Meillac vessels,
is entirely too early to permit diffusion to Cuba or Hispaniola as it went out
of style completely before 500 B. C.
There are no similarities between the stamped and incised potteries of
the succeeding Deptford, Weeden Island, and Safety Harbor periods of the
Florida Gulf coast or the check-stamped chalky ware of the east coast and
those of the Greater Antilles except for the Carrabelle and Keith Incised types
mentioned earlier. Their areas of distributionare separated from eastern Cuba
and Hispaniola by ceramic-less western Cuba, as was stated earlier, and also
FLORIDA AND THE WEST INDIES
by the ceramic Florida Glades and Keys areas where different pottery types
are found at these times.
The ceramic picture of the extreme southern part of Florida differs en-
tirely from that of the rest of the State (Bullen 1965). Here the Glades ceramic
tradition begins around A. D. 300-400 with sand-tempered simple bowls fre-
quently having inturned curving rims. Decoration when it first appears A. D.
500 (J.W. Griffin, personal communication), consists of simple diagonal lines
or narrow bands composed of short incised parallel lines or punctations or,
alternatively, simple straight lines bordered by ticks or short gashes. Sim-
ilar decoration is not known for Cuba.
Later, long incised arcades (Key Largo Incised; Fig. 2, Glades II,
middle) became the most popular decorated type. Key Largo Incised sherds
(Fig. 3, g-h), dated by trade sherds (St. Johns Check Stamped; Laxson1953)
to after A. D. 800 (Bullen and Sleight 1960) and by radiocarbon dating of Glades
sites in the Everglades National Park to an A. D. 805-830 (J.W. Griffin, per-
sonal communication), should be found in eastern Cuba or Hispaniola if any
communication between these regions and Florida occurred substantially be-
fore the time of the coming of Columbus. Similar specimens are not illus-
trated in the available Cuban literature but Rouse (1949:130), citing Morales
et al (1947), points out the presence of fairly similar incised arches below a
thickened rim on pottery from Cayo Ocampo, near Cienfuegos, Cuba (Fig. 3, C,
k-l). The similarity is not 100 per cent because of the thickened lip and the
fact the arches do not form arcades as in Florida, but it is suggestive.
Subsequently, Laxson (Bullen and Laxson 1954) uncovered another in-
cised type which bears a simple "zigzag" design which, perhaps significantly,
is also found at Cayo Ocampo. These similarities may be seen in Figure 3
where c-f show the Florida examples and i-j those from Cuba. As men-
tioned above for the curved designs, Florida zigzag incised sherds do not
have thickened lips like the Cuban ones but "d" does have an incised line at
the top giving an appearance very similar to that of the Cuban sherds.
These two suggestions of culture contact are tenuous but a little persua-
sive, especially in view of the fact that Cayo Ocampo is the "westernmost
ceramic site known to" Rouse (1949:130). If acceptable as such, they imply
that contact was made on a fairly simple level technologically or, perhaps
one should say, aesthetically. This is what might be expected if the contacts
were made by fishermen who might not have highly decorated pottery--at
least not in their canoes.
The final stage of Glades ceramics, Glades III, is typified by notched,
indented, or pinched lips on otherwise undecorated simple bowls (Fig. 2,
Glades III). This is a far cry from the overly ornate decoration found in
Hispaniola in late pre-Columbian times. It also differs considerably from
pottery in use in other parts of Florida at the same time. Such notches are
not known for the Greater Antilles although they may resemble faintly the
"finger-indented" rims of Suazey ceramics found in the Lesser Antilles after
A.D. 1200 (Bullen and Bullen 1972:153). Finger indented rims correlate with
Carib occupation of the Lesser Antilles but such rims are not known north-
west of the island of Sint Maarten.
One special Florida pottery type, apparently belonging to this period,
should be mentioned. Limited to the Florida Keys, the illustrated example
(Fig. 3, A) came from Howe Key near Key West but examples are also
known for Upper Matecumbe Key (Goggin and Sommer 1949:P1. 3, B-C).
Vessels have wide, somewhat triangular shaped, tabular handles which end
in a slightly incurving edge set off by lip notches from the rest of the lip re-
gion. Two or three parallel incised lines parallel lip edges. These sherds
remind me more of the Greater Antilles than they do of Florida, but dupli-
cate examples cannot be found in the Antilles. The deep lip notches, and
even the incised parallel lines, are a little suggestive of a sherd (Fig. 4, E,
middle of upper row) illustrated by Tabio and Rey (1966) for the Taino cul-
ture of eastern Cuba. At Howe Key the notched lip sherd was associated with
another non-Florida incised and punctated sherd (Fig. 3, B) which, however,
had a Florida Glades type of indented lip.
Except for the ceramic traits mentioned in the last few paragraphs,
large crude chert scrapers, and shell gouges, there are no typological
similarities in archaeological specimens from either the preceramic or ce-
ramic periods of Florida and of the West Indies. The lack of pottery in
the western half of Cuba and of preceramic materials in extreme southern
Florida also argue against any cultural connections between these regions.
In agreement with the proposition of no pre-Columbian contacts, the pre-
ceramic artifacts of the Greater Antilles can be traced back to Central
America and the pottery to South America (Bullen 1964a, 1964c). No Flor-
ida origins are evident. The similarity in shell gouges, noted by various stu-
dents, are probably a thousand or more years apart in time and easily ex-
plainable on the basis of similarity in raw material used for the same func-
tion. The structual peculiarities of shells limit the possible typological vari-
ation of such common tools as hammers, gouges, and celts. In Florida these
shell gouges are found in the St. Johns River region associated with well-
made shell celts (made from the lips of Strombus giges shells) in preceramic
deposits dating about 3000-2000 B. C. (Bullen 1955). In Cuba and the other
nearby islands shell celts are not found with the shell gouges as might be ex-
pected if the latter diffused from Florida.
