2000 Florida Anthropological Society Inc.
The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
all rights to the
and shall be
and images of
The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.
The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.
Contact the Florida Anthropological Society for additional
information and permissions.
THE FLORIDA ANtHROPOLOGICAL SOCIETY
./ 4. .. "
THE FLORIDA ANTHROPOLOGIST is pub-
lished quarterly in March, June, September, and
December by the Florida Anthropological Society
at the Department of Anthropology, Florida State
University, Tallahassee, Florida 32306. Sub-
scription is by membership in the society for in-
dividuals interested in the aims of the society.
Annual dues are $4.00; student memberships,
$2.00. Requests for membership and general in-
quiries should be addressed to the secretary;
subscriptions, dues, and back issue orders to
the treasurer; newsletter items to the president;
and manuscripts and books for review to the
editor. Second class postage paid at Talla-
OFFICERS OF THE SOCIETY
Ripley P. Bullen
Florida State Museum
2nd Vice President
Carl A. Benson
2310 Resthaven Drive
Orlando, Florida 32806
1st, Vice President
James W. Covington
University of Tampa
Tampa, Florida 33620
Howard A Chamberlen
Florida State Museum
Gainesville, Florida 32601
L. Ross Morrell
Florida Board of Archives and History
401 East Gaines Street
Tallahassee, Florida 32301
David S. Phelps
Department of Anthropology
Florida State University
Tallahassee, Florida 32306
David S. Phelps
Department of Anthropology
Florida State University
Tallahassee, Florida 32306
Robert C. Dailey
Department of Anthropology
Florida State University
Tallahassee, Florida 32306
L. Ross Morrell
Florida Board of Archives and History
401 East Gaines Street
Tallahassee, Florida 32301
Dan D. Laxson
231 West 41st Street
Hialeah, Florida 33012
William D. Goza
P. 0. Box 246
Clearwater, Florida 33517
Cliff E. Mattox
P. O. Box 521
Cocoa Beach, Florida 32931
William H. Sears
Department of Anthropology
Florida Atlantic University
Boca Raton, Florida 33432
Yulee W. Lazarus
Temple Mound Museum
P. O. Box 294
Fort Walton Beach, Florida 32548
Ripley P. Bullen
Florida State Museum
Gainesville, Florida 32601
Stanley J. Olsen
Department of Anthropology
Florida State University
Tallahassee, Florida 32306
Wilfred T. Neill
122 Homecrest Road
New Port Richey, Florida 33552
WILFRED T. NEILL
The Galphin Trading Post at Silver Bluff
South Carolina.......................................................................................................... 42
D. D. LAXSON
The Dupont Plaza Site.......................................................................................... 55
R. B. HOLLIMAN
Bird Hammock, Mound B, Revisited...................................................................... .... 61
The Palm River Midden, Hillsborough County, Florida................................................. 67
ALBERT C. GOODYEAR
Pinellas Point: A Possible Site of Continuous
Indian Habitation........................................................................................................... 74
LYMAN 0. WARREN
The Apollo Beach Site, Hillsborough County............................................................ 83
DONALD W. SHARON
A Lithic Dagger from Choctawhatchee Bay ................... ...... ................................ 89
RIPLEY P. BULLEN
Beveled Stemmed Points from Tampa Bay.......................................................... 89
ALBERT C. GOODYEAR, W. THOMPSON AND L. O. WARREN
Suwannee Style End Scrapers from Pinellas County.......................................... ...... 91
LYMAN 0. WARREN
Caladesi Causeway: A Possible Inundated Paleo-lndian Workshop............................... 92
RIPLEY P. BULLEN AND MARJORIE H. WING
A Scraper with Graver Spurs from Florida................ ............................................... 94
THE FLORIDA ANTHROPOLOGIST
Volumes XV through the current volume may be purchased from the Treasurer of the Society at $4.00 per
Volumes I through XIV may be purchased from Johnson Reprint Corporation, 111 Fifth Avenue, New York,
New York 10003. Prices are as follows:
Volumes I, V, X $6.00 per volume
Volumes II, IV, VI, IX, XI, XIV $8.00 per volume
PUBLICATIONS OF THE FLORIDA ANTHROPOLOGICAL SOCIETY
Numbers 1 through 5 are available from the Johnson Reprint Corporation at the following prices:
Numbers 1 and 3 $2.00
Number 2 $2.50
Number 4 $4.50
Number 5 $3.00
THE FLORIDA ANTHROPOLOGIST
THE GALPHIN TRADING POST SITE AT SILVER BLUFF, SOUTH CAROLINA
Wilfred T. Neill
Refuse pits at George Galphin's 18th Century
trading post yielded a small sample of Indian
pottery, hammerstones, food remains, and trade
goods. Contrary to a literature report, the abo-
riginal sherds are incised, not brushed. The
sherds should be likened to Ocmulgee Fields
Incised, not the Seminole brushed types. Gal-
phin dealt primarily with the Lower Creeks, not
with a small Yuchi band who lived for a brief
while near his post. The sherds are probably
ascribable to the Lower Creeks, not the Yuchi
as a previous report would have it. Galphin's
connections were, specifically, with the Coweta,
or "war town," group of the Lower Creeks.
Cofitachiqui, a proto-historic Lower Creek
town at Silver Bluff, is identified as Coweta,
not Kasihta as some would have it.
Pipestems from the Galphin Trading Post Site,
dated by stemhole diameter, need not be as-
cribed in any part to the Yuchi.
In latter 1949 I brought some sherds to the
attention of John M. Goggin. Writing from
memory years later, Goggin (1958) described
this material as a large sample of pottery,
probably Yuchi, from "Gallatin's" trading
post. He stated that vessels were globular,
with constricted mouths; and that decoration
was by heavy brushing. This description
erred in most particulars. The name of Albert
Gallatin is familiar to anthropologists; but
this brilliant statesman, who after retirement
founded the American Ethnological Society,
never operated a trading post. The site was
the trading post of George Galphin. While the
pottery could be Yuchi, it is more likely
Lower Creek. In fact it could be ascribed,
with considerable hope of accuracy, to the
Coweta group of the Lower Creeks. The sample
included but eleven sherds, these from six
bowls and three jars. Most importantly, no sherd
is brushed; decoration, when present, is by
incision. The sherds should be likened, not to
the Seminole brushed types, but to Ocmulgee
Fields Incised. The entire Silver Bluff sample
is illustrated (Fig. 1-2).
In proto-historic times, Silver Bluff was the
location of an Indian town called Cofitachiqui.
It has long been clear that this was a Lower
Creek town, either Kasihta or Coweta. A modern
look at the evidence negates the view, once held
by Swanton (1922), that it was Kasihta.
Also, misconceptions entered into an analysis
of pipestem fragments from Silver Bluff (Eaton
1962). The mean date of the fragments was sup-
posed to determine the temporal midpoint of the
post's occupancy, bui seemed much too early.
As shown later, the inconsistency can be re-
solved, and without invoking the Yuchi as Eaton
thought might be necessary. Contrary to several
authors, the Yuchi never lived precisely at
Silver Bluff; and their village, two miles away,
seems to have been occupied for a few years
The Galphin Trading Post Site is of consider-
able historical and archeological interest. It is
obviously desirable to clear up the confusion
that has surrounded colonial activities and
Indian occupations at or near Silver Bluff.
ACTIVITIES AT THE TRADING POST
George Galphin was an important trader of the
Georgia and South Carolina area during colonial
days. His post was at Silver Bluff, on the South
Carolina side of the Savannah River. The local-
ity was about ten airline miles from early Au-
gusta, Georgia, but nearly 20 miles by the wind-
ing stream. (The nearest modem community is
Kathwood, Aiken County, South Carolina.) Gal-
phin settled at Silver Bluff about 1739 (Williams
1930: 288). He dealt chiefly with the Creeks,
and was mentioned in specific connection with
the Creek Indian trade as early as 1744. In 1747
be bought from Lachlan McGillivray a 400-acre
tract which included Silver Bluff (Meriwether
1940: 69-70). The purchase in no way rules out
Galphin's earlier presence at the bluff since this
area, regardless of nominal ownership, func-
tioned primarily to further the aims of the colo-
nial government. Galphin soon received larger
land grants from South Carolina and Georgia. His
responsibilities went far beyond simple trading.
His task was that of encouraging Indian indebt-
edness, which might be paid off with land ces-
sions. Much of his revenue came from British
VOLUME 21, NOS 2 & 3
JUNE SEPTEMBER, 1968
subsidization of this work. Payments to him
were through the British governor of Georgia.
Galphin's theater of activities lay mostly to the
south and west of the post, extending as far as
Charleston, Savannah, St. Augustine, Pensacola,
and Mobile. It was at Galphin's establishment
that gifts and annuities were distributed to the
tribes. Indians from far away brought hides and
produce to the post, which was a major source
of firearms, gunpowder, hoes, blankets, salt,
knives, calico, and trinkets. Other traders, some
of them operating far from the Savannah, ob-
tained their wares from Galphin. Planters got
their tools and supplies from him. His post, a
way-station on the Savannah, offered barge
docks, cabins for the accommodation of visitors
(including an occasional governor), and holding-
pens for livestock. By 1757 Galphin owned 40
slaves, and no doubt acquired more as his for-
As the American Revolution became imminent,
Galphin lost the favor of some British officials
because he sympathized with the rebellious
colonists. The Creeks and the Cherokees to-
gether had owed Galphin the sum of 9,791
sterling. The Indians had paid their debts with
land cessions to the colonial government, from
which Galphin would then collect this sum. But
James Wright, British governor of Georgia, with-
held payment. The trader then filed a claim
against Wright for the aforesaid amount. (Chero-
kee indebtedness to Galphin does not mean that
these Indians were coming to Silver Bluff. James
Adair, trader to the Cherokee, was supplied
through Galphin.) Although becoming unpopular
in some quarters, Galphin in 1775 was appointed
Commissioner of Indian Affairs for the Southern
District. Henry Laurens (soon to become Presi-
dent of the Continental Congress) wrote a con-
gratulatory letter. When the Revolution broke
out, Fort Galphin was built to protect Galphin's
goods. More accurately, it was built to protect
the-Indians' annuities, which included precisely
the items needed by the poorly supplied revolu-
tionaries. Fort Galphin was a brick building,
stockaded (Jones 1873: 151-152). In 1780, at
the age of 71, the trader died (White 1855: 247),
and the next year, Fort Galphin fell to American
troops under Henry ("Light Horse Harry") Lee.
After the American victory in the Revolution,
Galphin's holdings were not confiscated but
were turned over to his heirs. The trading post
locality was named Galphinton, and in 1785 an
American treaty with the Lower Creeks was
staged there. Galphin's former claim against the
British governor of Georgia was re-filed by his
heirs against the United States government.
Galphin's descendants pursued the "Galphin
Claim," a legal tangle of exceptional complex-
ity, to a settlement in their favor in 1848. The
principal of the debt was paid by the United
States to the trader's heirs in 1849. The in-
terest--amounting to $191,352.00-was paid in
1850. The public outcry against the payment was
so great that the entire cabinet would have been
dismissed, had not attention been diverted by
the unexpected death of President Zachary Tay-
lor after little more than a year in office (Hamil-
ton 1940: 368).
Galphin's affairs, as I have outlined them,
are standard fare in many history textbooks
(Evans 1913). Also see McCall (1811: Appendix)
in addition to references cited. I summarize
these affairs in order to show that they were not
limited to dealings with an obscure little band
of Yuchi who lived for a brief while at a locality
about two miles below the post. Within Galphin's
arena of interest, the powerful, land-rich Lower
Creeks were always the people to be reckoned
with. These Indians, although concentrated
during Galphin's time in central and west-central
Georgia, had but recently relinquished claim
to the Savannah River valley. When James Ogle-
thorpe, founder of the colony of Georgia, visited
the Lower Creeks in 1739 to clarify their land
cessions on the Savannah, it was to Coweta
that he went (Lowrie and Franklin 1834: 786).
Coweta was the principal war town" of the
Lower Creeks; colonial officials looked on it
as the capital of the Creek Confederacy, and
always reserved the title "Emperor" for the
Coweta chief (Swanton 1946: 126). In 1761 there
was published in the Georgia Colonial Docu-
ments an official list of traders; George Galphin
was listed (VIII: 522) as being trader to the
Coweta. Although living on the Chattahoochee
at this period, the Coweta were coming to Gal-
phin with their trade goods, and to receive their
annuities. The Lower Creek Trading Path con-
nected the Lower Creek country with the Savan-
nah (Myer 1928: pl. 15).
The extent of Galphin's influence among the
tribes is revealed by an episode of 1777 (Van
Doren 1928: 355-356, 366). In that year the nat-
uralist-explorer William Bartram visited the
Muklasa town on the Tallapoosa River, Alabama.
Here the Indians had beaten an adulterous white
trader with hickory cudgels, and had further
determined to have either his ears or his entire
stock of trade goods. However, the Muklasa a-
greed to remit their demands if Galphin would
intercede for the offender. Bartram took the mes-
sage to Silver Bluff, where Galphin said he
would do whatever he could about the matter.
The Muklasa, connected with the Albama and
Koasati, were only loosely affiliated with the
Creek Confederacy, yet were willing to await
and abide by the ruling of Galphin on the far-
away Savannah River.
THE GALPHINS AND THE CREEKS
In the Southeast during colonial and early
American times, many Scotch-Irish immigrants,
involved with politics and the Indian trade,
made advantageous marriages into some tribes
(e.g.,Adair, Bernard, Burgess, Cornells Grierson,
Kinnard, MacNack, McGillivray, McGive, Mc-
Queen, Powell, Ross, Stuart, Thomas, Vann,
Watts, Weatherford). As noted by Milling (1940:
339, footnote), George Galphin was no excep-
tion to the general rule. I have found no specific
statement that the trader's wife, or wives, came
from the Coweta; but his half-Indian son, John,
lived among the Coweta group of the Lower
Creeks, acted as spokesman for them, and con-
sidered himself one of them. Excerpts from let-
ters by or about John Galphin (Kinnaird 1946:
161, 168, 219-220, 342) will document this state-
ment. (I modernize the spellings in these com-
In a 1793 letter to William Panton (of the trad-
ing firm Panton, Leslie and Company), John
Galphin stated, "The (Lower Creek) chiefs re-
quested that I should do all their business for
them in future. "To Georgia officials he sent a
1793 letter stating, "We view with astonishment
the steps taken by your people when sending
peace talks in our Nation continually...We act-
ually see our hunting grounds laid out into dis-
tricts, without considering us to have any right
or claim...talking at a distance will never settle
matters,so we shall expect you in twenty days."
In a 1794 letter John Galphin replied for the
Lower Creeks to a query from Spanish interests.
Addressing his remarks to James Burgess (Don
Diego Burges), J. Galphin wrote, "...the Amer-
icans are building forts on this side (the west
side) of the Oconee River. We are resolved to
dispute our rights with them...We cannot leave
our Nation...for if we should go to see our friend
(the Pensacola governor), the Americans...might
...carry off our women and children...We are de-
termined to go in a large body against these
people who have established themselves on this
side of the Oconee...Call Long Warrior and the
king of Mikasuki, and inform them...of what we
are going to do, and tell them to send the talk
to Payne at Alachua." (Payne and the Mikasuki
king were leaders among the early, Hitchiti-
speaking element of the Seminole. Long Warrior,
whom William Bartram called a "king of the
Seminoles," may have been a Lower Creek but
more likely was Hitchiti.)
In 1793, Louis LeClerc de Milford (who lived
among the Lower Creeks and aspired to lead
them) sent a letter to Francois Hector, Baron de
Carondelet (the governor of Louisiana and West
Florida). Milford wrote, "But there is another
rascal among the Lower Creeks. This is Gal-
phin, who, several days after robbing the Amer-
icans, stole something from two of your inhabi-
tants near St. Mary's. He robbed them of every-
thing -- blankets, sheets, mattresses, seven
Negroes, their cows, and horses. There was also
a little boy whom Galphin scalped."
Pedro Olivier (Spanish commissioner for the
Creeks), in a 1793 letter to Carondelet, describ-
ed a complaint from the governor of St. Augus-
tine to the Kasihta chief, concerning Lower
Creek raids led by J. Galphin, Olivier mentioned
"depredations committed by the Lower Creeks
on the settlements in Spanish territory near the
St. Mary's River ... The delinquents in these
robberies are the same who committed the mur-
ders and sacking of the store of the agent of the
United States, Mr. Seagrove." (Actually the
store,raided by J.Galphin and his Lower Creeks,
was that of Robert Seagrove, not the Indian a-
gent James Seagrove.)
Olivier continued, "Having deliberated on this
point, it was determined (by the Kasihta chief)
that the half-breed Galphin should be punished
by death." But Yahola Micco, chief of the Cow-
eta, wrote to Enrique White, governor of Pensa-
cola, "The Tukabahchee head warrior and the
head warrior of the Kasihta had agreed to kill
John Galphin on account of mischiefs at St.
Mary's. But I could not agree to any such do-
Obviously John, the half-Indian son of trader
George Galphin, was handling external affairs
for the Lower Creeks, with the approval of the
Coweta chief who, although no longer styled an
"Emperor," headed the most powerful Creek
faction, the Coweta "war town" group of the
Lower Creeks. J. Galphin's letters were sent
from Broken Arrow (Thlikatcka). This important
Lower Creek settlement was a recent offshoot
of Coweta, and occupied a former Coweta town
site (Swanton, 1946: 126-127). After Creek re-
moval to a western reservation, the Coweta and
Broken Arrow people settled near each other,
and somewhat apart from the other Creeks (Speck
1907: pl. 1). Among these matrilineal and matri-
local Indians, a child customarily lived with its
mother,and belonged to her town and clan (Swan-
ton 1928a), so we might guess that John Gal-
phin's mother was from those Coweta who formed
the Broken Arrow town.
