2000 Florida Anthropological Society Inc.
The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
all rights to the
and shall be
and images of
The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.
The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.
Contact the Florida Anthropological Society for additional
information and permissions.
THE FLORIDA ANTHROPOLOGIST
Vol. IV. November, 1951 Nos. 3-4
Origins of the St. John's River Seminole: Were They Mikasuki?
------------------- -Kenneth W. Porter 39
S. T. Walker, An Early Florida Archeologist
------------------- -Ripley P. Bullen 46
Archeological Notes on Lower Fisheating Creek
-------------------- -John M. Goggin 50
An Analysis of Belle Glade Plain Rim Sherds from Two Fisheat-
ing Creek Sites
---------------- -Rita Krestensen Porter 67
Excavations at Kolomoki, Season II-1950 by William II. Sears
------------------- -Ripley P. Bullen 76
A Survey of Indian River Archeology, Florida by Irving Rouse;
Chronology at South Indian Field, Florida by Vera Masius
- - - - -John M. Goggin 77
Contributors to This Issue
Published at the University of Florida
ORIGINS OF THE ST. JOHN'S RIVER SEMINOLE:
WERE THEY MIKASUKI?
Kenneth Wiggins Porter
The confusion and obscurity in regard to the origins and relationships of
the various groups and tribes usually referred to collectively as Seminole make
welcome any ray of light, however flickering and dim, particularly when it is
directed upon some of the more important chiefs and leaders.
The principal figures on the Indian side in the Seminole War, 1835-1842,
are usually identified in current literature either by tribes or by their relation-
ships to earlier chiefs whose tribal origins are well known. Head-chief Mikonopi
(Upper Chief) was the nephew and successor of Payne and Bowlegs, sons of
Cowkeeper, founder of the "Seminole nation." Holatoochee (Little Chief), some-
times called Davy, was Mikonopi's nephew, or possibly brother. Holata Micco
(Chief Governor), better known as Billy Bowlegs, was also Mikonopi's nephew.
Alligator (Halpata Hajo), also known as Eufala Micco, is said to have been a
brother of Billy Bowlegs. Yaholocchee (Cloud) was also apparently some kin
to Alligator. The above seem all to have been members of the Alachua group,
the "original" Seminole or Seminole "proper."
Sam Jones (Apiaca), John Hicks, otherwise known as Tuckabatchee Hajo
(Mad Tuckabatchee), Halleck Tustenuggee (Potato Warrior), and Chitto Tusten-
uggee (Snake Warrior) were Mikasuki. Echo Emathla (Deer Leader), Nethlock-
emathla (Big Lieutenant Leader), and the latter's brother Thlocklo Tusten-
uggee (Fish Warrior), better known as Tiger Tail, were Tallahassee. Jumper,
sometimes called The Lawyer, whose Indian name was Oti Emathla-Island
Leader or, perhaps, Home Leader--was an old Red Stick Creek. Asi Yahola
(BJack Drink Hallooer), better known as Osceola, and also frequently called
by his step-father's name of Powell, was also a Creek, from Georgia. Euchee
Billy and Euchee Jack were chiefs of the tribe from whom they drew their
nicknames. Chekika was chief of the "Spanish Indians," or Calusa.
This list, however, omits one of the most important chiefs and one of the
most important war-leaders of the period: King Philip (Emathla-Leader) and
his son Coacoochee (Wild Cat). King Philip, a man of 60 at the outbreak of
the war, was the principal chief of the St. John's River Seminole, and of the
"royal blood." IIe was married to one of Mikonopi's sisters. His second in
command was Coi Iajo (Mad Panther or Mad Partridge). His village, in 1823
and probably later, was Yulaka, on the west side of the St. John's, 35 miles
from Volusia. Iis favorite, though not his oldest son, was Coacoochee. He is
not, however, further identified, tribally or by ancestry (Sprague, 1848:97-101;
Lieutenant . 1836:19-20; Swanton, 1922:412).