Very late in prehistoric times, some form of ball game with its atten-
dant ceremonial objects (ball courts, stone yolks, etc.) was introduced into the
FLORIDA AND THE WEST INDIES
Greater Antilles and St. Croix from southeastern Mexico. Its lateness is
demonstrated by the shallow provenience of fragments of stone yolks at the
Salt River site on St. Croix (Bullen 1965c:438). Neither possible ball courts
or such ceremonial objects are reported for Florida.
In connection with this importation there must have been a consider-
able amount of water transportation. It occurred at about the same time as
the presumed date of manufacture of the few pottery types from the Florida
Keys and the Miami area which have been mentioned earlier as having poss-
ible Cuban affinities. I conclude that there was no pre-Columbian cultural
trait diffusion between Florida and the Greater Antilles except possibly very
late, ca. A. D. 1200-1400, when occasional contacts may have occurred be-
tween Cuba and the Florida Keys. One reason is the effectiveness of the Gulf
Stream as a cultural barrier.
Alegria, Ricardo, H. B. Nicholson, and Gordon R. Willey
1955 The Archaic tradition in Puerto Rico. American Antiquity,
vol. 21, pp. 113-21.
Bullen, Ripley P.
1955 Stratigraphic tests at Bluffton, Volusia County, Florida.
Florida Anthropologist, vol. 3, no. 1, pp. 1-16. Gainesville.
1964a The preceramic Krum Bay site, and its importance in the
peopling of the West Indies. Proceedings of the 34th Inter-
national Congress of Americanists, pp. 398-403. Vienna.
1964b The archaeology of Grenada, West Indies. Contributions of the
Florida State Museum, Social Sciences, no. 11. Gainesville.
1965a Archaeological chronology of Grenada. American Antiquity,
vol. 31, no. 2, pp. 237-41.
1965b Florida's Prehistory. Chapter XXIII in Florida from Indian
Trail to Space Age, vol. 1, by Tebeau et al. Delray Beach,
1965c The archaeology of Grenada, West Indies, and the spread of ce-
ramic people in the Antilles. Proceedings of the 36th Interna-
tional Congress to Americanists, vol. 1, pp. 535-39. Seville.
Bullen, Ripley P. and Harold K. Brooks
1967 Two ancient Florida dugout canoes. Quarterly Journal of the
Florida Academy of Sciences, vol. 30, no. 2, pp. 97-107.
Bullen, Ripley P. and Frederick W. Sleight
1960 Archaeological investigations of Green Mound, Florida. Wm.
L. Bryant Foundation, American Studies, no. 2. Orlando.
1964 The Krum Bay site: a preceramic site on St. Thomas, United
States Virgin Islands. Wm. L. Bryant Foundation, American
Studies, no. 5. Orlando.
Bullen, Ripley P., and Dan D. Laxson
1954 Some incised pottery from Cuba and Florida. Florida Anthro-
pologist, vol. 7, no. 1, pp. 22-23.
Coe, William R.
1957 A distinctive artifact common to Haiti and Central America.
American Antiquity, vol. 22, no. 3, pp. 280-82.
Cruxent, Jose M. and Irving Rouse
1969 Early man in the West Indies. Scientific American, vol. 221,
no. 5, pp. 42-52.
Fewkes, I. Walter
1914 Relations of aboriginal culture and environment in the Lesser
Antilles. Bulletin of the American Geographical Society, vol. 46,
no. 9, pp. 662-78.
Goggin, John M. and Irving Rouse
1948 A West Indian ax from Florida. American Antiquity, vol. 13,
no. 4, pp. 323-35.
Goggin, John M., and Frank H. Sommer III
1949 Excavations on Upper Matecumbe Key, Florida. Yale Univer-
sity Publications in Anthropology, no. 41.
1927 The northern and southern affiliations of Antilles culture.
American Anthropological Association Memoirs 35.
1943 The Antillean problem in Florida Archaeology. Florida His-
torical Quarterly, vol. 22, no. 2, pp. 86-91.
Hoffman, Charles H. Jr.
1970 The Palmetty Grove site, San Salvador, Bahamas. Contribu-
tions of the Florida State Museum, Social Sciences, no. 16.
Howard, Robert A.
1965 New perspectives on Jamaican archaeology, American An-
tiquity, vol. 31, no. 2, pt. 1, pp. 250-55.
FLORIDA AND THE WEST INDIES
Laxson, Dan D.
1953 Stratigraphy at a Hialeah midden. Florida Anthropologist,
vol. 6, no. 1, pp. 1-8.
1935 Origins of the Tainan Culture, West Indies. G'teborg.
Morales Patinos, Oswaldos, et al
1947 Cayo Ocampo: historic de un cayo. Revista de Arqueologia y
Ethnologia, 2 nd. ser., ano 2, nos. 4-5, pp. 55-123. Habana.
1941 Culture of the Ft. Liberte region, Haiti. Yale University
Publications in Anthropology, no. 24.
1949 The Southeast and the West Indies. In The Florida Indian and
his Neighbors (J.W. Griffin, ed.), pp. 117-37. Winter Park,
1952 Porto Rican prehistory. Scientific Survey of Puerto Rico and
the Virgin Islands, vol. 18, pt. 3, pp. 307-460. New York.
1958 Archaeological similarities between the Southeast and the West
Indies. Florida Anthropological Society Publications, No. 4,
1964 Prehistory in the West Indies. Science, vol. 144, no. 3617,
1939 The relationship of Florida archaeology to that of Middle Amer-
ica. Florida Historical Quarterly, vol. 17, no. 3, pp. 211-18.
Sturtevant, William C.