Thus George Galphin, in both his trade and
his personal life, was particularly associated
with the Coweta. Two other facets of the sub-
ject remain to be considered. These are, first,
the proto-historic Indian town of Cofitachiqui,
which stood at Silver Bluff; and, second, the
Yuchi who lived near the' bluff in Galphin's day.
Silver Bluff is the probable site of Cofitachi-
qui, a proto-historic Indian town mentioned in
narratives of the De Soto expedition and the
Pardo expedition, as well as in certain later
documents. Swanton (1946: 127, 143) identified
the town as Lower Creek, either Kasihta or
Coweta. (Kasihta was the principal "peace
town," as Coweta was the "war town.") Choice
between these two is simple if attention is fo-
cused on early records, not later embellish-
ments. De Soto met the Kasihta about 250 miles
west of Cofitachiqui. He should also have met
the Coweta; and Cofitachiqui is the only De
Sotan locality that could be associated with
Also, the simple linguistic evidence has been
ignored. The Spanish pronunciation of "Cofita"
is very close to the present-day Creek and
Hitchiti (Mikasuki Seminole) pronunciation of
Coweta (a name often rendered Covita, Cavita,
Cabita, or Caouita in early documents). If the
town name was provided in Hitchiti by one of
the Hitchiti-speaking guides whom De Soto had
recently obtained,"chiqui"could signify "dwell-
ing-place;" and if (as seems more likely) pro-
vided in Muskogee (Creek) by an inhabitant of
the town, "chiqui" could signify "terraced" or
"piled into a mound," with reference to cere-
monial earthworks. Such earthworks, an impress-
ive series of them, did exist at Silver Bluff, as
noted below. (Remarks on Muskhogean terms are
based on linguistic material collected by me
from the Cow Creek (Creek-speaking) and Mika-
suki (Hitchiti-speaking) Seminoles of Florida;
but also see Loughridge and Hodge 1914.)
Furthermore, Pedro de Torres visited Cofi-
tachiqui in the year 1628, and reported to the
Spanish governor of Florida that all the chiefs
of the country were under the chief of that town.
In 1670, Henry Woodward wrote that he had visit-
ed Cofitachiqui, "where ye Emperour resides."
In the early 1670's, both William Owen and John
Locke referred to the chief of Cofitachiqui as
the Emperor. (Early references to Cofitachiqui
are summarized in Swanton 1922: 216-220.) As
correctly noted by Swanton (1946: 126-127), the
whites regarded Coweta not Kasihta as the
capital of the Creek Confederacy, and applied
the title Emperor to the Coweta chief. In fact, in
early colonial documents, the name Coweta was
often used as a synonym for "Lower Creeks."
Also, the Cherokee of that period called the
Lower Creeks "Ani-Kawita," just as they called
the Upper Creeks "Ani-Kusa," after the princi-
pal towns of Coweta and Coosa respectively.
The early Coweta chief, Brims, was called Em-
peror by both the English and the Spanish.
Some time between 1628 and the early 1670's,
Cofitachiqui vanished from Silver Bluff. Appar-
ently, it removed to northern Georgia for a brief
while, and then lost its identity completely. Be-
fore this removal, all the evidence, both lin-
guistic and historical, identifies Cofitachiqui
as Coweta, a ceremonial and political center of
the Lower Creeks. No evidence, antedating the
removal of Cofitachiqui from Silver Bluff, would
associate this town with the Kasihta, who in
proto-historic times (as in colonial times) were
living more than 200 miles west of the Savannah.
How did "Cofita," in pronunciation so much
like "Coweta," ever come to be translated as
"Kasihta"? (The Creek pronunciation of the
latter cannot be indicated with English charac-
ters; it is approached by the British colonial
"Cassista" and the Spanish "Casixte.")
In 1674 a colonial document mentioned "Cussi-
taws," and in 1679 another mentioned "Cussa-
toes." Colonial officials had a very casual ap-.
proach to spelling. Also, at this period, in both
manuscript and type, the sound of "s" in certain
positions was represented by a symbol resem-
bling the letter "f." Perhaps some colonial
emissary, laboriously trying to pen the name of
an Indian town, wrote "Cuffitaw," which was
misread by some scribe as Cussitaw. In any
event, the documents said nothing about the
"Cussitaws" being at Silver Bluff or even on
the Savannah; they were simply considered as a
likely tribe to inflame and send against the
troublesome Westo, who had settled above Agus-
ta. The "Cussitaws" were described as "power-
ful" in the two aforesaid documents; and on this
meager basis alone, Swanton (1922: 217-218)
insisted that Cofitachiqui was Kasihta even
though he realized that this identification called
for two Kasihtas in simultaneous existence, one
on the Savannah and the other in Alabama! Nor
did Swanton explain why colonial officials
should- consider sending Kasihta, the peace
town, against the war-like Westo; yet should
make no mention of Coweta, the war town that
handled military matters for all the Lower Cre-
eks. Today it seems likely that Kasihta, which
received no more than bare mention in early
colonial documents, was not "powerful" in the
1670's. Emperor Chekilli, the Coweta cheif of
the Lower Creeks, told Oglethorpe that the Kas-
ihta "first saw the red smoke and the red fire,
and made bloody towns." Chekilli continued
that now (in 1773) the Kasihta "know the white
path was best for them" (Swanton 1928a: 38).
The remarks clarify a circumstance that baffled,
Swanton (1922: 222): map references to Kasihta,
in the latter 1600's, as "Gitasee." The latter
term is simply the Creek word for "red," and
its use emphasizes that Kasihta was not of the
"white" moiety, not a peace town, at this time.
William Bartram's informants looked on Apa-
lachicola as the true peace town of the Lower
Creeks (Van Doren 1928: 313), and it is likely
that Kasihta became a white or peace town only
after the (pro-Spanish) Apalachicola settle-
ments had been destroyed by the British in 1706
Swanton (1922: 216; 1928a: 33-78), trying to
identify Cofitachiqui as Kasihta, gave some
weight to Creek migration legends. These tales
have remarkably little in common with each
other. -They reveal little beyond the Creek al-
legiance to the home town, and the Indian tend-
ency to reject a factual explanation in favor of
mysticism and metaphor. In any event, none of
the tales identified Cofitachiqui as Kasihta; the
best any of them did was to have the Kasihta
leading the Coweta in an eastward movement
through some mythical land at some time in the
Swanton eventually came to modify his earlier
views. In 1946 (126-127), he admitted that Cow-
eta was the principal war town of the Lower
Creeks; that the whites regarded Coweta as the
capital of the Creek Confederacy; that the Creek
"Emperor" (who in several accounts was stated
to reside at Cofitachiqui) was the Coweta chief;
that Coweta (unless it be Cofitachiqui) is not
mentioned in early Spanish documents although
it should have been; that the Kasihta were in
Alabama in the time of De Soto and De Luna;
that the Kasihta could have been west of the
Coweta in spite of origin myths. Although ob-
viously aware that "Cofitachiqui" was com-
pounded from "Cofita" plus "chiqui," Swanton
never mentioned the possibility that "Cofita"
was a rather good attempt at the Indian pronun-
ciation of "Coweta;" but otherwise he marshal-
led all the impressive evidence showing Cofit-
achiqui to have been Coweta. Still, he chose not
to reject the guess he had made 24 years before.
But at least he admitted, "it is even possible
that they (the Indians of Cofitachiqui) were the
Coweta, and that the Kasihta are represented
by a tribe which the Spaniards met on the Ala-
bama" (Swanton 1946:143).
Thanks to Swanton's pioneering efforts, and
the foundation they have provided for later work,
we can now be more definite in ruling out the
Kasihta as former residents of Cofitachiqui at
Silver Bluff. In short, Cofitachiqui -- a name
probably meaning "Coweta earthworks" -- is
here regarded as a Coweta ceremonial and poli-
tical center of the Lower Creeks; and is specif-
ically identified with earthworks formerly evident
at Silver Bluff.
As noted by Swanton (1946:45), Cofitachiqui
is placed at Silver Bluff partly on the strength
of information obtained by George Galphin from
unspecified Indians. Galphin's connections with
the Coweta group suggest that these Indians
supplied the information. They are the most
likely people to have done so if Cofitachiqui
had been a Coweta town. We do not know the
exact year that Cofitachiqui removed from Silver
Bluff, or the exact year that Galphin arrived
there; but the two events could have been sep-
arated by no more than 70 years. The Coweta's
18th century interest in, and association with,
Silver Bluff probably did not come about solely
through the machinations of colonial officials.
The Coweta might well have retained a proprie-
tary interest in a former ceremonial center. (It is
also worth mentioning that Galphins' remarks
are supplemented by Spanish and British col-
onial documentary evidence that Cofitachiqui
stood at Silver Bluff.)
In 1776 William Bartram made his first visit to
Galphin's trading past. Nearby, he saw aborigi-
nal earthworks, including Indian conical
mounts, terraces, areas, etc., as well as remains
or traces of fortresses of regular formation"
(Van Doren 1920: 259).
This sounds much like an early Creek cere-
monial ground, of the kind described and illus-
trated by Bartram in an account of 1789. (This
latter account, published in 1853, is not widely
available, but an illustration from it appears in
Bushnell 1919: 75, and in Swanton 1928b: fig 3).
Bartram thought that "Mr. Golphin's (sic) build-
ings and improvements will prove to be ... of
infinitely greater celebrity and permanency"
than the Indian earthworks; but such was not
the case. The brick structure at Fort Galphin
was still standing in the early 1870's (Jones
1873: 152), but it later fell into ruins. By the
20th Century, Galphinton had vanished. The
exact location of the post was remembered, but
only as a place to pick up "Indian beads."
These were actually pipestem fragments, mis-
identified as beads. Galphin must have stocked
(and broken or failed to dispose of) vast quanti-
ties of kaolin trade pipes, to judge from the
countless fragments found at the site by curio
hunters. Bits of pipestem were still weathering
out in a circumscribed area atop the bluff when
Chapman Milling and his family visited the site
in 1939; they picked up "hundreds" there (Mill-
ing: 1940: 151).
THE YUCHI NEAR SILVER BLUFF
Eaton (1962) measured the stemhole diameter
of 268 pipestem fragments from Silver Bluff. He
analyzed these measurements in terms of Bin-
ford's (1961) formula, which is based on the
change, with time, of stemhole diameter. The re-
sulting calculation suggested the year 1744 as
the temporal midpoint of occupation (by pipe-
users) at Silver Bluff. The figure is too early
for the midpoint of Galphin's operations, and
Eaton suggested earlier occupancy of the site
by the Yuchi. He documented this suggestion by
Milling's (1940: 179) reference to "the Yuchi,
of Silver Bluff," and by Quattlebaum's (1956:
37) identification of Cofitachiqui as a Yuchi
town. However, Milling did not intend to place
the Yuchi exactly at Silver Bluff; he knew their
village actually was two miles below the bluff,
for he quoted (1940: 185) James Adair's state-
ment to that effect. (The trader Adair, inciden-
tally, was a close friend and admirer of Galphin,
and is the best authority on the exact location
of the Yuchi town (see Williams 1930). Also, the
identification of Cofitachiqui as a Yuchi settle-
ment was discredited by Speck (1909: 7) and
Swanton (1922: 287).
Nor can it even be shown that the Yuchi vil-
lage, two miles below Silver Bluff, was occupied
prior to Galphin's arrival. There was an early
Yuchi settlement on the Savannah River about
five airline miles above Augusta (Neill 1955: 5),
and another at "Savannah Town" (New Savan-
nah, Savaneton, site of Fort Moore and an old
Shawnee town) just below Augusta, on the South
Carolina side of the stream (Milling 1940: 182).
These two towns were, of course, well upstream
of Silver Bluff; and well downstream, there was
the Yuchi village of Mount Pleasant.Other Yuchi
settlements were scattered along the lower
course of the Savannah roughly from Mount
Pleasant to Ebenezer, the abandoned Salzburger
town (Speck 1912: 1004; Swanton 1922: 308).
Several vague, early references, to Yuchi on the
Savannah, cannot be held specifically applicable
to the village two miles below Silver Bluff.
Where not vague, the references pertain clearly
to the settlements on the outskirts of Augusta,
or to the ones far downstream. Until 1740, colo-
nial policy permitted the Yuchi and other Indians
to state their land claims on the Savannah. At
one time or another the Yuchi claimed the vil-
lage sites near Augusta, and those on the lower
Savannah; but they never claimed Silver Bluff.
Probably, then, they did not settle near the bluff
until some time after 1740.
In 1743 a trader warned that the Lower Creeks
were becoming aroused against certain Yuchi;
and the next year, a Yuchi delegation visited
Charleston to ask for protection (Council Jour-
nals of South Carolina, X: 34; XI: 286-287). In
1746 a Yuchi band, under British colonial pro-
tection, was attacked by unidentified Indians,
supposedly Creeks at French instigation (Public
Records of South Carolina, XXII: 151). These
references probably relate to the Yuchi band
whose history particularly concerns us here. It
is likely that, after this raid, these Indians set-
tled near Galphin's newly founded establishment;
for not until 1750 do we hear definitely of Yuchi
living two miles below Silver Bluff. Adair re-
ported a raid on these Yuchi in 1750, again by
Indians supposedly at French instigation. And
in 1751, James Glen, the governor of South Caro-
lina, reported that the Yuchi settlement near
Silver Bluff had been abandoned (Milling 1940:
185-186). I suggest that this band of Yuchi fled
to the Silver Bluff area after the raid of 1746,
and departed soon after the raid of 1750. The
circumstance (brief residence near Silver Bluff)
would explain why colonial documents have so
little to say about Yuchi in the vicinity of Gal-
phin's important establishment.
It is, in fact, illogical to expect the Yuchi near
Silver Bluff before Galphin was well established
there. Hostile toward the Shawnee, Cherokee,
Chickasaw, and Catawba, and at this period not
on the best of terms with the Lower Creeks, the
Yuchi originally moved into British territory for
protection; this is made clear by the writings of
several colonial officials, such as James Glen
(Milling 1940: 184). Protection could be had in
the immediate vicinity of Augusta, for this town
had, even before 1740, a good 600 people in-
volved with the Indian trade; and across the river
was Fort Moore, begun in 1716. Protection could
also be had on the lower reaches of the Savan-
nah, for this area was comparatively far removed
from the territory of the tribes with whom the
Yuchi were actually at war. Also, on the lower
Savannah or not far away were many colonial
settlements: Savannah, Yamacraw, Pipe-maker's
Bluff, Fort Argyle, Old and New Ebenezer, Old
and New Inverness, and others. On the lower
Savannah, too, were the expatriated Apalachi-
cola, on whom the colonial officials kept a
watchful eye; and the Yuchi had been encouraged
by these officials to settle beside the Apalachi-
cola. In contrast, Silver Bluff was unprotected
until Galphin began to build there. Over land or
by water, it was at least a hard day's journey
through wild and heavily forested country from
any colonial center. The locality was highly
vulnerable from a military standpoint: the bluff
was low, and easily flanked; there was no en-
circling swamp; comparatively high, forested
ground formed a broad highway to the river.
Silver Bluff was never protected, in colonial
times, by anything but colonial influence, and
such influence did not encompass the bluff until
Galphin began to build there. And as it hap-
pened, even near Silver Bluff the Yuchi were not
safe from the shadowy raiders with (supposedly)
PIPESTEM DATING AT SILVER BLUFF
With the Yuchi out of the picture, how can we
account for the fact that the pipestem date of
1744 is too early to represent the temporal mid-
point of Galphin's activities?
While there was some early influx of settlers
into the Savannah River valley from coastal
South Carolina (for example, to the Margravate
of Azilia and to the trading post at Fort Moore),
most settlement of the valley was not continuous
but rather by waves: Oglethorpe's first colonists,
Yoakley's first supply ship, the second supply
ship with a big load of settlers, the Israelites,
the Salzburgers, the Swiss and the Moravians,
the Scotch Highlanders, Oglethorpe's second
colonists, Whitefield and Habersham's group in-
cluding a troop of soldiers, the emigrants from
Massachusetts, etc. Ships brought not only colo-
nists but their supplies. Trade goods probably
arrived, not in a continuous flow but in bursts.
Also, colonial settlements in the Savannah
River valley were threatened by, and at war with,
the Spanish between 1739 and 1742. During this,
the War of Jenkin's Ear, Oglethorpe attacked but
failed to take St. Augustine, and then retired to
his own base, Fort Frederica on St. Simon's Is-
land, Georgia. From here he directed other forays
against the Spaniards. As emphasized by Manucy
(1962: 1), Frederica was the very heart of the
British frontier defense system. Here attack was
expected, and here it came, in 1742. A Spanish
fleet breached the harbor defense, and set troops
ashore; but their advances were thrice repelled,
and the Spaniards fell back to Florida.
With the Spanish repulsed, and the Lower
Creeks contented to abandon their claim to the
Savannah River drainage, the future of the colo-
nists looked bright; it is not surprising that
someone--perhaps a colonial official, or Lachlan
McGillivray, or George Galphin himself--was
emboldened to order a large shipment of kaolin
pipes. It must be recalled that the stemhole
diameter is determined when the pipe is made;
no mathematical treatment, however elegant, will
reveal when a pipe was traded or used, because
such an item might repose in storage for one
year or forty.
It is also possible that Galphin, at the start of
his operations, overstocked his post with trade
pipes; for there is no indication that the Lower
Creeks at this period were much given to smok-
ing, except in ritual connection.
The market for pipes was further limited, at
least for a few years, by the general absence of
Negroes from most of the area supplied by Gal-
phin's post; for when the trader began operations,
Georgia was the only colony that prohibited slav-
ery. It also prohibited sale of alcoholic bever-
ages at this time. As a result of these prohibi-
tions, new immigrants into America began to
avoid Georgia. At popular insistence, slavery
was legalized in this colony in 1749, and the
sale of alcoholic beverages soon thereafter; but
not until the latter 1750's did white and Negro
settlement begin to increase noticeably in the
area supplied by Galphin.