One is inclined, when no tribal designation is attached to a chief's name,
to assume that he was one of the Alachua group of "original" Seminole, and
Philip's marriage to a sister of the Alachua chief Mikonopi might seem to
support this assumption, though it should be noted in this connection that an-
other sister was married to the old Red Stick Creek, Jumper. Testimony recently
encountered, however, suggests that Philip, Coacoochee, and their people may
have been an offshoot of the Mikasuki.
A contemporary account, prepared at least as early as 1840, states: "In the
summer of 1837, Tuskeneha, a brother of King Philip .... having been at-
tacked with measles, committed suicide by shooting himself through the head"
(author's italics, Officer...., 1840:142). A chief named Tuskiniha (Sapsucker
Lieutenant; Read, 1934:77) was prominent during the 1820's. He signed as
"Tuskaneha" the Treaty of Fort Moultrie, September 18, 1823, also signed,
among other, by Mikonopi and Philip. He endeavored, in April, 1828, to enforce
the death sentence for the murder of a white man, and appeared at a council,
October 19, 1828, with the senior John Hicks-Tokosi Emathla (Mole Leader),
and others; he was referred to as "Tuskaneha, chief of the Mickasuky, or
Muscogee band." He could not have been both a "Mickasuky" and a "Mus-
cogee," as the Mikasuki, like the Alachua, were Hitchiti-speaking, but that
the former designation was correct is indicated by the fact that in 1823 "Tus-
kameha (Taski heniha)" was chief of "New Mikasuki, somewhere near Green-
ville, in Madison County, 30 miles west of the Suwanee River." "Tuskenaha"
was also at a council at the Seminole agency, January 14, 1829, with John
Hicks, Sr., and "Coahadjo" (Sprague, 1948:22, 49, 64, 67; Swanton, 1922:401,
The question naturally arises: Granted that King Philip had a brother by
the name of Tuskiniha, and that the Chief of New Mikasuki also bore that name,
how be sure that they are identical? Such a question is particularly pertinent
when one considers the rather unimaginative character of Seminole busk-names
and the excessive duplications which result therefrom. "Tuskinia, lieutenant
of Chatoackchufall (Chatuckchufaula)," in Georgia, appears on an 1806 list;
he may have been the same as the "late chief Tuskeneha," a Georgia Creek,
mentioned March 31, 1837. Neither, however, can be identified with a brother
of King Philip or a chief of New Mikasuki (Swanton, 1922:410).1
A letter from an unusually authoritative source probably gives the answer.
Capt. John C. Casey, Acting Seminole Agent, a conscientious officer who had
to be particularly well informed as to the relationships among his wards be-
cause of their effect on the property interests which it was his duty to safe-
guard, wrote on July 11, 1838: "The old Mickasukey king,....King Kinhaitsee
made his will, bequeathing all his negroes to his nephew, according to the old
Indian law of descent. After his death, Micco Potokee declined taking them at
first, and told the son (Tuska Gueha) and daughter of his uncle to keep them.
1American State Papers, Military Affairs (hereafter referred to as MA), vol. 7:868.
Tuska Gueha killed himself...." (author's italics, Poinsett, 1838:118-121).
"Tuska Gueha," to be sure,does not look a great deal like Tuskiniha, in
any of its various forms, but the poor penmanship of army officers and Indian
agents frequently and notoriously combined with government printers' ignorance
of Indian linguistics to make nonsense out of Indian proper names; in this
particular document, indeed, distortion is not confined to Indian names, Murray
being turned into "Munay" and Perryman into "Penyman." That "Tuska Gueha"
is, indeed, the Mikasuki chief Tuskiniha is confirmed by a documentary refer-
ence of August 15, 1850, to "Molly or Mah-kah-tish-chee, daughter of Tos-kee-
nee-hah a Mliccasuki chief and granddaughter of Kin-hij-chee former principal
Mlicasuki chief." (National Archives, War Dep't., I letters Rec'd., Aug. 15,
"Tuska Gueha" is thus definitely "Tos-kee-nee-hah" and it seems very
doubtful that two Seminole chiefs named Tuskiniha should have committed
suicide. If, then, one accepts the identification of King Philip's brother with
the son of the Mikasuki king, it would make Philip also a son of that multi-
named chief and would account for his "royal blood" without the necessity of
assuming a descent from the Alachua chief, Cowkeeper (see appendix).