1960 The significance of ethnological similarities between south-
eastern North American and the Antilles. Yale University
Publications in Anthropology, no. 64.
Tabio, Ernesto, and Jose M. Guarch
1966 Prehistoria de Cuba. Department of Anthropologia,
Academia de Ciencias de Cuba. Habana.
Willey, Gordon R.
1949 The Florida Indian and his neighbors: a summary. In The
Florida Indian and his neighbors, pp. 139-167 (J.W.
Griffin, editor). Winter Park.
Florida State Museum
Gainesville, Florida, USA
March 20, 1971
Revised October, 1973
PIECES ESQUILLEES IN THE SOUTHEAST
William H. Wesley
In early 1972 a student in anthropology at the University of Arkansas
contacted several north Alabama chapters of the Alabama Archaeological So-
ciety and expressed an interest in any examples of unfinished fluted projectile
points for use in a study of such material. The artifact shown in Figure 1,
thought to be an example of a tool made from an unfinished fluted point was
therefore mailed to the University of Arkansas for examination. This arti-
fact is fluted on both faces and, although no intentional flaking appears on the
concave end, it does slightly resemble the shape of a Clovis or Cumberland
projectile point. The specimen was returned in March of 1972 with the com-
ment, "I have shown this example to several of our Dalton experts, and all
are in agreement that it may be a pieces esquillee" (Jolly 1972).
The physical aspects that contribute to the probability that this imple-
ment is a pieces esquillee, are: the battered, straight working edge, with
overlapping, step-like, horizontal fractures; the thin, concave upper edge,
with flutes, showing prominent wave lines, representing the end being struck
during use; and the overall thick, wedge-like shape. This type of lithic tool is
described in reports on the Brand site in Arkansas and the Debert site in Nova
Scotia, as having been used as a wedge for splitting bone (Mc Donald 1965,
Morse 1973). One method used for producing the thick flakes used in this
process is described as "bypolar", employing direct hard percussion against
one end of a core or nodule, with the opposite end resting on a stone anvil.
MacDonald states in the Debert report that the bypolar method was used ex-
tensively to produce thick flakes utilized as pieces esquillds; but that any suit-
able, thick flake was occasionally used, including small, expended cores.
As an aid to recognizing pieces esquillee among debitage in collections,
it was felt that experimenting with the use of lithic flakes as wedges for bone
splitting would provide useful information. Figure 2 shows examples of pieces
esquillee resulting from personal experimentation and the appearance of the ma-
terial does show some of the characteristics attributed to pieces esquillees.
As for the items in Figure 2, the flakes were produced by the bypolar
method and then were lightly pounded in a lateral groove cut into the large sec-
tion of bone shown at the bottom left portion of the picture. An 8 inch long sec-
tion of bone, roughly 1 inch in diameter and weighing 7 ounces, was used as a
billet. A flute resulted along the entire length of the face of number 2, within
a natural striation pattern which was present in the Fort Payne chert core
used to produce the flakes. This pattern is present on number 3, although it
is not fluted. Each tool was struck at least 25 to 30 times. During this pro-
Florida Anthropologist, vol. 27, no. 4, December 1974
Fig. 2. Flakes, bone hammer, and grooved broken
bone used experimentally. Photograph by John Martz.
pieces esquillee, ...
Dr. A. C. Long.
Fig. 3. Sketch depicting a pieces esquillee
resulting from use of a small exhausted
core, which would appear more like a
"prepared" tool than would a similar item
resulting from use of a "raw" flake.
w u wt3 I 1 1 1 1 j ... ............ ...... ......... ........... ...... -
- I I 1 .6 -
l i I 1 .1 i r i L I ... 1 ., i .. i i I, I ., I A i
cess the bone split through the upper wall while the last of the three tools was
being used. The bone wall was 5 mm thick, and the entire bone diameter var-
ied between 1 and 1. 5 inches. When the implements resulting from this sim-
ple experimenting are compared to the Figure 1 example, the similarity in
the appearances of the critical areas involved are considered to be convincing
enough to add appreciably to the probability indicated by the University of Ar-
kansas personnel that the Alabama example was actually used in the same
In the experimental bone splitting work, it was noted that the prepared
groove in the surface of the bone was about the same width as the working
edge of the Figure 1 tool. Measurements were taken and the working edge of
the tool, at it' s extremity, was 3 to 4 mm thick, and this was also the width
of the groove, resulting in such compatibility that the tool stood balanced ver-
tically, supported only by the contact with the groove walls, as would be indi-
cated by the Figure 3 sketch, if the edge of the tool were shown in contact with
the edges of the groove. The thickest part of the Figure 1 tool is 10 mm and it
measures 38 mm long and 30 mm wide, The experimental examples are
Even if pieces esquillees do definitely exist on early sites in the Ten-
nessee Valley area of Alabama, this in itself would be nothing spectacular.
The fact that pieces esquillees have been determined to be a common tool in
the Dalton Culture as shown by work at the Brand site is an indication that
their presence in other areas where the Dalton Culture is evident is to be ex-
pected. The presence of this tool may even already be better recognized in
the northern Alabama area than this paper indicates, but, none the less, fur-
ther emphasis on its probable geographic spread should be helpful to ama-
teurs who may be in a position to contribute further information on spatial
A small sample of material from the Quad site (Soday 1954) near Deca-
tur, Alabama, has been examined to check for the presence of the stepped
fractures indicative of use as a bone splitting wedge. Several classic examples
were located in less than a hundred examples of unclassified debitage, surface
collected from the Quad site over a period of years. A very large quantity of
such material from the Quad site is available for study and would undoubtedly
prove very revealing. Excavated material from the Quad site would also prob-
ably furnish interesting information. Cambron and Hulse (1960), reporting on
work from 21 5x5-foot pits in 1959 classify as crude tools "slightly modified
pieces of flint, as well as rejects. Some of this lithic material may well
have pieces esquillee traits.