Thus, there is nothing perturbing in Eaton's
discovery that 90% of the pipestem fragments
from the trading post site seemed to date from
early in Galphin's career there.
I have followed other authors in assuming that
kaolin pipes, recovered from British colonial
sites in the southeastern United States, were
made in the British Isles. However, "Pipe-mak-
er's Bluff" was among the first Georgia local-
ities to be named in colonial times. The local-
ity is on Pipe-maker's Creek, a tributary of the
Savannah in Chatham County, Georgia. It is sh-
own as a settlement on Thomas Wright's 1763
map of Georgia and Florida, but was named much
earlier. When Oglethorpe visited the Coweta in
1739, to clarify their land cessions on the Savan-
nah, Pipe-maker's Bluff was mentioned in the
discussion (Lowrie and Franklin 1834).
Kaolin deposits, many of them mined commer-
cially, stretch across the Atlantic Coastal Pla-
in of Georgia, from the Ocmulgee to the Savan-
nah, and into Aiken County, South Carolina
(Veatch and Stephenson 1911). If someone made
kaolin pipes at Pipe-maker's Bluff on Pipe-mak-
er's Creek, he need not have followed English
pipe-makers in using progressively smaller stem-
hole-reamers as the years went by. His original
stock of reamers may have lasted, and been used,
THE TRADING POST SITE IN 1946
When I first visited Silver Bluff, in 1946, it
and the surrounding country were but sparsely
inhabited. Some Negroes, living in scattered cab-
ins, had put in corn patches here and there. Many
of these people, known locally as "Red Negroes
", showed the physical signs of some Indian an-
cestry; but they retained no tradition thereof.
Several of the women wore their straight, black
hair in long braids.
The general area consisted mostly of over-
grown fields, with isolated stands of second-
growth pines and hardwoods. Bartram's earth-
works had long since been plowed out of exist-
ence; but not far from the trading post site there
was a sizeable expanse with traces of aborigi-
nal occupation. A Negro, who could have passed
for Indian, occupied a small cabin at one side
of the trading post site. In the field between his
cabin and the edge of the bluff, kaolin pipe frag-
ments were still weathering out of the soil.
Soldiers, on training maneuvers during World
War II, had dug a ramp into the bluff, leading
from high ground to the river's edge. The ramp
and associated diggings had cut through refuse
pits full of charcoal, freshwater clam shell, vert-
ebrate remains, and artifacts of both Indian and
white manufacture. The vertebrate remains in-
cluded freshwater turtles (Pseudemys); and some
large bovine animals, either cattle or bison. (Bi-
son were present as far east as the Savannah in
Galphin's day.) I was not in the area for arche-
ological investigation, and spent but little time
at the pits. From their exposed faces, I extract-
ed the more conspicuous aboriginal sherds (Figs.
1-2); two scraps of hard-glazed crockery, yel-
low with brown under-glaze designs; a scraper
of dark green glass, made from a bottle fragment;
two quartz hammerstones, much battered; a few
pieces of kaolin pipes; a clasp knife and a pair
of scissors; both badly rusted; a handforged
nail; a similar nail bent into fish-hook shape;
and a sample of vertebrate remains. Some of this
material -- crockery, glass scraper, bovine re-
mains, and hammerstones -- is illustrated
FROM THE TRADING POST SITE
The 11 aboriginal sherds were all extracted
from the cut face of one refuse pit. They are hard
and well-made, tempered with grit and shell. On-
ly one sherd has much more than a trace of shell
tempering, however. Reddish inclusions and car-
bonaceous granules also appear in the paste.
The surface is buff, tan, or brown; it is well
smoothed, but interiors show tool marks. Inois-
ing has been lightly smoothed over. Paste cores
are dark. At least one sherd may have had a thin
clay wash. There is no indication that the ves-
sels were ever re-painted. Vessel shapes includ-
ed a cazuela bowl, and a jar with a short, con-
stricted neck. Jar rims are outflaring, and in one
case (Fig. 1, lower right) the lip is thickened
exteriorly. All bowl fragments (Fig. 2) are note-
worthy for a fillet, separating the upper, dec-
orated portion of the vessel from the lower, pl-
ain portion. The fillet is notched or indented at
close intervals. (Compare Willey 1949: 494, on
Ocmulgee Fields Incised, a Lower Creek pot-
tery type best known from central Georgia.)
Fig. 1. Incised (left) and plain jar sherds from Galphin's trading post.
*: ; *
Fig. 2. Bowl fragments, all incised, from Galphin's trading post.
r ;I '
Fig. 3. Left row: Two crockery sherds and (at bottom) a glass scraper. Center row: Lower jaw fragments
from cattle or bison. Right row: Quartz hammerstones. All from Galphin's trading post.
The predominance of rim sherds in the Silver
Bluff sample reflects my selectivity; many sm-
all, plain body sherds were left in place. Extra-
polating from the curvatures of the collected sh-
erds, jar mouths ranged in diameter from about
5'% to 7 inches. Jar body diameter exceeded
mouth diameter by about 2 inches. Bowl mouths
ranged in diameter from about 12/2 to 13 inches,
the bowl diameter exceeding mouth diameter by
about an inch.
Fundaburk and Foreman (1957: pl. 137) figured
a series of vessels from the Creek town of Tulsa
on the Tallapoosa River in Alabama. One of
their bowls (bottom row, right, in their illustra-
tion) duplicates two of the bowl fragments from
Silver Bluff. Another of their specimens (second
row from top, right), an asymmetric bowl, is cov-
ered with roughly parallel incisions that run di-
agonally from upper right to lower left. The same
decoration appears on a Silver Bluff rim sherd
(probably from one vessel, a jar).
The incised designs are simpler in the Silver
Bluff sample than in Ocmulgee Fields Incised
from central Georgia (Fairbanks 1956: pl. 25) or
in McKee Island Incised from east-central Ala-
bama (Morrell 1965: 28-29. 31). Both Ocmulgee
Fields Incised and McKee Island Incised are
likely to be. at least at their respective type
sites, several decades older than the Silver
Bluff sample. The circumstance may reflect a
continued loss of interest, on the part of the Cre-
eks, in the surface decoration of their pottery.
Such loss of interest may have begun in late pre-
historic times (compare Lamar pottery with its
presumed Macon Plateau antecedents), but was
probably accelerated in historic times by the
ever-increasing availability of crockery, iron
pots, glassware, and other trade goods.
George Galphin was involved with the Lower
Creek trade at least as early as 1744, and is
mentioned in connection with Silver Bluff from
about 1739. There he built a trading post, barge
docks, a way-station for river travelers, holding-
pens for livestock, and no doubt cabins for his
family, associates, and numerous slaves. A
stockaded brick building was added when the
American Revolution became imminent. Galphin's
activities went far beyond simple trading, and
his theater of operations extended to Charleston,
Savannah, St. Augustine, Pensacola, and Mobile-
in other words to the borders of Spanish and
For a short while, an obscure band of Yuchi
occupied a village two miles below Silver Bluff.
They may have arrived there after 1746. They
are definitely mentioned only as being present
in 1750, and as having departed in 1751. Earlier
references to Yuchi on the Savannah River pro-
bably relate to other villages, either near Aug-
usta or on the lower reaches of the stream.
Galphin's attentions were primarily directed
not toward any Yuchi, but toward the powerful
Coweta or "war town" group of the Lower Cre-
eks. He was, by official records, the trader to
and for the Coweta. He married into this group,
and his half-Indian son, John, attained consider-
able stature among them. The Coweta brought
their trade goods to Silver Bluff, where they and
other Indians also received gifts and annuities.
Aboriginal earthworks at Silver Bluff were
identified by Indians as Cofitachiqui, a proto-
historic town mentioned in narratives of the De
Soto and the Pardo expeditions, as well as in
later accounts. Cofitachiqui was a Lower Creek
town, and can be identified as Coweta. Its name
may mean "Coweta earthworks." Early colonial
documents state that the "Emperor" of the
Creeks resided at Cofitachiqui. Colonial writers
applied the term "Emperor" to the Coweta chief
only, and regarded his town as the capital of the
Creek Confederacy. Cofitachiqui vanished from
Silver Bluff some time between 1628 and 1670.
George Galphin's death, in 1780, did not termi-
nate activities at his post. His holdings were
retained by his heirs after the Revolution. In
1785, the Treaty of Galphinton was staged at
Silver Bluff; by its terms, two Lower Creek
chiefs (out of the many) ceded further lands to
the commissioners of Georgia. Thereafter, noth-
ing more is heard of Lower Creeks at Silver
It is not perturbing that pipestem fragments
from the trading post site, dated by stemhole
diameter, supplied not the temporal midpoint of
Galphin's occupancy but rather the beginning of
it. Galphin may have purchased a large shipment
of pipes when he first began operations; and
pipes can repose indefinitely in storage. It is
not out of the question that the pipes were made
in colonial Georgia, by someone using "antique"
equipment at Pipe-maker's Bluff.
Eleven aboriginal sherds, recovered from a re-
fuse pit at the trading post site, are close to
Ocmulgee Fields Incised, although exhibiting
minor departures from the norm thereof. The
sherds are most likely to be Coweta. They
should date between 1739 and 1785, and most
likely between 1750 and 1770 (the period of max-
ium Lower Creek activity at Silver Bluff). They
are more casually decorated than is type-site ma-
terial of either Ocmulgee Fields Incised or Mc-
Kee Island Incised. The Silver Bluff sample is
likely to have been made several decades later
than the aforesaid type-site material. The con-
tinued flow of trade goods probably resulted in
progressive loss of interest, on the part of the
Creeks, in pottery decoration.
From a geographic standpoint, Silver Bluff is
but a minor landmark; yet it has witnessed some
highly significant episodes in American history.
Cofitachiqui was an important center of the Cre-
eks, who in late prehistoric times had extended
their influence widely in the Southeast. The town
saw the arrival of early Spanish explorers, whose
coming portended a decline in the Indian way of
life. Galphin's trading post, a major center of
colonial trade and politics, saw the Revolution-
ary War, and the termination of British control
over the American colonies. Finally, near the
midpoint of the present century, access to the
site was terminated, at least temporarily; for a
few miles down-stream, a huge reservation was
being mapped out to encompass a hydrogen bomb
plant, symbol of yet another era.
Binford, Lewis R.
1961 A New Method of Calculating Dates from
Kaolin Pipe Stem Fragments. (Paper pre-
sented at) Second Annual Conference on
Historic Site Archaeology. Macon.
Bushnell, David I., Jr.
1919 Native Villages and Village Sites East
of the Mississippi. Bureau of American Eth-
nology, Bulletin 69. Washington.
1962 Pipe Stem Dating and the Date for Silver
Bluff, S.C. The Florida Anthropologist, Vol.
15, No. 2 (misprinted No. 1): 57-62. Talla-
Evans, Lawton B.
1913 First Lessons in Georgia History. Ameri-
can Book Company. New York.
Fairbanks, Charles H.
1956 Archeology of the Funeral Mound, Ocmul-
gee National Monument, Georgia. National
Park Service, U.S. Dept. Interior. Washington
Fundaburk, Emma L. and Mary D. F. Foreman
1957 Sun Circles and Human lands. Emma L.
Goggin, John M.
1958 Seminole Pottery. In: Prehistoric Pottery
of the Eastern United States. Museum of An-
thropology, University of Michigan. Ann
1940 Galphin Claim. In J. T. Adams and R.
V. Coleman (editors), Dictionary of Ameri-
can History, 2nd, edition, Vol. 2, p. 368.
Charles Scribner's Sons. New York.
Jones, Charles C., Jr.
1873 Antiquities of the Southern Indians, par-
ticularly of the Georgia Tribes. D. Appleton
and Company. New York.
Kinnaird, Lawrence (editor)
1946 Spain in the Mississippi Valley, 1765-
1794. Part 3. Problems of Frontier Defense,
1792-1794. Annual Report of the American
Historical Association for the Year 1945.
Loughridge, R. M. and David M. Hodge
1914 English and Muskokee Dictionary Col-
lected from Various Sources and Revised.
Westminster Press. Philadelphia.
Lowrie, Walter and Walter S. Franklin (editors)
1834 American State Papers (etc.). Class II,
Indian Affairs. Vol. II. Washington.
Manucy, Albert C.
1962 The Fort at Frederica. Florida State
University, Notes in Anthropology, Vol. 5.
1811-1816 The History of Georgia, Containing
Brief Sketches of the Most Remarkable
Events, Up to the Present Day. 2 vols.
Meriwether, Robert L.
1949 The Expansion of South Carolina 1729-
1765. Southern Publishers, Inc. Kingsport.
Milling, Chapman J.
1940 Red Carolinians. University of North
Carolina Press. Chapel Hill.
Morrell, L. Ross
1965 The Woods Island Site in Southeastern
Acculturation 1625-1800. Florida State
University, Notes in Anthropology, Vol. 11.
Myer, William E.
1928 Indian Trails of the Southeast. Bureau of
American Ethnology, Annual Report for
1924-1925: 727-857. Washington.
Neill, Wilfred T.
1955 An Historic Indian Burial from Columbia
County, Georgia. Southern Indian Studies,
Vol. 7: 3-9. Chapel Hill.
1956 The Land Called Chicora. University of
Speck, Frank G.
1907 The Creek Indians of Taskigi Town.
Memoirs of the American Anthropological
Association. Vol. 2, Pt. 2. Lancaster.
1909 Ethnology of the Yuchi Indians. Anthro-
pological Publications of the University
Museum, University of Pennsylvania, Vol. 1.
No. 1. Philadelphia.
1912 Yuchi. In: Frederick W. Hodge (editor),
Handbook of American Indians North of
Mexico. Part 2: 1003-1007. Washington.
Swanton, John R.
1922 Early History of the Creek Indians and
Their Neighbors. Bureau of American Ethno-
logy, Bulletin 73. Washington.
1928a Religious Beliefs and Medical Practices
of the Creek Indians. Bureau of American
Ethnology, Annual Report for 1924-1925:
1928b Aboriginal Culture of the Southeast.
Bureau of American Ethnology, Annual Re-
port for 1924-1925: 673-726. Washington.
1946 The Indians of the Southeastern United
States. Bureau of American Ethnology, Bul-
letin 137. Washington.
Van Doren, Mark (editor)
1928 Travels of William Bartram. Macy-Masius.
Veatch, Otto and Lloyd W. Stephenson
1911 Preliminary Report on the Geology of the
Coastal Plain of Georgia. Foote and Davies
1855 Historical Collections of Georgia. 3rd
Edition. Pudney and Russell. New York.
Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast.
Smithsonian Miscellaneous Collections,
Vol. 113. Washington.
Williams, Samuel C.
1930 Adair's History of the American Indian.
Watauga Press. Johnson City.
New Port Richey, Florida
VOLUME 21, NOS. 2 & 3
THE FLORIDA ANTHROPOLOGIST
THE DUPONT PLAZA SITE
D. D. Laxson
JUNE -SEPTEMBER, 1968
Representative Tequesta material was salvaged
from the area now occupied by the Dupont Plaza
complex at the mouth of the Miami River. Occu-
pation to Glades I times is possible and large
areas under present-day parking lots may be
excavated in the future.
The area of hammock and limestone at the
mouth of the Miami River as it enters Biscayne
Bay is of both prehistoric and historic signifi-
cance. In the last several thousand years it has
seen, along with geological and ecological ch-
anges, Indians, hurricanes, saints and sinners,
soldiers and sailors, hard working visionary
pioneers and some of the richest people in the
Since there are excellent primary and second-
ary sources available concerning the historical
period of the area, it is felt anything but a mini-
mum effort along these lines would be redundant.
Therefore, information will be limited to the pre-
historic, and the last 75 years of the late his-
While there are Indian kitchen middens along
the courses of all the streams such as Snapper
Creek (Goggin 1950), Arch Creek (Laxson 1957),
Little River (Laxson, 1959) and the Uleta River
(Laxson 1957) draining the Everglades, the large
village at the mouth of the Miami River is best
documented. The site is referred to as Miami
Midden No. 1 and has been the subject of several
articles (McNicoll 1941; Squires 1941; Goggin
1949: 83; and Laxson 1959).
Surface collections indicate an area roughly
bound by N. Miami Avenue on the west, Brickell
Park to the south, and NE 2nd Avenue to the
north still retaining representative material of
the early inhabitants, but it is difficult today to
positively determine the boundaries of the ori-
ginal Tequesta village.
Two large buildings have been constructed in
this area in the last 75 years. The inevitable
destruction of the Indian site began with the
first, the industrialist Henry Flager's Royal
Palm Hotel (Fig. 1, RP).
Ground that had been previously cultivated was
first broken on March 15, 1896 and the top taken
from what was then called a "Seminole burial
mound". Numerous bones, beads, and trinkets
were hauled away in wheelbarrows to be used
as fill in a large hole thought to have been on
the approximate location of the present day In-
graham Building at NE 2nd Avenue and NE 2nd
The hotel, a wooden, yellow and white, some-
what ostentatious structure, adorned with the
"gingerbread" of its time, opened its doors to
what was to be a long list of notables on Jan-
uary 15, 1897. With the help of the 1926 hurri-
cane and souvenir hunters, it was demolished by
It was announced in 1953 that the area, then
covered with parking lots and dotted with coco-
nut palms, had been leased. By August of 1956,
the Arkin Consturction Company had begun work
just south of the old Royal Palm grounds on the
Dupont Plaza complex (Fig. 1, d) which was
officially dedicated on March 16, 1958.
Through the efforts of the late Hugh Davis,
then President of the South Chapter of the Flo-
rida Anthropological Society, and Marvin Brooks,
the chapter Secretary, permission was granted to
start test excavations. These began on Wednes-
day, August 22 and ended September 30, 1956.
Since that time, several supplementary tests
have been made, and Mrs. Paul Agnew made a
surface collection of the area in 1964. The late
John Hackett and the author, in order to deter-
mine village boundaries, made surface collec-
tions along the river bank and the grounds of the
Towers and Granada Hotels. The results are
designated as Area A on both the pottery distri-
bution chart (Table 1) and Fig. 1. Tests made
in the vicinity of the now demolished Robert
Clay Hotel are shown as Area B on the distri-
butioi chart and in Fig. 1. A small test pit (Fig.
1. A small test pit (Fig. 1, p) was excavated by
the author on the J.C. Ray property located be-
tween SE 5th Street and the river bank south of
the Brickell Avenue bridge. The area was part-
ially paved for parking and planted with Austra-
lian Pine, palm, ficus, and gumbo limbo.
Truck loads of dirt were followed to the vicin-
ity of 2100 South Bayshore Drive and material
screened out as shown for Area C, Tables 1-2.
By February of 1957 the Dupoft Plaza material
had been washed, sorted and cataloged and
efforts were being made to determine its value.
I Flogler St. /
S.E. Fifth St. $
Fig. 1. The approximate location of Miami Midden
No. 1, the Royal Palm Hotel and the pre-
sent day Dupont Plaza with areas contain-
ing Tequesta material.
Surface collections made on the grounds of the
Towers and Granada Hotels (Fig. 1, a) turned up
late pottery types although earlier Key Largo
Incised and Ft. Drum Incised was found under
the river bank, down to an approximate depth of
The largest samples of Matecumbe Incised
came from the grounds around Dallas Park, Rob-
ert Clay Hotel (Fig. 1, b), but due to the fact
that the sherds were similar in color and temper
and were scattered over only a few hundred feet,
it is possible that they were all from the same
The most sherds from any test came from the
dirt taken to the vicinity of 2100 S. Bayshore
Drive to be used as fill. Their tabulation is
shown under Area C, Table 1. All Glades pottery
types were represented, including two sherds
that might be classified as semi-fiber-tempered
since their centers showed signs of burned out
vegetable material. Unfortunately, neither their
depth in the midden nor their association with
other artifacts could be determined.
The wall of a large foundation pit (Area D) near
the present-day north facade of the building was
worked, but material was under what resembled
old paving and tamped-down limestone rock.
Stratigraphy was poor as most material was "up-
side down," although the largest collection of
Belle Glade sherds came from this part of the
midden. However, the west wall of a pit adjacent
to this area was excavated satisfactorily when
an undisturbed pot-hole was found and screened
to approximately 30 inches. The results are
shown in Tables 1-2.
The small pit (Fig. 1, p) excavated on the
river's south bank yielded satisfactory strati-
graphy and a single Opa Locka Incised sherd,
elaborately decorated with parallel, vertical
rows of thickly nested horizontal arcs or semi-
circles. Belle Glade pottery was represented by
a single sherd in each of three levels.
As could be expected considering the environ-
ment,shell tools were plentiful with the Strombus
scraper, a basic tool, most plentiful. The amount
of Busycon tools, which included a single con-
cave scraper, hints at the possibility that this
shell was previously found in greater numbers
than at present in this locale. In the area exca-
vated, Macrocallista shells were strangely ab-
sent from the tests. A single "plumb bob" vari-
ety of shell plummet was found in the Bayshore
Shark vertebrae were numerous with some
worked to be strung as beads or possibly net
handlers. A projectile point and human, turtle,
and deer bones were salvaged from most of the
As for historical artifacts, Krag carbine cart-
ridges found were probably dropped by troops
billeted in the area during the Spanish-American
War. Broken tableware was common, as was
typical brown "jugware." Two pieces of majol-
ica were found, one in the fill and one in the
6-12 inch layer of the wall of the large pit.
Champagne bottles, assumed from their shape,
were 9 inches long and 3 inches wide at the
center. The bottom was drawn out to a sharp
point, presumedly to be stored flat to keep the
cork wet, and their only identification was the
letter "K" near the bottom. The other bottles
were 9 inches long, 21/2 inches in diameter, the
bottom rounded somewhat but able to stand.
Their only identification was "JK&S 2117"
TABLE 1. CERAMIC DISTRIBUTION
A B C D
St. Johns C. St.
Opa Locka Incised
Ft. Drum Incised
Belle Glade Plain
Zig-zag on rim
Miami Incised var.
Finger nail Incised
Key Largo variants
St. Johns Plain
Key Largo Incised
SE 5th St. Pit
! I I N
Ul 0 '0 rH
8 51 41 11
1 4 4 1
1 1 1 -
- 1 -
- 1 1 1
1 1 -
- 1 8
e e 3
N 0 q
H- r- N
20 9 5
4 1 2
- 1 -
- 1 -
3 1 -
TABLE 2. ARTIFACT DISTRIBUTION
Artifacts Areas SE 5th St. Pit facade pit
A B C D
Buscyon scrapers 1 -
Busycon pick 1 2 1 -
Strombus scrapers 1 1 1 7 1 1 11 1
Busycon receptacle 1 1 1 -
Strombus pounder 1 -
Shark vertebrae 1 1 1 6 1 1 11 1 1
Manatee bone 1 -
Seminole beads 6 2 -
Bone points 1
Bone china 2 -
Majolica 1 1 -
Jugware 1 2 -
30-40 Krag shell 2 -
Champagne bottle 2 1- -
Beer bottle 1 1- -
Soft drink bottle 1 -
Sandstone ball 1 -
Limestone hone 1 -
RR ballast 2 2 -
Bricks 1 1 3- -
Broken chinaware 13 22 9 7- -
Cast iron fragments 1 1 1 1 1 -
Shell plummet 1 -
Greenstone celt 1 -
Coal 4 2 2 1 -
Fig. 2. Members of the local chapter of the Florlaa Anthropological Society screen fill from the Dupont
around the bottom rim, with the letter "W" in
the center of the bottom. Necks were molded for
Bricks, railroad ballast, coke, coal and several
large timbers resembling split pilings were also
found. On the eastern rim of the building site
large patches of gray clay were uncovered.
Some things became apparent almost immed-
iately: (1) The site was prolific (Fig. 2); (2)
there was a continuation of the midden on the
south bank of the river (Fig. 1, BP); and (3) the
site was occupied not only during Glades III
times as was supposed (Goggin 1949: 83), but
extended to the Glades II period.
The SE 5th Avenue pit and the large pot hole,
with the usual Glades Tooled and St. Johns
Check Stamped stratified above Key Largo In-
cised and plain sherds at a depth of 30 inches,
suggested the ill-defined Glades I period.
Clay samples test fired proved too calcareous
to be used for ceramics.
Considering the location of the site near the
ocean and its proximity to heavily wooded ham-
mocks, subsistence could hardly have been a
Belle Glade pottery indicated trade up-river to
the region of Lake Okeechobee, but for the most
part incised designs were confined to basic
types. Considering the location and size of the
midden, there were few variants of the usual
types found and very few trade sherds. Small
camp sites in the Everglades often turn up more
variant sherds and trade sherds, while the large
villages show greater sherd counts of basic de-
signs. The occupants of Miami Midden No. 1
probably did not have to forage the deep Ever-
glades until fairly late, but small groups "living
out" (with possibly a less rigid pottery tradi-
tion) would move more in a large area able to
support more transients.
One bright spot remains; under the now sealed
parking areas immediately to the north of the
present Dupont Building, there are still many
examples of representative material that might
be excavated under better disciplined conditions
in the future.
Special thanks is due Hilda and the late Hugh
Davis, and Marvin Brooks, for catching the site
in time, and for their help with the excavations.
Appreciation is expressed to the Arkin Con-
struction Company which allowed Chapter men-
bers to be "under foot" for a month, and also to
Mr. Bert Thomas, owner of the Bayshore prop-
erty, for letting members screen the fill.
Others who earned our gratitude were Mr. Rip-
ley Bullen of the Florida State Museum for his
suggestions, the late John Hackett who made the
surface collections in Area A, Bob Masters,
Charles Coomes, his son Andrew, Noel Herrman,
his daughter Virginia and his son Berry, Wayne
Allan, Dick Kotil and Charles Brookfield, who
all helped in one way or the other. Barton Hick-
man made the photograph shown in Fig. 2.
Goggin, John M. and Frank H. Sommer
1949 Excavations on Upper Matecumbe Key,
Florida. Yale University Publications in An-
thropology, No. 41. New Haven.
Goggin, John M.
1950 The Snapper Creek Site. The Florida An-
thropologist, Vol. 888, Nos. 3-4. Gaines-
Laxson, Dan D.
1957 Three Small Dade County Sites. The
Florida Anthropologist, Vol. X. Nos. 1-2.
1957 The Arch Creek Site. The Florida An-
thropologist, Vol. X, Nos. 3-4. Gainesville.
1959 Three Salvaged Tequesta Sites in Dade
County. The Florida Anthropologist, Vol.
XII, No. 3, Gainesville.
McNicoll, Robert E.
1941 The Caloosa Village Tequesta. Tequesta,
Vol. 1. Miami.
1938 John Sewell's Memoirs and History of
Miami. Franklin Press. Miami.
1941 Pre-Columbian Man in Southern Florida.
Tequesta, Vol. 1. Miami.
VOLUME 21, NOS. 2 & 3
THE FLORIDA ANTHROPOLOGIST
BIRD HAMMOCK, MOUND B, REVISITED
Rhodes B. Holliman
JUNE SEPTEMBER, 1968
A salvageable portion of Mound B (8WalO),
Bird Hammock, Wakulla County, Florida, was
surveyed and excavated. The artifacts recovered
include a cache of 7 chert knives; a small, pol-
ished Weeden Island I bowl containing 4 caramel-
colored, chert projectile points and a knife of
the same material; a small, plain tetrapodal bowl;
2 caches of sheet mica; and miscellaneous pro-
jectile points, knives and celts. Sherds recover-
ed are Swift Creek Complicated Stamped of the
Weeden Island Complex. These finds confirm
earlier dating of the mound as Weeden Island I
In 1959, the site called Mound B (8Wa10) in
Bird Hammock, Wakulla County, Florida, was
resurveyed (Fig. 1). Moore (1918: 561-564) re-
corded partial excavation of two mounds in this
site: Mound A (8Wa9), still visible, lies about
335 yards northeast of Mound B. Moore stated
that Mound B had an eastwest diameter of 75 feet
and a north-south diameter of 56 feet. His exca-
vation began with a trench from the east side,
starting with a width of 75 feet and projecting
into the mound for 21 feet, where the trench had
a width of 38 feet. The original height of the
mound was 5 feet, 9 inches. There were two
ramps associated with the mound in early times;
one from the southeast, 23 feet long and 24 feet
wide, and one from the southwest, 38 feet long
and 30 feet wide. The southeast ramp has been
obliterated, possibly by Moore (1918: 563), but
the southwest ramp is readily visible and shows
The west border of the mound today shows the
least intrusion by vandals and, for the purpose
of measurements, will be considered the true
base of the original structure. An east-west
measurement from this point gives a diameter of
54 feet, which could correspond almost exactly
to the original mound dimensions less that por-
tion destroyed by Moore (1918: 563). The north-
south diameter is still about 56 feet. Several
large trees are standing on the western half of
the mound, and on the southwest ramp, showing
that no concerted effort has been made in many
years to excavate completely these remaining
portions. Many shallow intrusions in the mound
and ramp, however, attest to the workings of
vandals and, therefore, prompted this excavation
Willey (1949: 295) postulated a Weeden Island
I dating of the mound from a sherd analysis from
fresh excavations. That is confirmed by this
writer from 22 miscellaneous sherds recovered
from undisturbed sand below a depth of 3 feet in
the mound. All were Swift Creek Complicated
Stamped of the Weeden Island Complex. In addi-
tion, investigation of several low hammocks to
the northeast of Mound B yielded large quantities
of Swift. Creek Complicated Stamped sherds in
oyster shell kitchen refuse. These middens were
also reported by Moore (1918: .564) and Willey
The mound was surveyed in 0.5 foot contours
and staked out in 10 feet x 10 feet squares on a
magnetic compass bearing. Depth measurements
on features were made from present day surface
levels. Square designation is made from the
southwest corner of each square. The first trench
included squares 20N 40E, 30N 40E and 40N
40E. The second trench included squares 50N
20E and 50N 30E (Fig. 1). The presence of
large trees made other excavation prohibitive.
All features were found in a stratum of charcoal
stained sand which was approximately 1 foot
thick. In most cases, undisturbed yellow sand
was encountered after removing 2-2.5 feet of
Feature 1 (Fig. 2). Depth 4.1 feet in 40N 40E.
A cache of seven stone knives was found. Each
was made of gray, amorphous, water-deposited
chert. These show rough, percussion flaking
with some secondary, pressure flaking of the
Feature 2. Depth 4.1 feet in 40N 40E. One pro-
jectile point of translucent, amorphous chert was
Feature 3. Depth 5.0 feet in 40N 40E. One
stone knife of gray, silicified limestone was
0 10 20E 30E 40E 50E 60E 70E 80E 90s
BIRD HAMMOCK CONTOUR INTERVAL 0.5 FOOT
Fig. 1. Horizontal and vertical plans of Mound B. Shaded area in vertical profiles indicates trench. 62
Fig. 2. Cache of stone knives from Feature 1.
Fig. 3. Polished bowl from Feature 5 with remnant of painted figure.
A 3 C D
Fig. 4. Cache of projectile points and knives from Feature 5.
Fig. 4. Cache of projectile points and knives from Feature 5.
Fig. Cache of mica with covering stone and incised alligator tooth from Feature
Fig. 5. Cache of mica with covering stone and incised alligator tooth from Feature
Fig. 6. Plain, tetrapodal bowl from Feature 8.
Feature 4. Depth 4.1 feet in 40N 40E. Two
sheets of mica each measuring 12.5 cm. long by
9.0 cm. wide were unearthed under a thin piece
of sandy slate measuring 16.0 cm. long by 9.5
Feature 5 (Figs. 3-4). Depth 4.1 feet in 30N
40E. A small, plain, polished Weeden Island I
bowl, 12.3 cm. in diameter and 3.3 cm. deep was
found containing four projectile points and a
knife (Fig. 4 a-e) made from a fine caramel color-
ed, translucent chert. The points and knife were
apparently from the same core and show excel-
lent secondary chipping. The interior of the bowl
was originally decorated with a black painted
figure and radiating black lines along the rim.
The figure is nearly obliterated but a bird form
can be postulated from the remnants of paint
(Fig. 3). The bowl was cracked from soil press-
ure but had not been "killed." In addition, the
cache included three other points and a knife
(Fig. 4 f-i), all made from white, silicified lime-
Feature 6. Depth 5.3 feet in 40N 40E. A broken
greenstone celt, a small slate celt measuring
6.5 cm. long and 4.4 cm. wide, one flint knife
blank, and five pieces of sandstone were found.
Feature 7 (Fig. 5). Depth 7.2 feet in 50N 20E.
Eight sheets of mica (Fig. 5b) each measuring
about 7.5 cm. long by 7.0 cm. wide were found
stacked under a piece of sandy slate (Fig. 5a)
measuring 9.0 cm. long by 7.3 cm. wide. In, ad-
dition, an incised alligator tooth (Fig. 5c) was
Feature 8 (Fig. 6). Depth 8.2 feet in 50N 30E.
A small, plain, tetrapodal Weeden Island I bowl
measuring 9.2 cm. wide at the rim and 7.7 cm.
deep was found in a badly crushed condition. It
was probably "killed" ceremonially. The bowl
showed heavy use by its fire-blackened and erod-
Feature 9- Depth 3.2 feet in 50N 30E. Seven
sheets of mica, each about 10 cm. in diameter.
were neatly stacked.
Moore (1918: 563) reported a few vessels of
plain ware or coarsely decorated design that
were found singly in this mound and all were
"killed" by a basal perforation. No illustra-
tions were given. He also found 15 burials in a
very poor state of preservation, some of which
had associated artifacts. He recorded mica from
Mound A but not from Mound B. To these data
we can now add an "unkilled" vessel, a vessel
"killed" by crushing, and an abundance of mica.
It would appear from our limited survey that
the aboriginals constructed a primary mound
about two feet above the existing forest floor
into which they placed their mortuary offerings.
Beach sand was then deposited on the primary
mound to form the present elevation and burials
and additional offerings were deposited therein.
A salvage operation was deemed important to
recover rapidly diminishing information from a
much-dug, but little reported site. Appreciation
is extended to Mr. Edward Dolan for his assit-
ance in the project.
Moore, Clarence B.
1918 The Northwestern Florida Coast Revis-
ed. Journal of the Academy of Natural Sci-
ences of Philadelphia, Vol. 16, Part 4.
Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast.
Smithsonian Miscellaneous Collections, Vol.
Virginia Polytechnic Institute
THE FLORIDA ANTHROPOLOGIST
THE PALM RIVER MIDDEN, HILLSBOROUGH COUNTY, FLORIDA
Tests pits dug into a shell midden in the Tampa
Bay area produced ceramics indicative of a Safe-
ty Harbor Period occupation. The site is assign-
ed to the pre-contact portion of this period due
to the complete lack of European items. A work-
ed beaver incisor and two pottery types evidence
trade with peoples to the north.
The Palm River midden (Hil08) consists of a
low marine shell deposit situated directly on the
south bank of the Palm River, just below the
city of Tampa, in Hillsborough County, Florida.
The site is approximately one quarter of a mile
upstream from McKay Bay, in the NW /4 of the
SE /4 of Section 21. Twn.29 S Rng.19 E. Here
the river is quite broad, and is affected by the
action of the tides. The midden has been un-
disturbed by pot-hunters, but the brackish water
is slowly eroding it away by undercutting the
shell stratum. The existing shell deposit covers
an area roughly 350 feet in length and up to 70
feet in width.
A narrow drainage ditch bisects the site. To
the east of the ditch the area is undisturbed and
covered by a dense growth of weeds, shrubs,
palmettos and cabbage palms. A few deciduous
trees are also represented. A thin band of man-
groves grows along the river bank. Here, the
midden surface is irregular and dotted by several
depressions. The western half of the site has
been cleared of vegetation, except for the palm
trees. It also appears to have been leveled and
graded. The greatest vertical concentration of
shell borders on the river, being up to almost
four feet in depth. From this point the elevation
decreases gradually towards the south.