Bullen and Beilman (1973) in the report on the Nalcrest Site in Florida
compare small core tools from that site to similar examples from the Brand
and Quad sites. Further observed similarities of material in north Alabama
to that found on the Brand site should be no surprise. Prior to 1972, when
the tentative pieces esquillee identification of the Figure 1 artifact was made,
no definite Dalton period material had been taken from the same site as that
item. Most of the lithic dibitage was highly patinated, however, and several
steep edged uniface scrapers, typical on early sites, along with very early
Archaic projectile point types, had been found. Then in March of 1974 a small
Dalton projectile point was found on this site, indicating Dalton period activity.
The foregoing comments offer little in way of specific conclusions. Even
this meager information, however, when compared to that available on the
subject from Louisiana, Arkansas, Florida, and elsewhere, does make specu-
lation tempting. Therefore, an armchair model by way of a prediction is pre-
sented: that substantial widespread recognition will develop, and that "pieces
esquillees" will become a familiar name in the lithic terminology of south-
eastern United States in the not too distant future. The emphirical aspect of
this prediction lies with the results to be produced by amateurs and profess-
sionals who, hopefully, will be especially alert to the possibility that exam-
ples of lithic wedges for bone working may show up on future projects.
Bullen, Ripley P. and Laurence E. Bellman
1973 The Nalcrest Site, Lake Weahyakapka, Florida. Florida
Anthropologist, Vol. 26, no. 1, pp. 1-22. Gainesville.
James W., and David C. Hulse
An Excavation at the Quad Site. Tennessee Archaeologist,
Vol. 16, no. 1, pp. 14-26. Knoxville.
Jolly, Arthur F. III
1972 Personal Communication.
McDonald, George F.
1968 Debert, A Paleo-Indian Site in Central Nova Scotia, Anthro-
pology Papers, No. 16, National Museum of Canada. Ottawa.
Morse, Dan F.
1973 Dalton Culture in Northeastern Arkansas. Florida Anthro-
pologist, Vol. 26, no. 1, pp. 23-38. Gainesville.
Soday, Frank J.
1954 The Quad Site, A Paleo-Indian Village in Northern Alabama.
Tennessee Archaeologist, Vol. 10, no. 1, pp. 1-20. Knoxville.
July 3, 1974
A POVERTY POINT OWL AMULET FOUND IN FLORIDA
Paul M. Lien, Ripley P. Bullen, and Clarence H. Webb
Paul M. Lien of Tampa, a member of the Pinellas County Chapter
(Suncoast Archaeological Society, Inc. ) found the red jasper owl amulet
illustrated in Figure 1 (bottom row) when he was diving in the Withlacoo-
chee River several years ago. It came from about 20 feet of water, some
300 feet west of Highway 200, in the main part of the river. At the Sept.
20, 1974 meeting of the Chapter, he showed it to Bullen who recognized it
as a product of the Poverty Point lapidary industry.
Subsequently, Lien brought the specimen to the Florida State Museum
where it was compared with an amulet (Fig. 1, middle) in the Alexander
Collection from Poverty Point. Later, Bullen confirmed in conversation
with Webb at the SEAC meeting in Atlanta that Lien's amulet had all the
typical characteristics of such objects as found at the Poverty Point site
on the Mississippi River in middle Louisiana. The photographs in Fig-
ure 1 were made by Kay Purinton of the Florida State Museum. Pictures
of the Hebe site amulet (Fig. 2) were made by Robert C. Morris and sup-
plied by Haag.
The Florida specimen--right, front, and left side views--is compared
with the Florida State Museum's example (FSM-103338) from the Poverty
Point site in Figure 1 while the Hebe specimen from Mississippi is illustrated
in Figure 2. Dimensions of three non-Poverty Point site finds are compared
with ranges of specimens from the Poverty Points site inself in Table 1.
Close similarity is evident with Lien's amulet appearing slightly closer to
those from Poverty Point than does the one from the Hebe site.
Lien's specimen came from Withlacoochee River and its left side, left
part of the front, and part of the back aspect are pitted or eroded. Probably,
it lay on its right side in sand or dirt while the exposed left side was scoured
by water-borne materials. The spot on its left side is yellow in color and an
inclusion in the rock. Smoothing striations are found where the original fin-
ished surface is stillpresent. The right eye shows the drilled dot. The sus-
pension hole, in the neck, is a little off center, drilled from both sides, and
tapers inward from both sides. Lien feels that on the right side a slight facet
is present to represent a wing. This is too faint for positive identification.
The bottom exhibits two areas separated by a groove. The smaller, nearer
the front, extends further downward than does the larger and represents the
feet or talons of a perched owl (Fig. 1). The larger area, to the rear, is a
flat base. It shows striations made, apparently, by string sawing. The Al-
Florida Anthropologist, vol. 27, no. 4, December 1974
i i '
M trn iC
Fig. 1. Right side, front, and left side views of owl amulets.
Top row, Poverty Point on left, Florida on right respectively, all
approximately full-scale; middle row, Poverty Point, bottom row,
Florida, both five time actual size.
LIEN, BULLEN, AND WEBB
exander Collection amulet has been mounted on styrofoam and hence these
details cannot be checked on it. They show up rather well in Figure 2.
The above description agrees closely with that of typical Poverty Point
owl amulets by Webb which follows.
At the Poverty Point site, among 47 bird effigy pendants, beads, beads,
or reforms of polished stone, 20 are of owls (Webb, Ford, and Gagliano 1974).