Three stratigraphic test pits were excavated
in the overgrown area using arbitrary six inch
levels. The cleared section was neglected be-
cause permission to dig in this area was not
granted. The first test was dug in September of
1967, the others in March and April of 1968.
Test A, a 5 x 10 foot pit, was placed in an area
of heavy shell concentration at a point half way
between the east and west limits if the undis-
turbed midden deposit. It was laid out parallel
to the river, 6.5 feet from the eroding bank. The
upper four inches consisted of black, sandy loam
and shell. Beneath this was a layer of clean
shell. This contained lenses of both sandy loam,
and gray ash, sand, and crushed, burned shell.
The lenses contained the majority of the arti-
facts and bones recovered from this level. Shell
mixed with black sand comprised the fill between
13-20 inches. The shell deposit ended at a depth
of twenty inches in the east of the pit, and at
24 inches in the west end. Below this was dark
sand which gradually graded into light colored
sand and represents the original ground surface.
The black sand contained some artifacts, bone
fragments and scattered oyster shells. The gray-
ish-white sand was completely sterile. A light
colored, mottled clay hardpan was encountered
at 31 inches.
Test B was a five foot square pit situated 56
feet to the west of the first test. It was six feet
from shore, in another area of thick shell. The
stratigraphy of this test (Fig. 1) very closely
approximates that of Test A. The upper six
inches consisted of shell and black earth. Below
this, to a depth of 24 inches, was clean shell
with lenses of black soil, charcoal, ash, and
numerous fragmentary fish bones. The fill be-
tween 24-36 inches was black soil and shell
which, unlike the equivalent layer in Test A,
was so hard and compact that it could only be
removed with a great deal of effort. The shell
disappeared at a depth of 36 inches and was re-
placed by sterile, brownish sand. Mottled gray-
ish-brown clay appeared at 44 inches.
Test C, also five feet square, was placed forty
feet to the east of Test A. It was 23 feet inland
from the shore line. Here the shell was covered
by a 1.5 inch thick loam layer. The shell bearing
level, consisting of compacted shell and black,
organic soil, ended eight inches below the sur-
face and was underlain by dark sand. Artifacts
appeared only in the upper six inches.
Oyster shell comprised the major portion of the
VOLUME 21, NOS. 2 & 3
JUNE SEPTEMBER, 1968
SOIL & SHELL
I CLEAN SHELL
SAND & SHELL
Fig. 1. The north profile of Test B at the Palm
midden fill. The next most common molluscs
were Melongena corona and Polinices duplicate.
Other shellfish (Table 1) appeared less frequent-
ly. Strombus alatus and Fasciolaria gigantea
were each represented by only a single hammer.
In Test A the frequency of Melongena lessened
as depth increased, whereas Venus became more
frequent. In Test B the percentage of Melongena
remained about the same but the number of Venus
shells increased with depth.
Vertebrate animal remains (Table 2) were num-
erous and representative primarily of fish, turtles
and the white-tail deer. The latter was the sole
representative of the mammals with the exception
of a cotton rat which may be intrusive into the
midden material. The remains of alligators were
very rare in the first test but fairly common in
the second. Scute fragments from Test B were
carbonized on the exterior surface indicating
that these reptiles were probably roasted en-
cased in their hides. Turtles were represented by
almost every genus inhabiting the area today.
Of them all, the rarest was the softshell turtle.
Bird remains were scarce and fragmentary. The
most frequently encountered fish remains were
those of the two marine catfish. Of the two,
Galeichthys was by far the more common. The
next most popular fish was Caranx, the Jack.
The remaining fish appeared in about equal num-
What appears to be the head of a proximal hum-
an phalange was recovered from the 12-18 inch
level of Test B. It may represent an accidental
amputation or it may have been in some material
transported to the midden from a burial area.
NON CERAMIC ARTIFACTS
In all three tests the upper two inches con-
tained items such as nails and metal scraps.
All of these objects appear to have been de-
posited during storms since the area near the
shore is littered with an assortment of debris.
There was no evidence of any disturbance below
this depth. Also common to all excavations were
pieces of unaltered, porous limestone and rough-
surfaced chert "growths", none of which ex-
ceeded three inches in diameter. These were in
almost every level.
Although no chipped stone artifacts were en-
countered, chips and small cores of local, impure
chert were found at various depths, though they
were most numerous in the upper six inches.
The lack of stone tools was sharply contrasted
by the great quantity of shell artifacts. The most
common of these were Melongena hammers.
These comprised just over 71% of the total shell
artifact assemblage. Strombus (Fig. 2 ), Busycon
and Fasciolaria hammers were very rare.
Test A yielded the greatest assortment of arti-
facts (Table 3). The most interesting of these
is a worked beaver incisor (Fig. 2i-k) recovered
from the 18-25 inch level. It is 25 mm. long,
8 mm. wide, and 5 mm. thick at the edge. The
posterior portion of the tooth has been removed,
leaving the sides enclosing a hollow pulp cavity.
One end is beveled sharply producing a keen
biting edge. The other end is also beveled some-
what, but the base has a crescent-shaped con-
cavity in it, leaving the sides protruding as
blunt points (Fig. 2j). This concavity appears
to be the result of a break, rather than an in-
This artifact is interesting in two respects.
First, it brings to mind the beaver incisor chis-
els characteristic of several cultures attributed
to the period of transition from Middle to Late
Woodland in the northeastern United States and
south-central Canada (Griffin 1952: 63, 78, 93,
104). The incisors used as chisels associated
with the Arvilla Focus of northern Minnesota are
cut on the lingual surface to sharpen the biting
edge (Griffin 1952: 119). This treatment closely
resembles that of the tooth from Hi108, and it
may well have served as a chisel.
Second, the Tampa Bay area, along with most
Florida, is not the natural present-day habitat
of the beaver (Hall 1959: 549). According to
Olsen (1964), it is also absent from archeologi-
cal sites in the state. However, Neill (1956:
391) describes unworked beaver remains from
TABLE 1. MOLLUSCS REPRESENTED AT THE
PALM RIVER MIDDEN
Lobed Moon Shell
Banded Tulip Shell
Great Heart Cockle
Sun Ray Shell
VERTEBRATE ANIMALS REPRESENTED AT THE
PALM RIVER MIDDEN
Malaclemys terrapin (?)
Hispid Cotton Rat
TABLE 3: STRATIGRAPHIC SEQUENCE OF ARTIFACTS AT PALM RIVER MIDDEN
LEVELS (6" each)
1 2 3 4 5 1 2 3 4 5 6 1
Pinellas Plain 33
Belle Glade Plain
Lake Jackson Plain 2
St. Johns Check Stamped
Clay Tempered Plain 5
Fossilized Bone Frag.
Fossil Shark Tooth
Sandstone Abrader Fraqs. 2
Worked Beaver Incisor
Small Battered Columella 2
Melongena Hammers 4
Worked Busycon Frags.
33 37 19
1 1 1
2 7 7 1
several preceramic and early ceramic sites in
east central Florida, including a midden in Vol-
usia County which is possibly assignable to the
St. Johns II period. This indicates that the bea-
ver once inhabited at least a portion of the state.
Therefore, the worked incisor indicates either
the presence of this animal in the Tampa Bay
area at one time, or contact with people of re-
gions which were inhabited by beavers. Of the
two possibilities, the latter seems the more
The only other bone artifacts consist of two
specimens of what are probably pins. Both were
manufactured from a deer long bone, and the cut
surfaces are smoothed. One, from the 6-12 inch
level, is a medial fragment (Fig. 2 ) measuring
76 mm. in length, 6 mm. in width, and 4 mm. in
thickness. The other (Fig. 2n) appears to be com-
plete and measures 79 mm. in length and 4 mm.
in thickness. One end tapers slightly, ending in
4 mm. wide, blunt tip. The other end has been
broken through a very shallow, transverse gro-
ove cut into the bone to form a flat base 7 mm.
A chert nodule four inches long and 2.5 inches
wide served as a hammerstone. One end is un-
altered, but the other is chipped and battered
from use. Two cubic fragments of a sandstone
abrader were located in the upper six inches.
A couple of small pieces of red ocher-cemented
sand were between 12-24 inches. Three frag-
ments of a large Busycon shell, found between
6-18 inches, exhibiting grinding and smoothing
along one edge, along with unworked pieces of
the same shell, probably represent a cup or dip-
per. A columella chisel (Fig. 2g) was the only
non ceramic artifact recovered from the sand
layer below 24 inches.
Test B. Excluding pot sherds, artifacts were
rare except for the everpresent Melongena ham-
mers. These were relatively common in the upper
eighteen inches, became very scarce in the 18-
24 inch level, and were replaced by columella
pounders in the lowest levels. This test also
yielded a small piece of red ocher-cemented sand
and a fragment of unaltered, fossilized bone.
Test C produced a 6.5 cm. long, fossilized shark
tooth (Fig. 2h), as well as two hammers and a
columella pounder (Fig. 2m).
The midden yielded a total of 324 pot sherds.
All of these were undecorated, except for a few
which were check stamped. The pottery was not
very varied with only five recognizable types
being represented in the collection. The types
and their quantities are listed below. The per-
centages given indicate what part of the total
ceramic inventory each type represents.
Pinellas Plain...................Total: 264 (81.5%)
Rims.....32 Body Sherds.....232
Belle Glade Plain...................Total: 19 (5.9%)
Rims.....0 Body Sherds.....19
Lake Jackson Plain................Total: 15 (4.6%)
Rims.....1 Body Sherds.....14
St. Johns Check Stamped..........Total: 13 (4.0%)
Rims.....2 Body Sherds.....11
Pasco Plain........................... Total: 8 (2.5%)
Rims.....0 Body Sherds.....8
Unclassified Clay Tempered Plain
......................................... Tota 5 (1.5%)
Rims.....1 Body Sherds.....4
The sherds identified as Pinellas Plain are,
for the most part, fine sand tempered. A few
coarser grained specimens are also represented.
However, they all fit into Willey's (1949: 479)
definition of this type. In 70% of the sherds the
paste is contorted, laminated, and crumbly. The
surfaces are crackled. This is especially true of
the sherds from Test B which tended to disinte-
grate when removed from the midden and dis-
solved when placed in water. Two rim sherds
from the upper twelve inches of Test A have
rough, "pebbly" surfaces (Fig. 2a). The re-
maining sherds have a harder, more compact
paste. Nevertheless, the majority of these lam -
ination of the paste in the uneven, layered pro-
jections protruding from the broken sherd edges.
Rim sherds represent large, open bowls. None
of the rims bore any appendages. Lips were
round: 53.1%; flat-round: 28.1%; flat: 18.8%. Flat
lips were confined to Test A. All lips on sherds
from Test B were round. The only rim from Test
C was flat-round. None of the lips bore any
crimping or indentations on the exterior edge.
One rim sherd from Test A witn a vertical,
triangular node on the lip was identified as Lake
Jackson Plain (Fig. 2f). All body sherds with a
similar paste and surface finish were classified
as such also and probably came from the same
vessel. The paste is gray, somewhat contorted,
with a light, fine sand-clay temper. It is much
harder and compact than that in the Pinellas
Plain sherds. Sherd exteriors are much smoother.
In all, it is a better quality ware.
The St. Johns Check Stamped sherds represent
two vessels. One, with a gray-brown paste, bears
five square checks per linear inch (Fig. 2b). The
other has nine square checks per inch on a buff-
The unclassified sherds are heavily tempered
with pieces of crushed-clay which extrude onto
the surface (Fig. 2e). Sherd exteriors are'well
smoothed, but are crackled slightly. The rim is
6 mm. thick; the lip is flat. The sherds represent
a single vessel. Bullen (personal communication)
does not feel that they represent a variety of
Pinellas Plain pottery, but, rather, that they are
trade sherds. However, the point of origin is
The predominance of Pinellas Plain sherds at
the Palm River midden indicates a Safety Harbor
period occupation (Sears 1967: 25; Willey 1949:
482). The other identified pottery types corrobo-
rate this date. The St. Johns Check Stamped
and Belle Glade sherds are associated with the
Safety Harbor Complex although they are not
diagnostic markers for this period (Willey 1949:
479). Pasco Plain sherds are found in small
quantities in some Safety Harbor sites. The Lake
Jackson Plain pottery is assignable to the con-
temporaneous Ft. Walton period of the northwest
coast. Since no items of European origin were
recovered from the site, it is presumed to be
pre-contact and should be dated as early Safety
Harbor ca. A.D. 1350-1513 (Bullen 1965: 306).
Trade with peoples to the north is suggested
by the presence of the clay tempered sherds and
the worked beaver incisor. The Lake Jackson
Plain sherds probably also came from an im-
ported vessel, as did most of those at the Tierra
Verde site in Pinellas County (Sears 1967: 58).
However, those at Tierra Verde were imported
for ceremonial use and not for utilitarian pur-
poses as they were at Palm River.
5 C M.
Fig. 2. Artifacts from the Palm River midden. a, Pinellas Plain; b, St. Johns Check Stamped; c, Belle
Glade Plain; d, Pasco Plain; e, unclassified clay tempered plain; f, Lake Jackson Plain; g, col-
umella chisel; h, fossilized shark tooth; i-k, worked beaver incisor (three views); 1, Strombus
hammer; m, columella pounder; n-o, bone pins.
A heavy dependance on hunting, fishing and
gathering to supplement an agricultural diet is
indicated by the great number and variety of
vertebrate and molluscan remains in the midden.
The variety also indicates that the inhabitants
consumed almost every available form of animal
life. In this respect the lack of mammal remains,
excluding the white-tail deer, is interesting. It
may imply that other mammals were not readily
available, or that the bone sample was not large
enough to show more diversity.
Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast.
Smithsonian Miscellaneous Collections,
Vol. 113, Washington.
Marie Laski and Stephanie Pierog deserve
special thanks for permitting the excavation of
the site situated on their property. Thanks are
also extended to Nancy Hubble, Janet Adams,
Mike Gadd and Stuart Otto for helping when they
could with the digging and screening. Bill Harri-
son is thanked for photograghing the artifacts.
Gratitude is expressed to Ripley P. Bullen for
his assistance and suggestions concerning the
pottery, as well as to Dr. Elizabeth S. Wing and
Dr. Walter Auffenberg for helping with the identi-
fication of the vertebrate remains.
Bullen, Ripley P.
1965 Florida's Prehistory. In Florida from
Indian Trail to Space Age, Vol. 1, pp. 305-
316. Southern Publishing Co. Delray Beach.
Griffin, James B. (editor)
1952 Archeology of Eastern United States. The
University of Chicago Press. Chicago.
Hall, E. Raymond and K. R. Kelson
1959 The Mammals of North America, Vol. 2.
The Ronald Press Co. New York.
Neill, WilfredT., H. J. Gut, and P. Brodkorb
1956 Animal Remains from Four Preceramic
Sites in Florida. American Antiquity, Vol.
21, No. 4, pp. 383-395. Salt Lake City.
Olsen, Stanley J.
1964 Mammal Remains from Archaeological
Sites, Part 1. Southeastern and Southwestern
United States. Papers of the Peabody Mus-
eum of Archaeology and Ethnology, Harvard
University, Vol. 61, No. 1. Cambridge.
Sears, William H.
1967 The Tierra Verde Burial Mound. The
Florida Anthropologist, Vol. 20, No. 3-4,
pp. 25-73. Tallahassee.
THE FLORIDA ANTHROPOLOGIST
PINELLAS POINT: A POSSIBLE SITE OF CONTINUOUS INDIAN HABITATION
Albert C. Goodyear
A report of a midden beach on the southern
coastal tip of Pinellas County. The analysis
was made on the surface collections of six in-
dividuals. who have been collecting sherds, pro-
jectile points, and other artifacts of Indian man-
ufacture from a badly eroded shell midden which
has been, and is presently being, washed by
wave and tidal action. This site is of particular
archaeological interest since representative
artifacts from practically all the Gulf coast cult-
ural periods have been recovered there.
The Pinellas Point Site (Willey 1949: 335) is
readily accessible and locatable since it lies
within the city limits of St. Petersburg. The mid-
den can best be reached by following 10th Street
South, another synonym for the site, till it term-
inates at the midden beach itself. Roughly, the
area between 14th Street South and 7th Street
comprises the Pinellas Point Site (Fig.l). The
actual site, however, is strictly limited to the
The physical features of Pinellas Point are
typical of many of the middens that are located
along the coastline of southern Pinellas County.
The site is backed by a low lying flatland on
which forests of long leaf pine (P. palustris)
occur. This pine forest which extends back be-
hind the midden for approximately 100 yards,
terminates abruptly because of a housing devel-
The length of the midden was measured and
found to extend for some 750 yards. The width
of the midden varied from place to place, depend-
ing on the portion to be measured. On the west-
ern end of the site the midden was approximately
25 yards wide. This is the widest part of the
midden as it gradually diminishes in an easterly
direction, tapering down to a few feet of water-
worn shell in its eastern-most extreme. Until a
few years ago there was a shell mound located
directly in the middle of the beach strip (Fig.
1). This mound was unfortunately bulldozed
flat in order for a house to be built in the ele-
vated area. Though no excavations were made
in this mound, the author personally recovered
three projectile points that can be directly attri-
buted to it. Since all portions of the midden have
been severely tumbled and disturbed by water
action, no portion 1of the midden was suitable
for stratigraphic tests. Thus, the investigation
of the site has been limited to surface collect-
In analyzing the artifacts from Pinellas Point,
the collections of Mark Brooks, Frank Bushnell,
Phil Jordan, Richard McDonnel, and the author
were utilized to classify and categorize by ty-
pology the various sherds and projectile points
found at this site.
A problem immediately encountered in the an-
alysis of the pottery was the extreme scarcity
of decorated sherds. In all, eight decorated types
were found, and usually a type was not repre-
sented by more than two or three specimens.