Nineteen are carved in the round and one is a flat silhouette. Eighteen of the
20 are of red jasper, one of galena, and one of amethyst or fluorite. Typical
red jasper owls have a flat base, with tail and two small feet, so that they
can be placed in upright stance. They have curved beaks, drilled eye dots
seated in bilateral flat facets, transversely grooved neck lines, and counter-
drilled suspension holes placed at the posterior neck or upper body area.
Individual variations include an owl head without the body, a head and upper
body with perforation through a narrow base so that it suspends head down-
ward, and a owl with a decorated circular flat area atop the head.
Recorded sizes for typical owls are given in Table 1. From Poverty
Point are three owl pendant reforms, blocked out of red jasper, showing the
beak-eye bevels, neck folds, feet and general outline, but without the perfora-
tions, the eye drillings, or the final polishing (Fordand Webb 1956: Fig. 38h).
The red jasper, from which these and many other objects at Poverty Point
were made, is thought to derive from northwestern Alabama or northeastern
Heretofore, red jasper owl pendants have been found at only two other
sites, both of the Poverty Point culture. At the Aaron Place, a satellite site,
located on the Arkansas River cutoff only seven miles northwest of Poverty
Point, a typical specimen was found on the surface by Charles Aaron. The
second is from the Hebe site in the Yazoo Basin, Washington County, Mis-
sissippi. It was a surface find made by Robert C. Morris of Leland, Mis-
sissippi, through whose courtesy we show it in Figure 2. The object came
from a small area of the Hebe site that also produced typical Poverty Point
baked clay objects and other cultural material of the Poverty Point culture.
The owls from the Aaron and Hebe sites are thought to have been manu-
factured at, and traded from, the Poverty Point site. This conclusion seems
similarly applicable to the Florida specimen in view of the distinctiveness of
the artifact and the characteristic attributes of the object. This conclusion
does not strike us as inappropriate, despite the distance, in view of many
evidences of widespread trade and interaction during Poverty Point times.
Specifically: (1) Small (1966) has reported baked clay balls of characteristic
Poverty Point types from Tick Island, Florida; (2) Bullen has identified
untempered incised pottery [St. Johns Incised] from the Poverty Point site
as trade items originating in north or northeastern Florida around 1000 B. C.;
Table 1. Owl amulet dimensions (mm).
Height Width Depth
Florida 18.5 7.0 8.0
Hebe 16.0 6.3 9.5
Aaron Place 28.0 16.5 17.0
Poverty Point 11-29 6-16.6 6.5-18
IIIIIININNIIII IINIINIINIjW H Fig. 2. Right side, front,
--- and left side views of Hebe
amulet. (Scale is metric)
(3) Webb has confirmed, in the David C. Reichelt Collection from Chocta-
whatchee Bay sites, the presence of red jasper pendants and two-hole gor-
gets, lamellar blades, cores and "Jaketown perforators, large points of
Arkansas novaculite, a Motley point of gray chert, and the joint appearance--
at Choctawhatchee Bay sites and at the Claiborne Poverty Point site--of
steatite vessels, fiber-tempered pottery and Levy-like large Archaic projec-
tile points made of Alabama orthoquartzite. Finally (4) Bullen (1959) has
shown that the Transitional period of Florida, between 1000 and 600 B. C.,
was a dynamic time that involved movement of people and ideas, experimen-
tation in ceramics, and interaction with the interior and along the Gulf Coast
to the Mississippi Valley.
Lien's owl amulet from the Withlacoochee River is anther example
of such interaction. It may be pointed out that this river is the only water-
way leading directly into the central lake district of Florida from the Gulf of
Mexico. To a traveller from the west, it offers a much shorter route than
does Tampa Bay and the Hillsborough River.
Bullen, Ripley P.
1959 The transitional period of Florida. (15th) Newsletter, Southeast-
ern Archaeological Conferences, vol. 6, pp. 43-53.
Ford, James A. and Clarence H. Webb
1956 Poverty Point, a late Archaic site in Louisiana. Anthropological
Papers of the American Museum of Natural History, vol. 40, no. 1.
Small, James F.
1966 Additional information on Poverty Point baked clay objects. Flor-
ida Anthropologist, vol. 19, nos 2-3.
Webb, Clarence H. James A. Ford, and Sherwood M. Gagliano
1074 Poverty Point culture and the American Formative. Unpublished
Tampa and Gainesville, Florida
STONE MORTARS IN FLORIDA
Ripley P. and Adelaide K. Bullen
This paper has been written to record the presence in Florida of shallow
stone mortars used, presumably, in the reduction by grinding of corn or nuts.
Such objects are known about, informally, but are not specifically recorded in
the archaeological literature of Florida.
Fragments of presumed grinding stones are recorded for late sites such
as Lake Jackson (Griffin 1950:103) north of Tallahassee, the Ft. Walton zone
at Site J-5 beside the Chattahoochee River (Bullen 1959:346) and Safety Harbor
(Griffin and Bullen 1950:P1. IV, _) beside Tampa Bay. These, however, are
more-or-less parallel-sided flatish fragments with at least one rather smooth
surface which tends to become a little concave in the case of the more used
The mortar illustrated below (Fig. 1) is quite different. Approximately
15 pounds in weight, it is about 12 inches across, fairly thick, and made of
cherty limestone. The concavity is about as wide and deep as the stone permits.
That it was intentionally shaped is evident by the chipping scars along the side
which expose the cherty nature of the specimen.
This mortar was found by Kim Borders of Hawthorne in a cultivated field
in the Micanopy-Rochelle area of Alachua County. Preceramic Archaic pro-
jectile points as well as Deptford, Weeden Island, and Alachua Tradition sites
are known for this region. St. Johns Series pottery is frequently found at
these sites although they are well removed from the St. Johns River region.
Interestingly, a larger but similar mortar (FSM-96339) was found by
A. K. Bullen on the surface of the Ross Hammock site, beside the Indian River
in southeastern Volusia County, and brought back by the authors and William J.
Bryant in 1965. Occupation at Ross Hammock started early in the St. Johns I
period and extended up into the early part of the St. Johns II period.
This second mortar was broken crosswise at what appears to be the cen-
ter. The available "half" is 16 inches across and 12 inches long suggesting the
mortar may originally have been 24 inches long. The ovate grinding hole is 3
inches deep while the maximum thickness of the mortar is 4. 5 inches. It was
made of coquina.
Florida Anthropologist, vol. 27, no. 4, December 1974
Griffin, John W.
1950 Test Excavations at the Lake Jackson site. American
Antiquity, vol. 16, no. 2, pp. 99-112. Menasha.
Bullen, Ripley P.
1958 Six sites near the Chattahoochee River in the Jim Woodruff
reservoir area, Florida. River Basins Surveys Papers, No.
14. Bureau of American Ethnology, Bulletin 169, pp. 315-57.
Griffin, John W. and Ripley P. Bullen
1950 The Safety Harbor site, Pinellas County, Florida. Florida
Anthropological Society, Publication No. 2. Gainesville.
Fig. 1. Two views of Kim Borders mortar.
CAGLES HAMMOCK, CORAL SPRINGS SITE NO. 5
Broward County Archaeological Society Project
prepared by Bert Mowers and Wilma B. Williams
Cagles Hammock, known locally as the "Big Head", is situated in the
wild southwest corner of the City of Coral Springs, in the N.E. 1/4 of Sec-
tion 31, Township 49S and Range 41E., about 800 feet south-southwest of the
section corner. It is a rather small mound on the north end of the wooded
hammock (Fig. 1). The usual ficus tree clump indicates the slight elevation
that made habitation possible in the wet Glades, before modern drainage.
The rest of the hammock is a cypress swamp.
After a datum was established, a row of 5- by 5-foot squares was
staked out for excavation (Fig. 1). Supplementary squares brought the total
to 15 as numbered on the excavation plan. Artibrary 6-inch levels were used,
and all material was screened. All artifacts were washed, classified and cat-
The first four inches of Level 1 was dark sand with organic humus. A
few modern artifacts such as cartridge cases were found in the first few
inches. A considerable amount of potsherds were found which, as the pre-
liminary survey indicated, belonged almost exclusively of the Glades II peri-
od as shown by the incised designs (Table 1). Level 2 contained even more
sherds with a striking increase in Miami Incised and plain rims. Rock ap-
peared from 10 to 20 inches below the surface with solution holes and deep
cracks between rock masses of the underlying Miami Oolite limestone. The
holes and cracks were dug out carefully, on the principal that the oldest ma-
terial was the deepest, and holes were obvious places to dump refuse.
The deep humic area, quite unusual in hammock sites, indicates an al-
most complete lack of disturbance for a long time, either by white man, Sem-
inole, or late prehistoric occupants. Pottery time-markers for the Glades III
period were virtually absent, indicating little use of the site during the last
millennium (Table 1).
The only exception to the undisturbed condition of the site, was a burial
uncovered in Square 3. Here, the original digging done to inter the corpse,
inverted the usual sherd count, more being found in Level 1 than in Level 2.
Florida Anthropologist, vol. 27, no 4, December 1974
172 CAGLES HAMMOCK
(_ N FLORIDA HOLLY )
DETAIL OF BURIAL
Fig. 1. Upper, excavation plan; lower,burial details.
MOWERS AND WILLIAMS
In Square 3, Level 3 (12-18 inches), a hard, gray layer of stoney ma-
terial was struck at depths ranging from 14 to 18 inches below the surface.
This layer contained some midden material--bones, shell bits, etc. --and
hence it was not true rock but a concreted mass formed after deposition of
the earliest midden was completed. This layer was continuous across this
square, following the slope of the surface.
However, this was found to be but a thin layer, variable in thickness
and much softer than the Peace Camp concretion (Mowers and Williams 1972).
It did not extend across the entire site, although isolated lumpy masses were
Obvious physical differences, allowing stratigraphic digging, were not
observed at this site, unlike Peace Camp where almost all excavation fol-
lowed stratification. Also, the characteristic "zone of incipient concretion, "
as it was called at Peace Camp, did not occur here. This was a thin level of
pebbly, grayish colored midden aggregates, grading down into the solid stoney
concretion. Apparently, all the conditions leading to formation of a solid con-
cretion were not present at Cagles Hammock; for instance, the great amount
of hauled-in sand and marl found at Peace Camp, was not duplicated here.
Dark midden was found below the thin, stoney layer, and bedrock of
limestone was found at 23 inches below the surface.
Below the concretion, only one small sand-tempered sherd was found.
Food bones were fairly numerous, and one columnella tool with ground tip,
one broken bi-point, several Busycon tool fragments, an unworked antler tip,
and turtle shell, testified to human occupation.
For the most part, all of the 5280 sherds found at Cagles Hammock could
be classified in one of the following wares: (a) St. Johns, a chalky, mostly un-
tempered, paste; (b) Belle Glade, a chalky but slightly sand-tempered paste,
usually quartz sand, with a "scratched surface", but occasionally fine sand
and unscratched; (c) Goodland, a heavily sand-tempered paste with a soft, open
textured appearance, and color ranges from light tan to orange-red, to brown
and black, with fine to medium coarse beach sand as temper; and (d) Glades
Gritty ware, the usual sand or sand and grit tempered ware characteristic of
the Glades area. It must be emphasized that these four wares have ranges of
physical characteristics, and occasionally grade into sherds that are inter-
mediate, making classification difficult. Considering that the occupancy of this
site spanned several hundred years, and in view of the varying conditions of clay
sand and grit materials plus the limited skills and technical difficulties of fir-
ing pottery under primitive circumstances, it is indeed remarkable that only
four categories of paste are necessary. A few sherds included shell-temper-
ing but other ingredients in the paste allowed for classification within the above
four categories. No fiber-tempered or limestone-tempered sherds were found.