Though pottery sherds are common at the Pin-
ellas Point Site, 95% of them are a dark, sand-
tempered, residual ware that is very indefinite
as to time and type in this area. This pottery
closely fits the description of Glades Plain,
Thick Sand Tempered Plain, and Sand Tempered
Plain cited by William H. Sears from the Bay-
shore Homes Site, also located in St. Petersburg
(Sears 1960: 4-7). Similar to the Glades ware
at the Bayshore Homes Site, the sherds from
Pinellas Point have a thickness range of 3/8 to
3/4 inch, and also run from reddish-brown thr-
ough gray and black in color.
With the advent of Bullen's (1968) projectile
point guide, an attempt was made to classify
the stone projectile points of this site by types
and to supplement the pottery evidence in order
to better reconstruct the cultural sequence of
In total, 285 representations .of projectile
points were counted and from these 125 were
reasonably classifiable by type. The realm of
unclassifiable consisted not only of whole points
but also tips, stems, bases, and any other por-
tion that would indicate a projectile point was
intended. The 160 remaining unclassifiable pro-
jectile points were not figured in the percent-
ages, but should be noted as a testimony to the
large use of projectile points by the inhabitants
of this site. Of the 285 specimens examined,
none were made of any foreign stone materials
(such as quartzite) and all seem to be made of
typical Tampa Bay silicified coral or limestone
VOLUME 21, NOS. 2 & 3
JUNE SEPTEMBER, 1968
SCALE IN FEET
0 500 1000 1500 2000
Fig. 1. Midden location in southern St. Petersburg.
Percentages were constructed for the 125
classifiable points and seem to illustrate heavy
concentrations of cultural periods which were
previously expected to predominate (Table 1).
The sole evidence of Paleo-Indian contact on
Pinellas Point is centered around one projec-
tile point (Fig. 2a): This point, from the Mark
Brooks collection, complies closely with the
Suwannee definition (Bullen 1968: 48), and ex-
hibits limited basal thinning and grinding. Su-
wannee points are not uncommon to the Tampa
Bay area as they are occasionally found on
dredged and pumped-up bay bottom fills along
this coast (Warren 1964).
A variety of point types comprise the Archaic
period in its earliest stages. One Santa Fe-like
point has been recovered (Fig. 2b), by the
author, but is somewhat deviant in that it has
ears which hang down. This point is too badly
patinated, almost to the point of chalkiness,
which precludes any information to be obtained
through ground bases and edges. It is, however,
quite thin basally. A second point probably of
this same time period (Fig. 2c), is somewhat
unique in that it possesses the concave and
pointed eared base of a Santa Fe point, but is
also short and lightly waisted characteristic of
a Beaver Lake. It is also ground on its edges
and base. The possibility of this point and other
points which demonstrate a combination of traits,
being regional variants or transitional experi-
ments in passing from one point style to another
should be considered.
Among the Bolen series four have been found.
Two are beveled and two are plain. Among the
Bolen Plain, one of these (Fig. 2d) demonstrates
large size, severe patination, and specialized
notches, all of which Bullen has discussed
(1968: 42). The other Bolen Plain is an example
of a subtype 2 (Fig. 2e). The remaining two
Bolen Beveled are made of the same type of
chert and show a moderate amount of patina. The
smaller of the two (Fig. 2f) represents a subtype
3, or high notched. The last Bolen Beveled
(Fig. 2g) is probably a subtype 4 due to its size
and the remnants of barbs once present on the
top of each notch.
ARCHAIC AND LATE ARCHAIC
A predomination of Archaic period projectile
points (33.6%, Table 1) which closely resemble
the outlined forms set forth by Bullen (1968:
28-30), constitutes the evidence of an Archaic
period on Pinellas Point. If, as Bullen states,
the four classic types, Putnam, Levy, Marion,
and Alachua were made over the longest period
of time and hence do represent the most common
Florida point, then the majority of the projec-
tile points found at this site would naturally
cluster in this category. Among the Putnams,
Levys, and Alachuas there were no significant
differences in quantity, but the Marions out-
numbered any one of the other three types three
Proceeding to the later Archaic, the Newnan
point is fairly well represented on Pinellas
Point. Of the 125 identifiable projectile points,
7.2% were identical to the model constructed by
Bullen (Fig. 2h). One of these had its distal
end beveled slightly and probably served as a
The Culbreath point makes up the third largest
category with 11.2%. The actual number of Cul-
breaths (14) were equally divided between sub-
types 1 and 2. The original cutoff date for Cul-
breaths has been set at 1,000 B. C. (Bullen
1968: 28), but recent excavations at the Canton
Street midden by Frank Bushnell, Walter Askew,
and the author demonstrate that Culbreath points
probably have a longer tradition than originally
expected, as Culbreath points have been repeat-
edly found in Perico Island zones at this site.
This Canton Street midden is located about two
miles due west from Pinellas Point and is lo-
cated on the same coastline.
The establishment of an Orange and Transi-
tional period is considerably hampered by the
lack of any fibertempered sherds as well as any
St. Johns and Pasco Incised. One questionable
sherd of St. Johns Incised has been found, but
its paste seems too gritty for chalky ware. One
good example of a steatite sherd has been found
and is probably Transitional in date. Again refer-
ring to the Canton Street midden, several sherds
of steatite have come from the bottom of this
deposit which is mostly of the Orange and Trans-
itional period in its bottom layers. Two orna-
ments of steatite have also been recovered from
Pinellas Point, one tubular bead and the head
of a grooved pendant (Fig. 3). These ornaments
may also represent re-worked steatite sherds by
later cultural periods rather than artifacts of the
Fig. 2. Projectile points from Pinellas Point.
TABLE 1. CLASSIFIABLE PROJECTILE POINTS BY PERCENTAGE.
TYPE PERCENT OF TOTAL
Beaver Lake 0.8
Santa Fe 0.8
Morrow Mountain 1.6
Stemmed Archaic Types
(Total Sample: 125 specimens) 100.8
TABLE 2 SHELLS FROM PINELLAS POINT MIDDEN IN DECREASING
ORDER OF PREVALENCY.
1. Virginia Oyster (Crassostrea virginica Gmelin)
2. Southern Bay Clam (Venus Mercinaria campechiensis
3. Lightning Conch (Busycon perversum Linne)
4. Bay Scallop (Pectin irradians Lamark)
5. Horse Conch (Fasciolaria gigantea Kiener)
6. Florida Conch (Strombus alatus Gmelin)
7. Tulip Shell (Fasciolaria tulipa Linne)
8. Sun Ray Clam (Macrocallista nimbosa Solander)
9. Crown Conch (Melongena corona Gmelin)
10. Great Heart Cockle (Dinocardium robustrum Solander)
11. Bloody Clam (Arca pexata Say)
12. Lobed Moon Shell (Polinices duplicate Say)
In the substantiation of cultural periods on
Pinellas Point through the use of decorated
pottery, the Deptford and Perico Island periods
benefit the most. From the limited amount of
designed sherds recovered from this site, four
were Deptford, three of which were Deptford
Linear Check Stamped, one unique stamp, and
three were Perico Linear Punctated. Deptford
and Perico Island specimens have been lumped
together in this section since Bullen feels that
Perico Island is definitely in a Deptford time
horizon temporally (Bullen 1966).
By the use of projectile points, Hernando and
Citrus specifically, further demonstration of a
Perico Island occupancy can be accomplished.
Using the percentages previously cited, 24% of
the 125 were definitely Hernandoes while only
1.6% (2) could safely be considered Citrus
points. Again, referring to the Canton Street
Midden, Hernando and Citrus points have been
found in association with Perico Island ceramics
and the same is true for many other sites of
this time period (Bullen and Askew 1965). It
should also be mentioned that 13 of the 30 Her-
nando points were Hernando C subtypes (Bullen
and Bullen 1963: 90-91), and that Pinellas
Point has long been known locally as site pro-
ducing these variations in Hernandoes. Also the
limited amount of Citrus points seems unusual
since they are considered contemporaneous
with Hernando points (Bullen 1968: 24).
SWIFT CREEK AND WEEDEN ISLAND
A single sherd of Swift Creek Complicated
Stamp, Early Variety, has been recovered (Fig.
4), and constitutes the only decorated sherd of
this period. Six Duval points have been found
(4.8%), and may, as Bullen (1968: 17) states,
belong to both the Weeden Island periods as well
as Swift Creek.
Evidence for Weeden Island Occupation is
also rather weak. The most common Weeden
Island sherd is St. Johns Check Stamped, five
of which have been recovered. Two pieces of
West Florida Cord Marked have been found, as
well as one questionable sherd of Pasco shell
impressed (?). None of the Weeden Island In-
cised or Punctated type sherds have ever been
seen from this site.
Three Columbia (2.4%) projectile points have
been found but are too small to serve the dag-
ger function suggested by Bullen (1968: 21).
From the standpoint of ceramics, evidence
for a Safety Harbor occupancy is also scant.
Three Pinellas Plain sherds have been recover-
ed, made of typical Pinellas contorted paste,
but none with notched rims. William H. Sears
also encountered some of this plain ware at the
Lighthouse Mound (Sears 1960) without much
other Safety Harbor pottery and it may be sug-
gestive of late Weeden Island, virtually transi-
tional to Safety Harbor (Bullen, personal com-
However, though the Safety Harbor pottery re-
covered is somewhat limited, projectile points
presumably of this period are well represented.
Nine Tampa points (7.2%) and one Pinellas
point have been recorded, and are very close
in design to similar points found by Frank Bush-
nell at the Maximo Point Site, a site of the Safe-
ty Harbor period (Bushnell 1962: 100).
q 6 4 g
Fig. 3. Ornaments from the Pinellas Point Site.
a-d, shell plummets; e-h, shell beads;
i, shark tooth bead; j, galena pendant;
Four artifacts, an ovoid blue glass bead, a
gunflint, olive jar sherds, and a clay pipe stem
comprise the evidence of direct or indirect (tr-
ade) European contact on Pinellas Point (Fig.
6 a-c). Of these specimens, probably only the
blue trade bead could safely be considered a
possession of the Indians. Sometime in the
1700's or before, Indians in the Tampa Bay area
were using firearms as a major hunting device.
Celi of the Royal Spanish fleet, in 1757, twice
encountered Indians who were using firearms,
but he fails to tell what type (Ware 1968: 42).
The gunflint from Pinellas Point, though prob-
ably belonging to a settler or fisherman, may
also have been Indian. It is, however, chipped
of a flint not indigenous to Florida. Since Span-
ish fishing haciendas were located in coastal
southern Pinellas County quite frequently up into
the 1800's, the presence of olive jar and clay
pipe stems would seem natural. It should be re-
membered that while these items were of an
European or North American origin, they could
have been used by both settler and Indian.
Fig. 4. Decorated sherds. a, olive jar; b, St.
Johns Check Stamped; c, drilled sand-
tempered plain; d, Deptford Linear
(unique); e, Cord marked; f, indistinct
stamp; g, Perico Linear; h, Swift Creek
Complicated Stamp, early variety; i, St.
Johns Incised (?); j, steatite sherd.
Fig. 5. Bonework and raw materials, a, peg-top
pin; b-c, cut deer bone; d, bonebead; e,
drilled bone gorget.
Fig. 6 European contact artifacts, a, blue gun-
flint; b, blue glass bead; c, clay pipe
.- .- -w ..- -
Fig. 7. Microliths from Pinellas Point. a-d, plain,
slender liths; e-g, expanding base; h-i, T
As previously shown the investigation of the
Pinellas Point Site is considerably handicapped
by the small quantities of decorated pottery re-
covered. When considering the relatively small
amount of designed sherds in comparison to the
proportionately larger undecorated sherds, a
problem of distribution arises. Why are designed
sherds so scarce when plain ware from this site
is so plentiful? The nature of the site itself, that
of a midden, is of course, part of the explana-
tion, but many other middens on the Gulf coast
such as the Wash Island Site (Bullen and Bullen
1963), produce large amounts of decorated 'pot-
tery in spite of being isolated from any mortuary
construct. One other factor that must be mention-
ed is the discriminating, method of surface col-
lecting that has been utilized by the local indi-
viduals who have made collections from this
site. When walking this site, the collection of
sherds has been usually neglected except for an
obviously decorated example. If more attention
had been directed to the collecting of sherds, it
is possible that more decorated pieces would
have been recovered.
Another problem that presents itself is the ap-
parent absence of the plano-convex scraper.
Wherever Archaic and early Archaic projectile
points are encountered on the Gulf coast, plano-
convex tools invariably accompany them. Such
is not the case at this site, in fact the author
has only seen two specimens that could be
classified as such. It is also important to note
that thumbnail or flake scrapers, produced us-
ually by all periods, are quite rare here.
This lack of scraping tools is somewhat uni-
que, however, since plano-convex tools are
pumped up in profusion from local submerged bay
bottom sites. The Maximo Point fill, approxi-
mately two miles to the west following the same
coastline, offers interesting contrast. On this
fill Bolen Plains were found as well as Green-
briar-Daltons and Clovis-like blades. On this
fill no Bolen Beveled points were ever found but
the most common Archaic artifact was the plano-
convex scraper. The lack of scraping tools on
Pinellas Point is quite problematical and is thus
heightened by the comparative conclusion that
most other sites yielding Archaic and early Arch-
aic projectile points have plano-convex tools
occurring simultaneously. This suggests that the
early Archaic occupation of Pinellas Point was
probably more sporadic and of shorter duration
than the usual early Archaic sites which seem to
have a more complete tool inventory. Since Bol-
ens, Greenbriars, Daltons, and other more lan-
ceolate points occur quite frequently with plano-
convex tools in the Tampa Bay fills and dredges
and since both the tools and the points of these
sites show very similar discolorations and pa-
tina, it is felt that in this bay area these two
types of artifacts are probably coeval.
An interesting phenomena that occurs quite
abundantly at this site is the microlith (Fig. 7).
Several hundred of these small implements have
been found and some variety of design has been
noted. Presumably some of these were hafted
since ears, notches, expanding bases, and T
shaped bases were fashioned. The exact use of
these microliths is not known, but since few
surviving artifacts from Pinellas Point demon-
strate perforation, with the exception of an occa-
sional shell or bone bead, some perishable mate-
rial such as skin or wood is .implied. James A.
Ford, of the Florida State Museum examined
some of the microliths in 1966 and stated that he
thought they were similar to those from Jaketown
in use and manufacture. Whatever the use in-
tended, these small, drill-like tools were chipped
in great quantities and undoubtedly constitute
the most common worked stone from this site.
THE PINELLAS POINT ENVIRONMENT
The Pinellas Point Site should also be con-
sidered in light of its geographical and ecologi-
cal setting. Geographically it is in a prime loca-
tion from the standpoint of travel and communi-
cation. It lies on a promontory which is one of
the southeastern most portions of Pinellas Coun-
ty and potentially a natural landfall for bay-going
dugouts traveling up Tampa Bay or out into the
Gulf of Mexico. By being situated on the coast-
line it was no doubt easily accessible to people
who were using the seashore or open water as a
travel route. The recovery of such foreign trade
objects as steatite and galena gives an indica-
tion as to the northern traveling and trading in-
fluences experienced by the Pinellas Point
Since the site is situated on an exposed coast-
al beach, the effects of hurricanes and heavy
storms should be considered when discussing
the badly eroded and tumbled midden present
there today. The coastline of Pinellas Point
lies in confrontation with lower Tampa Bay and
also the Gulf of Mexico. The low lying flatland
that the site is located on might be in danger of
flooding, as it sometimes is today, if hurricanes
and squalls occurred and would possibly make
prolonged occupation unfeasible.
The ecological factors of a site on a bayshore
are obvious. Dependence on marine life such as
shellfish and fish must have been constant.
Frank Bushnell, biologist at St. Petersburg Jun-
ior College, has kindly compiled a list of shells
from this midden which he considers representa-
tive of the inhabitants diet. This list (Table 2)
is presented in a decreasing order of quantity.
Oddly, few of the shells such as Busycon per-
versum, Strombus alatus, and Melongena corona
show signs of alteration for such tools as adzes,
hammers, picks, and gouges. The most common
worked shell artifact is in the form of ornaments
Though wild game must have been plentiful,
as indicated by the large presence of projectile
points, few animal bones have survived from
this midden. The most common food bones are
turtle, deer, and fish. Some of the bone was
worked into pins, bird bone beads, and drilled
sharks teeth. Modern examples of cow and pig
have been found and demonstrate a large amount
of deterioration presumably due to the non-pre-
serving nature of the salt water and midden.
Regarding sources of fresh water, there are
two artesian wells present, one at 14th Street
and one at 7th Street (Fig. 1). These now have
metal pipes sunk into them and it is possible
that these outflows were present long enough
ago to be utilized by the occupants. In addition
to these springs, a fairly large lake is located
about one and half miles to the northwest of
Pinellas Point and may also have served as a
source of freshwater. Flint spalls have been
found along the edges of this lake.
The Pinellas Point Site should be considered
in view of its longevity as a site of repeated
prehistoric and historic occupation. Though sim-
ilar in many respects to other middens along the
Tampa Bay coast and the Gulf coast in general,
Pinellas Point does seem somewhat unique in
a practically complete representation of all Gulf
coast archaeological periods. Beyond the ob-
vious desirable qualities for survival such as
fish, shellfish, wild game, and probably fresh
water, this author feels that this location was
particularly advantageous or strategic since
it was contacted by so many people over such a
large span of time.
Our knowledge of Pinellas County and of this
bay area as a whole, though limited by the lack
of stratigraphic data, is nevertheless increased
by this present survey and hopefully it will
benefit archaeology elsewhere in Florida.
I wish to express my appreciation to Walter
Askew, Mark Brooks, Richard McDonnel, and
Phil Jordan for the use of their collections and
their helpful discussion. I would particularly
like to thank Frank Bushnell for his assistance
in measuring the site, analyzing the food re-
mains, and taking the photographs accompany-
ing this paper. Finally, I wish to thank Ripley
P. Bullen who identified the sherds and some
of the projectile points, and who was always
ready to help.