The incised designs found on the sherds of Cagles Hammock are those of
the Glades II period as described by Goggin (1949, 1952, 1964): Key Largo,
Dade, Opa Locka, Miami, Matecumbe, Cane Patch, and Sanibel Incised and
St. Johns Check Stamped (Table 1). Infrequently, a sherd of Goodland paste
was found with one of these patterns, but normally they were found on Glades
Gritty ware. Occasionally, a rim tooling appeared on an incised sherd. Of
doubtful classification are the pinched or tooled rims not recognized as be-
longing to any recorded pattern.
The variants of Key Largo designs and Opa Locka are also difficult to
classify. For instance, is Figure 3, c a two loop Key Largo or a two stacked
Opa Locka variant? A number of unclassified incised sherds (Fig. 3) were
found and are being studied. A sherd of overall punctation is possible an im-
port from Englewood. An unusual pot of Goodland paste, but with a tooled rim
resembling Glades Tooled, was found in association with the burial in Square 3.
It may be a prototype of Glades Tooled or represent an intrusive burial. These
irregular shaped sherds, of a paste similar to that of Goodland Plain, were
parts of a heavy based pot with square corners like a griddle (Fig. 3, g-h).
Possibly they represent parts of a Weeden Island compartmental vessel.
Sherds by levels are given in Table 1. Four pits showed more sherds
in Level 1 than in Level 2. Apparently, most of the sherds were between
depths of 4 to 6 inches. Level 1 had a total of 102 incised sherds, all of Glades
II period (some unclassified incised sherds may be an exception to this). Level
2 had a total of 137 incised sherds, all but 3 (2 Cane Patch and 1 Sanibel In-
cised) undoubtedly Glades II period, with the possibility of some unclassified
sherds being of the Glades I period. These figures seem to indicate over-
whelming Glades II period for this site. Very little evidence of Glades III usage
(7 sherds), or of Glades IB (3 sherds), exists (Table 1). Artifacts other than
pottery hint only at a limited Archaic and/or Transitional occupation of the
site. This hammock, then, is probably as near to a closed site as we are
likely to get.
In a comparison of relative amounts of Belle Glade Plain, Glades Plain,
and Goodland Plain, slightly more Belle Glade was found in Level 1 than in
Level 2. More Goodland Plain was found in Level 2 than Level 1. This seems
to indicate that the influences from Belle Glade peoples was later than those
from the Goodland Point area. Of the 1092 sherds of Belle Glade-Goodland
found, only 78 or 7. 2% were Belle Glade Plain. Therefore, it would appear
MOWERS AND WILLIAMS
Fig. 2. Photographs of unusual artifacts.
a, bone dagger; b, steatite pendant; c,
shell pendant;d-e, tip of and complete
greestone celt; f, stone from burial.
Table 1. Pottery distribution by 6-inch levels.
0-6 6-12 12-18
St. Johns Check Stamped
St. Johns Plain
Key Large Incised
Opa Locka Incised
Cane Patch Incised
Belle Glade Plain
Totals 2245 2742 293 5280 100.0
*From one vessel.
that Goodland influences were dominant during Glades II times and even more
so in Glades I times.
Glades Plain sherds cannot be used to indicate chronology with any accu-
racy at this site. An undecorated rim sherd may be either Glades I, II, or III.
All Glades Gritty ware body sherds without markings are counted as Glades
Plain, although many may be pieces of incised pots. Rims and body sherds
have been matched, where possible, then reported as individual pots. It is be-
lieved that this method gives more and better information than simply counting
rims and body sherds as such. But serious errors still exist. There is no good
marker for Glades IA. Glades Plain rim sherds in Level 1 totaled 175, 7.8%
of the whole. Glades Plain rim sherds in Level 2 totaled 219, 8. 3% of the whole.
These differences are too small in the light of known errors in classification
and digging from which to draw valid conclusions.
The assemblage of non-ceramic artifacts from this small site is typical
of the Glades tradition (Laxson 1962, Mowers and Williams 1972): 50 shark
vertebrae (both perforated and unperforated), 234 shark teeth, 38 bipointed
bone tools, 4 socketed bone points, 4 bone gouges, 7 Strombus celts, 3 Busy-
con tools, 5 columnella tips, 2 fragments of worked shell, 2 examples of
worked bone, a worked stone, 3 instances of imported stone, a piece of flint,
and, with the burial, an antler rack.
The shark teeth found were both large and small, perforated and unper-
forated. This would indicate a common trait of knife making and ornamenta-
tion. The shark vertebrae found were probably used as necklaces or ear plugs.
A single Stingray spline and a single Parrot fish mouth plate were also found.
After pottery, bone artifacts as listed were the most common tools. All
were made from deer bone. One bone bead was found and several specimens
identifiable as dog bones. Shell tools, also listed above, were less frequent
than usual in Glades sites. Noteworthy is a plummet (Fig. 2, c).
A fragment of a small greenstone celt was found in Level 2. It had been
ground and polished to a fine edge (Fig. 2, d). Two matching fragments of a
steatite gorget were found, which when put together formed only half the arti-
fact (Fig. 2, b). This is the first steatite ornament found in Broward County,
to our knowledge. A rather thick miniature greenstone celt (Fig. 2, e) was 4
centimeters wide by 5 centimeters long, and ground to a thick but sharp edge.