Bullen, Ripley P.
1968 A Guide to the Identification of Florida
Projectile Points. Florida State Museum.
1966 Burtine Island, Citrus County, Florida.
Contributions of the Florida State Museum,
Social Sciences, No. 14. Gainesville.
Bullen, Ripley P. and Walter Askew
1965 Tests at the Askew Site, Citrus County,
Florida. The Florida Anthropologist, Vol. 18
No. 4. Gainesville.
Bullen, Ripley P. and Adelaide K. Bullen
1963 The Wash Island Site, Crystal River,
Florida. The Florida Anthropologist, Vol.
16, No. 3. Tallahassee.
1962 The Maximo Point Site 1962. The Flo-
rida Anthropologist, Vol. 15, N5. 4. Talla-
Sears, William H.
1960 The Bayshore Homes Site, St. Peters-
burg, Florida. Contributions of the Florida
State Museum, Social Sciences, No. 6.
Ware, John D.
1968 A View of Celi's Journal and Chart of
1757. The Florida Historical Quarterly, Vol.
67, No. 1. Gainesville.
Warren, Lyman O.
1964 Possibly Submerged Oyster Shell Middens
of Upper Tampa Bay. The Florida Anthro-
pologist, Vol. 17, No. 4. Gainesville.
Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast.
Smithsonian Miscellaneous Collections,
Vol. 113. Washington.
THE APOLLO BEACH SITE, HILLSBOROUGH COUNTY
Lyman O. Warren
The Appolo Beach site consists of cultural
material dredged from Tampa Bay and deposited
as fill for a real estate development. During more
than two years of collecting from the site, a
number of chipped stone artifacts including pro-
jectile points, scrapers, blades, and cores were
reclaimed as well as a range of pleistocene fos-
sils. Sherds collected from the dredge fill ranged
from fiber tempered types through Papys Bayou
specimens of the Weeden Island phase.
The real estate development known as Apollo
Beach is located on the east shore of Tampa Bay
in Ruskin, on land formerly planted in tomatoes
and owned by the Dickman family. The only
identified Indian site was uncovered in the ex-
treme northwest corner of the development, about
two hundred yards off shore from the present
upland, and about 1V mile south of where New-
man's Branch opens into the Big Bend section
of Tampa Bay. The sand for this artificial site
had been dredged and pumped from bay bottom
and deposited as fingers or fill on three sides
of a common channel which opened to the north
on the bay. The depth of the water at mean high
tide before dredging had been about three feet.
The depth from which the artifacts came was not
known. The channel was said to be 18 feet deep,
and here, as elsewhere at dredged up sites, did
not offer itself readily to underwater investiga-
tion of vertical site placement because of mud
and sand washdown along the edge of the fill.
The Indian site was first brought to my atten-
tion in 1961 by Mr. and Mrs. Dean H. Thompson
of New Port Richey. The Thompsons had walked
the greater part of the development on many
occasions and had amassed an outstanding col-
lection of Pleistocene fossils from near the In-
dian site and elsewhere. They had found no other
sites than the one reported here, and had noted
the significance of the nearby Newman Creek for
an aboriginal population. The concentration of
spalls and artifacts in the area suggested to
them that a dig might be worth while, and, as
they did not feel up to it themselves, they sug-
gested that I might undertake one. The Thomp-
sons have recently presented me with a list of
the Pleistocene fauna which they found at Apollo
Beach (Table 1).
About 300 yards west of the pumped up site
there is now (1968) located a Holiday Inn and a
Digging was done by Major George Robinson,
Francis Bushnell, Al Goodyear, Mark Brooks,
Walter Askew, Robert Ebersole, and the entire
Warren family. Gerald Spence was especially
active and encouraging, and generously pre-
sented a number of his finds to the author. The
dig went on for over two years, a somewhat un-
usual feature for this type of salvage archaeo-
logy, where sites are generally uncovered and
covered up again by the land moving machinery
in the space of a few months, weeks, or even
days. The prolonged study of this site was made
possible because the development corporation
went into bankruptcy and earth moving came to a
The topography of the site was obviously,
since dredged up, artificial. The eye was pre-
sented with an expanse of whitish sand, gener-
ally level, sprinkled with finely crushed beach
shell and pebble phosphate, and sparsely with
Pleistocene fossils and Indian artifacts. The
former were generally black; the latter often
brown in color. There was a tendency for fossils
and artifacts to be concentrated on shallow sand
moundlets. These low elevations had been form-
ed by the deposition of pumped up material at
the outlets of the dredge pipes. Here, the dis-
charged material had been subjected to a concen-
trating process. Heavier material like artifacts,
fossils, lumps of gray clay and bog iron, and a
heavy, coarse, brown sand had layered out at
and near the center of the little mounds, and the
water and lighter solid material had flowed on
down hill to the periphery. Occasionally a small
piece of rusty dredge pipe or fragment of broken
glass would be found. The greater part of the
Apollo Beach assemblage came from these hil-
locks, and generally was recovered by shaving
with trowel or flat bladed shovel along the lami-
nae of coarse brown sand which lay close to
the grayish clay.
Before describing the finds it might be well to
mention briefly what was not found. There were
no ground objects of stone or shell; no con-
sciously asymmetrical, askew or lop-sided
points; no Bolens, Suwannees, Arrodondos,
Greenbriers, Hernandos, Daltons, or Pinellas
points. There were no Clear Fork Gouges or
let the clear glassy silicifed coral shine through
Of the projectile points, 34 of a total of 37
were angle-notched (or stemmed points with deep
barbs), fairly well made, symmetrical bifacially
worked, and without any significant amount of
secondary retouching. They average 4.5 to 7
centimeters in length, and are rather broad, with
a maximum width of 3.5 to 4 cms. In thickness,
they average 8-10 millimeters. The faces show
neither a longitudinal medial ridge or crest nor
any plano-conves features. The edges are excur-
vate (convex) and the stems generally contract-
ing, but occasionally expanding around the broad
angle notches. The stem bases are usually con-
vex or slightly so, and the convexity of the
C d bases together with the excurvate edges of the
blade suggests that the points were chipped out
of lanceolate or "tear drop" shaped blanks by
removal of wide angle notches. This may be
visualized in Figs. lb, and 3a and c, where 3b
represents the blank.
Fig. 1. Typical Apollo Beach points.
"Horse's Hoof Scrapers." There was no evi-
dence of midden soil or midden shell and no d
bone points or objects of worked bone, nor was
there any discernible human skeletal material. c
A few crude sandstone objects were present,
but no grinders, manos, or metates.
The significant finds were chipped stone arti- f
facts and sherds. The former fell into 12 cate-
gories: (1) corner notched points; (2) stemmed
points (rare); (3) thick scrapers; (4) thin disk
scrapers; (5) hafted scrapers (popularly known
as "stunners"); (6) ovate knives; (7) large
knives; (8) bifacial core "choppers;" (9) uni-
facial core planes; (10) spalls; (11) spalls with
worked edges; and (12) cores.
The stone material was either of silicified
limestone or silicified coral, and while many
of the artifacts have a matt finish, others have
a sort of lustre or gloss, almost as if they had
been polished in a tumbling machine. Some are
outwardly chalky except where recent breaks Fig. 2. Typical projectile points.
TABLE 1. PLEISTOCENE FOSSILS IN THE THOMPSON COLLECTION
FROM APOLLO BEACH
(Courtesy of Mrs. D. H. Thompson)
Alligator, scutes and teeth
Armadillo, scutes and tooth (?)
Florida Glyptodon, scutes
Cat Family, several claws and teeth
Bird, many bones
Giant Beaver, teeth
Camel, teeth and bones
Deer, teeth and antlers
Fish, armor plate of sturgeon-like fish(?), unlike
those found in phosphate pits of Bartow
crushing plates of drum fish
crushing plates and skull piece of ray
Horse, teeth, many different sizes and patterns
Mammoth and Mastodon, teeth and tusks
Sloth, teeth and claw of smaller sloth
Whale, teeth and ear bones
Poipoise (?), ear bones similar to whale
Shark, teeth, several different kinds
Turtle, shell, many patterns and bones
Walrus (?), tusks of different sizes
Unidentified, many small teeth, small vertebrae, and
claws of various sizes
TABLE 2. SHERD COUNT, APOLLO BEACH
(Courtesy of Ripley P. Bullen)
Orange Plain, 16; some have a black, shiny appearance
like that found on many of the fossils.
Semi-fiber-tempered plain, 9.
Pasco Plain, 12.
Sand Tempered Plain, 33, sometimes called Glades Plain.
Saint Johns Plain, 6.
Clay-tempered plain, 2.
Pinellas Plain, 1; this does not imply a Safety Harbor
Deptford Linear Check Stamped, 2.
Deptford Linear Check Stamped or Deptford Simple, 1.
Papys Bayou Plain, a smooth, Weeden Island-like chalky
Perico Linear Punctuated, 1, probably finger-nail im-
Eighty-four of these sherds cover the Orange
(fiber-tempered), Transitional, Deptford, or Perico
Island time periods. The other two sherds may be more
or less disregarded as of no statistical importance.
They may represent odd variants of other types.
A second type of point (3 of a total of 37), was
characterized by a contract-stem with convex
base; edges straight, incurvate or slightly re-
curvate, and most significantly with square
shoulders set smartly at 90 degrees to the long
axis of the point (Fig. 3d).
"Thick scrapers," a total of 17, are plano-
convex in cross section; in outline they may be
circular, elliptical, semicircular, ovate, elon-
gate or irregular. They run about 3-4 cms. in
length or diameter, by about 8-10 mms. thickness
at the crown. Their distinguishing features are
(1) a sharply convex or snub nose bit as seen
in profile; (2) the bit and often the entire peri-
phery is very nearly pressure flaked; (3) they
are usually made of coral and have a more gem-
like lustre than the points, which are more often
of limestone with a matt finish; (4) despite the
elegance of most of them, at times the rough
cortex of the coral has been left on.
"Thin disk scrapers," 5 in all, are similar in
outline, quality, and technique of manufacture
to the thick scrapers with which they overlap
typologically. They are only 4 mm. or so in
thickness, and therefore do not have the snub-
nosed character of their bit as well developed
as the thick scrapers. If anything, thin scrapers
are more aesthetically pleasing, and, when they
first come out of their coarse sand or clay mat-
rix in a moist and glossy state, have an almost
"Hafted scrapers" are shown in Figs. 4a-d.
It can be seen from the photographs, I believe,
that there is a typologic overlap of hafted scrap-
ers with the thin and thick scrapers on the one
hand and the common Apollo Beach point on the
other. The faces may be either planoconvex or
"Ovate scrapers" resemble thin disk scrapers
but have an ovoid or egg shape. Their dimen-
sions are about 4-5 cms. in length by about 3-4
cms. in breadth. A total of eight are in the col-
lection, and two are bifacially worked.
A "drill" was reclaimed which, like the com-
mon points, has wide angle notches.
Pottery (Table 2) was scarce in the excavated
artificial sand tumuli, but surface sherds--along
a shallow ridge of oyster shell and coarse yellow
sand not more than a hundred yards away--were
notable in that the highest proportion of them
were plain semi-fiber tempered and black in col-
or, at times encrusted with rusty sand. One sherd
had a possible Safety Harbor type rim incision
on a St. Johns soft chalky ware.
A fragment of petrified turtle shell in a little
bolus of silicified clay was found. This is a
composite concretion possibly formed by the
tumbling action of an early river bed, probably
I I 'D C
Fig. 3. Apollo Beach points.
The Apollo Beach assemblage does not give
the impression of a broad time spectrum; on the
contrary, the typology seems remarkably circum-
scribed. If we may rely on the typology, the
angle-notched projectile points are coeval. And
the scrapers, thin disk and thick, are so similar,
one to the other, that they, too, must be of a
piece. They and the angle-notched points seem
to be tied in by the hafted scrapers, which have
the cameo-like qualities and snub-nosed bit of
the scrapers and the deep barbs of the project-
Where the larger tools and pottery fit in is,
of course, conjectural. It is possible that more
than one time period is represented. Data from
elsewhere suggest a late Archaic or Transition-
al time period for the Apollo Beach site. Ripley
P. Bullen has described a similar stone assem-
blage from Zone 9, a transitional or late fiber-
tempered zone, at Site J-5 on the banks of the
of the Chatahoochee River, with a radiocarbon
date of 1200 B.C.+ -150 years (Bullen 1958).
Francis Bushnell and Al Goodyear have found
similar points and scrapers associated with
Perico ware near Canton Street in St. Petersburg,
near Maximo Point. Mark Brooks has had observ-
ed in a narrow sand zone an association of simi-
lar points and scrapers with St. Johns Incised
chalky ware at Flint Ridge, New Port Richey.
At Culbreath Bayou in Tampa, Bullen, Warren,
and Thompson have described a similar complex
of chipped stone tools in a non-ceramic context,
undoubtedly a stone workshop (Warren, Thompson
and Bullen 1967). Typical.points from this site,
named "Culbreath," seem closely related to
points from Apollo Beach, Canton Street, and
Bullen, Ripley P.
1958 Six Sites near the Chatachoochee River
in the Jim Woodruff Reservoir Area, Florida.
Bureau of American Ethnology, Bulletin 169,
River Basin Survey Papers, No. 14. Wash-
Warren, L.O., W. Thompson and R. P. Bullen
1967 The Culbreath Bayou Site, Hillsborough
County Florida. The Florida Anthropo-
logist, Vol. 20, Nos. 3-4, pp. 146-163.
St. Petersburg, Florida
Fig. 4. Halted scrapers and points.
THE FLORIDA ANTHROPOLOGIST
A LITHIC DAGGER
FROM CHOCTAWHATCHEE BAY
Donald W. Sharon
This unique specimen (Fig. 1) was recovered
from the shallow waters of Choctawhatchee Bay,
Florida, by Joyce Sharon, wife of the author, on
site 8W61 (Eden), which is known to contain
Weeden Island II and Ft. Walton phase ceramic
materials. The shallow waters adjacent to the
site have produced numbers of projectile points,
two of which are considered by the author to be
good examples of Daltons. No lithic artifacts
have been found in situ on the site. In view of
the foregoing it is apparent that the specimen
in question cannot be assigned to any specific
The dagger measures 137 mm. in overall length.
The stem side edges and base are ground. The
material is light gray in color and appears to be
flint which is similar to the well known Duck
River, Tennessee deposits. This material is of
course an import and is considered exotic for
this area of Florida.
The function of this specimen is also prob-
lematical, it could have been utilized for a pro-
jectile point, knife, or as I have chosen to call
it, a dagger.
It is hoped that this short note may be of some
interest to the student of lithic materials.
Ft. Walton Beach, Florida
BEVELED STEMMED POINTS
FROM TAMPA BAY
Ripley P. Bullen
In 1954-55 the arrow and spear points in the
Florida State Museum's research collections
were divided into various typological groupings
in a preliminary attack on the problem of pro-
jectile point classification. Beveling of blade
edges was used as a primary sorting criterion
and, as a result, all beveled points were put into
one drawer. After this was done, it was noticed
that the collections included only five stemmed
points with beveled blades. This sorting indi-
cated the great scarcity of such points in Flor-
Fig. 1. Obverse (left) and reverse (right) sides
of lithic dagger from Choctawhatchee
Bay, Florida. Length of specimen, 137 mm.
Recently, Lyman C. Warren of St. Petersburg
found three similar points, the base of a Suwan-
nee point, and a St. John's Incised sherd in
spread oyster shell which came from the Benton
shellyard on U. S. 19 north of St. Petersburg.
JUNE SEPTEMBER, 1968
VOLUME 21, NOS. 2 & 3
Fig. 1. Beveled stemmed points and base of a
Suwannee point from Tampa Bay.
Due to the rarity of these points, he gave his
collection to the Florida State Museum (cat. no.
The three beveled points (Fig. 1, upper) are
fairly thick and clearly, albeit crudely, beveled.
In the largest example, the tang or stem is round-
ed while in the smallest, the tang may be called
squarish or rounded depending on the bias of the
examiner. The smallest point resembles the Stan-
ley Stemmed point of Alabama (DeJarnette, Kur-
jack, and Cambron 1962: 67).
Interestingly, the two smaller points have the
bevel on the right hand sides of their faces
(looking at the points from the top or tip end)
while the largest has the bevel on the opposite
side. Similarly, four of the five points in the
Museum's collection have right hand bevels. The
exception is an extremely broad point which is
unique in several respects.
Three of the points in the Museum's collection
exhibit contracting stems similar to but not as
rounded as the larger one in the Warren collec-
tion. Another has a squat, straight-sided stem
and barbs at the corners of the blade. The fifth
has an expanding tang and a somewhat side- or
corner-notched appearance. The fairly wide bot-
tom of the stem of this point shows basal grind-
Beveled stemmed points have not, I believe,
been previously discussed in The Florida An-
thropologist. Apparently, they are a rare item
archaeologically. We have no stratigraphic data
which might permit us to relatively date these
points. The following clues suggest, but it is
only a suggestion, that they may belong fairly
early in the Preceramic Archaic period of Flo-
Of the five points in the Museum's collections,
three are cataloged from the vicinity of Gaines-
ville and two from the eroded beach at Philippi
Hammock on the northwestern side of Tampa
Bay. The three in the Warren collection came
from Benton's U. S. 19 shellyard as mentioned
earlier. Shell in this yard had been dredged up
from the east central part of Tampa Bay. Other
specimens from this shell deposit range in time
from the Paleo-Indian to the early post-Orange
Transitional ceramic period.
These beveled stemmed points are thick and
crudely made. They do not approach the relative-
ly thinner and reasonably well chipped points of
the Archaic period of Florida. Single edge bevel-
ing is common on Bolen points which have been
placed in an early, post-Paleo-Indian, Archaic
period. Ground bases are characteristic of Bolen
points and one of the beveled stemmed points
from the Gainesville area has a smoothed base.
There is some slight suggestion that rounded
bases like that of the largest point in the Warren
collection may be early in the stemmed point tra-
The lack of beveled stemmed points in known
Late Preceramic deposits and the possible clues
given in the previous paragraphs suggest such
points belong in the little known Early Precera-
mic Archaic period. As we gain more knowledge
about that period, it appears to have more variety
in projectile point types and to have been a much
more dynamic period of culture change and vari-
ation than previously believed.
DeJarnette, David L., Edward B. Kurjack, and
James W. Cambron
1962 Stanfield-Worley Bluff Shelter Excava-
tions. Journal of Alabama Archaeology, Vol.
8, Nos. 1-2. University.
Department of Social Sciences
Florida State Museum
side. The end is rounded or snub-nosed to pro-
duce an effect like an end scraper. Whether it
was originally made as an end scraper, or wheth-
er a broken Suwannee was reworked into a hafted
end scraper is uncertain. A slight flaw in
the chert suggests the latter may be the correct
Fig. 1. Possible Suwannee end scraper from Cur-
lew Road (length of specimen, 8 cm.).
SUWANNEE STYLE END SCRAPERS
FROM PINELLAS COUNTY
Albert C. Goodyear, William Thompson
and Lyman O. Warren
The Curlew Road in Pinellas County was built
a couple of years ago. At its western end it a-
buts on U.S. 19, where its westerly extension is
known as the Caladesi or Dunedine Beach
Causeway, the site of a possible Paleo-Indian
chert workshop dredged from the bottom of St.
Joseph's Sound. As it proceeds to the east, it
winds along, following, in a general way, the
course of the Curlew Creek to its south. It cros-
ses U.S. 19, the "Sunshine Skyway." and circles
through rolling sand dune country now planted to
citrus. On one of these dunes, shortly after the
road was put through, Thompson found an inter-
esting artifact (Fig. 1) which is depicted in this
paper. At the suggestion of Ripley P. Bullen, the
scraper was photographed, and its characteris-
It is well proportioned and aesthetically pleas-
ing artifact, measuring about 9 cms. in length
and about 4 cms. wide, with a cream and orange
color, patinated, chalky, basally ground, tanged
or eared with one ear unfortunately broken off
and reconstructed to match its mate as shown in
the photograph (Fig. 1). The edges of the blade
appear to be pressure flaked; the faces are bi-
facially worked; there is no fluting or removal
of a longitudinal basal flake, but the basal half
is appreciably thinner than the distal. The work-
ing edge appears to be the end rather than a
Fig. 2. Possible Suwannee
end scraper from
In March. 1966. a similar end scraper was found
by Goodyear on a sand dune overlooking the har-
bor of Tarpon Springs. This artifact (Fig. 2). al-
though similar in shape to the other, is a little
smaller and has no ears. It does not have
the same degree of calcination, and the lower
edges and bases show no appreciable grinding,
but. on the contrary, are quite sharp.
It is possible that these Suwannee style end
scrapers were part of the tool kit of the Paleo-
Indian of Florida, but our present data does not
rule out the possibility that they were Suwannee
points reworked in Archaic times. This paper
may stimulate others to record similar "Suwan-
nee End Scrapers."
St. Petersburg. Florida
CALADESI CAUSEWAY: A POSSIBLE
INUNDATED PALEO-INDIAN WORKSHOP
Lyman O. Warren
The Caladesi Causeway, recently drag-lined
and pumped out of the bay bottom of St. Joseph's
Sound, connects the City of Dunedin with Dun-
edin Beach on Honeymoon (formerly Hog) Island.
Much of the fill material was a light gray mixture
of marl, sand and fragmented shell, mostly scal-
lop. Included were a great many nodules of sili-
cified limestone, generally spheroidal in shape,
which ranged from golf ball to basketball size.
Silicified coral heads were much rarer, unlike the
situation at the Trouble Creek workshop east of
Bailey's Bluff, located a few miles to the north
of the Caladesi Causeway, where silicified coral
heads greatly outnumbered silicified limestone
boulders. Again, in contrast, some of the Cala-
desi coral heads were unusually large, weighing
over 300 pounds. It seemed obvious that drag
lines and dredges had reached deep through the
mud and sand of the bay bottom to bring to the
surface great volumes of the underlying Haw-
In March of 1965, Mr. Charles Greth and Mr.
Edward Dunn informed me of chert debitage lo-
cated a couple of hundred yards to the northwest
of the third and most westerly bridge of the
causeway. At this location, both along the
newly created beach and up on the surface of
the fill, the ground was littered with countless
spalls and cores of silicified limestone. In gen-
eral, they measured 2-6 inches in length, and
were, on the average, larger than the chips and
chunks at most workshops in the Tampa Bay
area. Where waves and tides had eroded the
north-facing beach before the sea wall went in,
a scarp, some two feet in height, was formed,
and in its face spalls were abundant from top
to bottom, suggesting that the fill, in this area
of over 100 yards in diameter, was extremely
rich in artifacts. Test pits at three arbitrary
locations on the upland surface of the fill and
on the beach showed that this was indeed the
case, and in two pits, the spalls were so thick
they impeded the digging.
In this welter of surface debris, scores of
tools could be found by patient and repeated
searches. They were all heavily patinated, as
were the spalls. They were rust-brown or pearl
gray in color, or a mixture of browns and grays,
except where recent breaks disclosed the under-
lying blue-black of the original pre-patinated
flint. While many of these objects were presum-
ably blanks, others, more finished, appeared to
be completed tools. They were of crude manu-
facture, percussion flaked, inelegant, yet read-
ily discernible when spotted, from the hundreds
and thousands of spalls present. There were
infrequent hammerstones of varying shapes, in-
cluding a discoidal hammerstone, pounded on
the edge, with a scintillating geode on one flat
surfact, and a silicified hemisperical coral head
pounded on the convex surface, or pole.
Fig. 1. Bifacial choppers, or scraper-knives, from
the Caladesi Causeway
There were also about 20 percussion flaked,
circular, ovate, triangular, and rectangular kni-
ves, measuring about 2-3 inches in length, and
distinguished from choppers by their relative
thinness. Heavy plano-convex scrapers or scrap-
ing planes presenting crowns of various shapes
and degrees of convexity, with plane surfaces
of an oval or elliptical outline were more num-
erous. Bifacial choppers, or scraper-knives,
were numerous. Some were of considerable sym-
metry and attractiveness; others, a total of
twenty, had characteristic kidney shape (Fig.l).
These "kidney choppers" seemed quite unique
to the Caladesi site. Their kidney shape shows
up much better in nature than in the photos. One
or two football sized boulders, trimmed on sev-
eral faces, and pounded on one, suggested huge
sledge hammers or possible anvils, but were
probably extra large cores.
Over several months, the collection of crude
objects grew. Dozens of field trips brought back,
with monotonous regularity, only more examples
of the above tool types. We kept looking for
some sort of index artifact' which might tie down
in time or culture this unusual workshop; un-
usual in part because it was so devoid of pro-
jectile points. Finally, in July of 1965, the base
of a Suwannee point turned up.
This base (Fig. 2, center) was lying in the
eroded face of the spoil bank facing the beach
and had been exposed after an unusually high
tide. It was crudely made, percussion flaked,
unfluted, blue-gray in color with rust-brown
stains, parallel edges, heavy basal grinding,
and rather thick or disc-shaped in cross sec-
tion; it would have been, if complete, much
longer than the average Suwannee of the Tampa
Bay area. A finely made, pressure flaked Suwan-
nee from the Kellogg Fill in Boca Ciega Bay is
shown for comparison (Fig. 2, left). The Cala-
desi Suwannee base seemed a sort of an oddity,
and yet its general appearance and heavy-handed
manufacture was similar to the other tools of
the site. Nearby was a typical "kidney chopper"
of the same color and patina and in the same
Fig. 2. Suwannee point (left); base of Suwannee
point (center); expanded stem point
In February, 1966, at the annual meeting of the
Florida Anthropological Society in Clearwater,
John White showed me a point (Fig. 2, right)
which he had found in the same sort of debris
described above, on the beach not far from where
the Suwannee point was collected. It was of an
expanded stem variety, with grinding on both
edges of the stem, and with beveling on one
edge of the blade. It was rather wide in stem
and blade, 2.7 and 3.8 cms., respectively, in
relation to its length of 6 cms. In cross section,
this point is disc shaped and not bifacially con-
vex. It is patinated, rusty in color, and looks as
if it were made of the same material as the other
artifacts from the site. It reminds one of a bevel-
ed Bolen point trying to evolve into a stemmed
As far as I know, these two projectile points
are the only ones found at the Caladesi site and
they probably represent many hours of searching.
Mr. White also found a mineralized bone point
on this site. The presence of the two early
points depicted in this paper, in the absence of
anything from a later period, suggests, although
it does not prove, that we are dealing with a
Paleo-Indian or very early Archaic workshop.
The profusion of stone debitage is, of course,
in favor of a workshop, rather than a habitation
or habitation-workshop site. With the exception
of Mr. White's bone point, there have been no
animal, shellfish, or midden soil residues.
The workshop was attractive to the early peo-
ple who used it because of the readily acces-
sible outcropping of workable chert. It would
seem that the site was abandoned in late Paleo-
Indian or early Archaic times, some eight or nine
thousand years ago, when St. Joseph's Sound
was first formed by the rising waters of the Gulf
I wish to acknowledge my indebtedness to Mr.
Charles Greth and Mr. Edward Dunn for informing
me of the site and for allowing me to inspect
their collections and to Mr. John White for per-
mission to publish the photograph of his pro-
jectile point. Mr. Ripley P. Bullen gave many
helpful suggestions in preparing and editing the
manuscript. The photos were taken with a Kodak
A dredge captain employed by the Webb and
Knapp Corporation told me that he had dredged
down to a depth of 18 feet below mean high tide,
and that in his opinion most of the stone mate-
rial had come from the ship's channel under the
bridge. It is impossible to tell from just what
depth this assemblage came. The bay bottom in
this section of St. Joseph's Sound was originally
only two or three feet below the surface of the
St. Petersburg, Florida
3 42 INC
3 4 5 CM
Fig. 1. Spurred scraper from Florida
(Lower photo enlarged X 2.5)
A SCRAPER WITH GRAVER SPURS
Ripley P. Bullen and Marjorie H. Wing
One of the characteristic tools of the Paleo-
Indian period is a small end scraper with graver-
like spurs at the outer ends of the main scrap-
ing edge (Byers 1954: Fig. 92a). Fig. 1 illu-
strates the first tool of this type found in Flo-
rida. These specialized scrapers permit scrap-
ing and incising without the necessity of chang-
In 1958 Bullen directed excavations at Bolen
Bluff on the south side of Paynes Prairie, a
little south of Gainesville, Florida, as part of
the Florida highway salvage program. This site
was known to have produced several Suwannee
points and it was hoped excavation would re-
veal a Paleo-Indian zone. While fragmentary
Suwannee points were uncovered, no such zone
was found. The earliest excavated level, which
was on the west side of the road, produced what
are known locally as Bolen points (Bullen 19-
58), typologically similar to the beveled Big
Sandy 1 points of Alabama (DeJarnette, Kur-
jack. and Cambron 1962). Presumably, Bolen
points would date to about the same time period,
or around 7300 B.C. CDeJarnette, Kurjack, and
Cambron 1962: 85-86).
Work on the western side of the road was lim-
ited in area due to the small space between the
edge of the old road cut and that of the new
right of way. Subsequently, when the space be-
tween the excavation and the old road was re-
moved by large dirt-moving machinery, Wing
examined the material being removed and col-
lected, along with other specimens, the spurred
scraper illustrated in Fig. 1. In 1966, while
preparing for an eighteen months stay in Brazil,
Wing re-examined her collection and brought
this scraper to the Florida State Museum for
safe keeping during her absence from the coun-
try. Because of the apparent importance of this
specimen, both authors thought it desirable to
record it without further delay.
The scraper is made of a local chert which
contains geode-like inclusions (Fig. 1). Time
has given it a white-ivory color like that of
many Suwannee points, but originally it must
have been much darker. The back exhibits a
typical unworked concoidal fracture surface and
part of the striking platform is still present at
the narrower or butt end (Fig. 1, right). The
greatest thickness, 8 mm., is near the center a
little towards the narrower end. One chip was
removed from the top to make the wide or scrap-
er end about 5 mm. thick. Sides have been shap-
ed in a scraper-like fashion but the only real
scraping edge is at the wide end between the
two spurs. This edge has some fine, step-like
fractures, undoubtedly the result of use. Of the
two spurs, the one on the right was intentionally
left as a point during manufacture. The one to
the left has been generated by chipping and is
the more protruding of the two (Fig. 1, left).
Edges of chipping scars of the large left hand
spur are dull or rounded. It seems reasonably
certain that this part of the scraper was used as
This spurred scraper closely resembles those
from the Bull Brook (Byers 1954: 348, a), Shoop
(Witthoft 1952: P1. 4, 13-17), Williamson (Mc-
Cary 1951: Fig. 9, 1-12, especially 9), Harda-
way (Coe 1964: Fig. 64c), and Quad (Soday
1954: 12, 23-25) sites in the eastern United
States, and the classic Folsom site in Colorado
(Roberts 1936: P1. 8). The only difference be-
tween spurred scrapers from the above sites and
the Florida specimen is that the spurs on the
last are more exaggerated.
The Florida spurred scraper came from a site
which has produced nearly 20 whole or fragment-
ary Suwannee points which represent the region-
al equivalent of a late Clovis or Paleo-Indian
time period. This additional evidence strongly
implies the presence in Florida of man during
the Paleo-Indian period.
Bullen, Ripley P.
1958 The Bolen Bluff Site on Paynes Prairie,
Florida. Contributions of the Florida State
Museum, Social Sciences, No. 4. Gaines-
Byers, Douglas S.
1954 Bull Brook--a Fluted Point Site in Ip-
swich, Massachusetts. American Antiquity,
Vol. 19, No. 4, pp. 343-351. Salt Lake City.
Coe, Joffre L.
1964 The Formative Cultures of the Carolina
Piedmont. Transactions of the American
Philosophical Society, N.S., Vol. 54, Pt.
5, pp. 1-130. Philadelphia.
DeJarnette, David L., Edward B. Kurjack and
James W. Cambron
1962 Stanfield-Worley Bluff Shelter Excava-
tion. Journal of Alabama Archaeology, Vol.
VIII, Nos. 1-2. University.
McCary, Ben C.
1951 A Workshop Site of Early Man in Din-
widdie County, Virginia. American Antiqui-
ty, Vol. 17, No. 1. pp. 9-17. Salt Lake City.
Roberts, Frank H. H.,Jr.
1936 Additional Information on the Folsom
Complex. Smithsonian Miscellaneous Col-
lections. Vol. 95, No. 10 Washington.
Soday, Frank T.
1954 The Quad Site, a Paleo-Indian Village
in Northern Alabama. Tennessee Archaeo-
logist, Vol. 10, No. 1, pp. 1-20. Knoxville.
1952 A Paleo-Indian Site in Eastern Penn-
sylvania: An Early Hunting Culture. Pro-
ceedings of the American Philosophical
Society, N.S., Vol. 96, No. 4, pp. 464-
Florida State Museum
INFORMATION FOR AUTHORS
The Florida Anthropologist publishes original papers in all subfields of anthropology, with an
emphasis on archaeology. Contributions from allied disciplines are accepted when they concern
anthropological problems. The geographic scope of the journal, in order of importance, includes
Florida, the Southeastern United States and related regions, North America, and other areas.
Manuscripts will be published in order of receipt within the limits of space and issue balance,
except those contributions which are solicited by the editor. Authors will generally be notified
of acceptance, rejection, or need for revision within 3 to 5 weeks following receipt of a manuscript.
Preparation of Manuscripts
All manuscripts submitted for publication should be typed double spaced on one side only of a
good grade of white bond paper, leaving more than one inch margin on all sides of the page.
Pages which have numerous corrections should be re-typed. All parts of every manuscript, in-
cluding text, quotations, tables, abstract, lists, acknowledgment, and references should be typed
double spaced. All manuscripts longer than four pages must be accompanied by a brief abstract
written in an informative and succinct manner; manuscripts of four pages or less will be included
under "Brief Notes."
Reference and Style
Contributors are referred to American Antiquity, the journal of the Society for American Archaeology,
for matters of style and reference. All manuscripts, with the possible exception of "Brief Notes,"
should list references at the end of the text, and should limit the list to only those references
cited in the text. All books, monographs, and articles should be listed alphabetically by author,
and chronologically by work. Do not abbreviate the names of series or journals. The listing for
a book should include the full title, publisher's name, and place of publication.
Notes and Acknowledgments
Footnotes should not be used unless absolutely necessary. If the information is pertinent to the
discussion, include it in the text. Acceptable footnotes are those stating that the paper is a
revision of one offered at a professional meeting, or acknowledging permission to publish. General
acknowledgments should be included following the text.
Twenty-five reprints without covers are furnished free to each author. Those authors desiring
more than this number will be charged for the .cost of printing involved.
Illustrations and Tables
All illustrative material, drawings, maps, diagrams, charts, plates, and photographs should be
included in a single numbered series and designated "Figures," abbreviated as Fig. (Figs.)
in the text and legend. Never use the term "Plate."All tabular material should be included in
a separately numbered series of "Tables." All figures must be of a size suitable for maximum
reduction to( 6 X 8V2" including caption. If the original illustration is too large to be mailed,
a photograph suitable for reproduction should be submitted instead of the original. Photographs
should have a glossy surface and be sent unmounted. Drawings should be made with india ink
on a hard, smooth-surfaced, white board. When several objects are shown in a single illustration,
each object of the figure should be designated with a lower case letter, not a number or a capital
letter. Each figure and table must be accompanied by a brief descriptive caption. Captions should
be typed double spaced on a separate sheet of paper. Specific references in the text should
indicate what portion of the manuscript the illustrative material is intended to support.