Between the femurs of the burial, a roughly ovate stone (Fig. 2, f) and a
broken piece of deer bone were present.
MOWERS AND WILLIAMS
Midden burials are the exception in Glades hammock sites but, after
troweling the surface of Square 3, a pelvic bone was uncovered. Careful ex-
posure revealed other bones of an extended human burial. This was pedes-
taled for photographs.
The skeleton was extended in a NNE-SSW direction, skull to the north,
lying on its back, skull turned to the left, the right arm diagonally across
the midsection and leg bones and feet extending into Square 14 (Fig. 1, lower).
Bones were in extremely bad condition, soft and fragile, with many breaks.
The skull was crushed and filled with dirt. The mandible was out of position
to the right side of the skull. The rib cage was crushed on the left side.
In association with the body were sherds of a large Glades Tooled pot
made of a Goodland type paste. It was located about one foot west of the skull.
Between the broken pot and the rib cage was the left rack of a deer antler. A
very unusual stone object (Fig. 2, f) and a small broken piece of deer bone
were found between the femurs (Fig. 1, lower). Chemical erosion has re-
moved all traces of work performed in shaping the object, its usage is un-
This was possibly a ceremonial burial, with the body laid in a shallow
grave dug into the midden material, and accompanied by a killed bowl of food,
a symbolic antler and some of the deceased' s worldly goods. Bits of other pots
and some shark teeth, found in association with the bones, are believed to be
components of the midden into which the grave was dug.
The 48 sherds of the pot found near the skull were the only Glades
Tooled sherds found on the site. It is logical to believe that the body was that
of a traveler who died here long after the settlement had been abandoned,
perhaps several centuries later.
Summary and Conclusions
This small Glades hammock was little used for the past thousand years,
and then only as an overnight campsite. Prior to that, it was an extended hab-
itation for family use for a long time which would include full occupancy dur-
ing the Glades II times, as shown by the many incised sherds of that period.
The undecorated Glades Gritty ware rim sherds, the steatite gorget and other
stone tools hint at an even older occupancy.
The large number of shark teeth and vertebrae indicate full participa-
tion in a coastal economy. The sea provided great resources for food and tool
making materials. The Strombus gigas celt, the Busycon tools and utensils,
the Pleuroploca hammers and chisels, shark teeth knives, the occasional
stingray spline and the pumice abrader, all indicate a sea oriented life.
Artifacts other than pottery (deer bone and antler) reflect the inland loca-
tion of the site. However, they still used the shell celt and adze to make
their dugouts, and the highly sophisticated shark tooth knife for smaller work.
The relative abundance of Belle Glade and Goodland wares is especially
interesting. Table 1 shows a higher percentage of Goodland ware than is indi-
cated in the literature for the surrounding region. The whole question of
Glades Gritty ware, and particularly of Goodland variants, is being inten-
sively studied, with the object of creating an orderly arrangement of the
Glades ceramics, at present poorly understood.
This culture may have had its beginning in the Archaic period on the
shell mounds along the St. Johns River and the Indian River lagoons. People
gradually filtered down from there along the coast and the Glades water-ways,
to the tip of the Florida Keys. At Peace Camp, near Davie, a pre-ceramic
level dated at 3050 years B. P. (1100 B.C.) plus or minus 140 years (Mowers
and Williams 1972) yielded a great number of Strombus gigas celts and other
Permission to excavate this site was given by Coral Ridge Properties,
Inc. a subsidiary of Westinghouse Electric Corp. through Mr. Gordon Ickes,
who has cooperated with the Broward County Archaeological Society in its re-
ports on five other sites within the Coral Springs area. Our appreciation is
also due Mr. Russ Cagle who arranged transportation by swamp-buggy and
jeep. Members participating in the excavations include: Lorenzo, Judy, and
Juanita Mixon, Kipp White, John Betts, Dorothy and Bert Mowers, Robert
D'Andrade, Jo and Wilma Williams, Merle Fairbanks, Deborah Jean Smith,
Todd White, Wayne Lafferty, Leon Reyniers, Lucille Miller, Fred Scheidler,
James and Eleanor McAndrew, and John Sauls. Photographs by Dr. Stephen
Goggin, John M.
1949 Cultural Occupation at Goodland Point, Florida.
Florida Anthropologist Vol. 2, No. 3-4.
1952 Space And Time Perspective in Northern St. Johns Archeology,
Florida. Yale University Publications in Anthropology,
1964 Indian And Spanish Selected Writings. University of Miami
MOWERS AND WILLIAMS
Mowers, Bert, and Wilma B. Williams
1972 The Peace Camp Site, Broward County, Florida.
Florida Anthropologist Vol. 25, No. 1, pp. 1-20.
Fig. 3. Unusual sherds from Cagles Hammock.
a, unclassified punctated; b, e-f, unclassified incised; c, Key LaTgo variant;
d, St. Johns Simple Stamped; g-h, two sherds of square-cornered bowl.
In accordance with U. S. Postal Regulations No. 132. 622, the following
statement of OWNERSHIP, MANAGEMENT AND CIRCULATION for THE
FLORIDA ANTHROPOLOGIST is included in this issue.
Florida Anthropological Society, Inc.
c/o Ripley P. Bullen, Resident Agent
Room 102, Florida State Museum
Gainesville, FL. 32611
Officers and Executive Committee
See inside front cover for names and addresses.
Total no. copies printed Average no. copies Sept., 1974 issue
Varies from 1000 each issue Vol. 27, no. 3
to 1150 1100 1150
Sold, UofF Library-Foreign 220 255*
Mail Subscriptions 820 803
Total Paid Circulation 1040 1058
Free Distribution 3 3
Total Distribution 1043 1061
Office Use, Left-Over 57 89
Total 1100 1150
*35 copies sold authors, this issue only.