The case for the identity is not water-tight. That King Philip had a brother
named Tuskiniha is definitely attested by only a single anonymous authority.
That a chief named Tuskiniha was closely associated with Philip, whether his
brother or no, is, however, abundantly witnessed. "The principal chiefs on the
St. John's, Tuskinia and Emathla, (Philip)" are mentioned on March 26, 1837,
and Tuskiniha continues to be mentioned throughout the summer, until as late
as August 20, always in association with one or more of the well-known St.
John's River chiefs, Philip (Emathla), Coi I!ajo, and Wild Cat (Coacoochee);
frequently, perhaps it should be added, he is mentioned in association with
Mikasuki. One such grouping, indeed, seems to turn back the pages of history
by over eight years. "Coe Hajo, a principal chief, Tuskaenaha, a sub-chief of
Philip's, and Ilicks, a sub-chief of the Mickasukies" are mentioned as being
together on August 20, 1837. The three signers of a "talk" at the Seminole
agency, January 14, 1829, were John Hicks, "Coahajo," and "Tuskenaha."
The Hicks of 1837 was the son of the John Hicks of 1829, but it seems more
probable than not that the Coi Hajo and Tuskiniha of 1829 and 1837 were iden-
tical (MA, vol. 7:835, 870-871, 844, 846, 879; Sprague, 1848:67).
If King Philip had a brother named Tuskiniha who committed suicide, it is
still not absolutely certain, though very likely, that he is to be identified with
the Mikasuki chief of that name who also made away with himself. One definite
incongruity should be mentioned. King Philip's brother is said to have killed
himself in the summer of 1837, whereas, according to Casey, the son of the
Mikasuki king would seem to have committed felo de se at least two years
earlier, since it is said that Micco Potoka, who "by the double claim of the
will and relationship, took possession of the negroes....caused the matter to be
examined by the agent, the late Wiley Thompson," killed on December 28, 1835,
at the beginning of the Seminole War. It would be easy, however, to be confused
about the agent before whom the deceased's property was claimed or, indeed, as
to the time and circumstances of the claim, which might have been in anticipa-
tion of Tuskiniha's decease rather than pursuant thereto.
If we accept Tuskiniha, chief of New Mikasuki, 1823, as identical with the
St. John's River chief Tuskiniha, 1837, the interim should be accounted for. The
name Tuskiniha does not appear, to my knowledge, between 1829 and 1837, and
is not definitely identified with New Mikasuki after 1823. The Treaty of Fort
Moultrie, 1823, however, gravely dislocated the Indian population of Florida,
and it would not have been surprising if a chief of New Mikasuki, west of the
Suwanee, should have been forced thereby to remove himself to the St. John's
and to associate himself with a better established kinsman, perhaps in a sub-
Other indications, in addition to the identification of the St. John's River
chief Tuskiniha as King Philip's brother and as the whilom chief of New Mik-
asuki, contribute to the possibility that the St. John's River Seminole had
Mikasuki affiliations, though these are at the most confirmatory and would hard-
ly be noticed without the leads furnished by the 1840 article and Capt. Casey's
One of Philip's sons stated in 1837 that "his father had required Abiaca,
(Sam Jones), chief of the Micasukies, to join him" (MA, vol 7:835, Mar. 26, 1837;
867, Apr. 9, 1837; 870-871, May, 1837). Why was it over a Mikasuki chief that
Philip asserted such an authority? Was it as grandson of the old Mikasuki king?
General Jesup asserted that "Coacoochee, John Cavallo,...with several
others, carried the hostages off;" Captain John Page stated that 200 Mikasuki
carried away the hostages Jumper and Mikonopi. 2 Coacoochee is assigned prin-
cipal responsibility by the general; the captain assigns the actual agency to
Mikasuki. A party of Indians friendly to the whites, visiting among the hostiles
about this time, reported that "we reached a party of Coe Hajo's people....we
saw, very few men, as they had gone to take Micanopi" (MA, vol. 7, 875). Coi
Hajo was the St. John's River second chief; were these men of his the "Mik-
asuki" reported to have carried away the hostages and, if so, had they done so
at the orders of the head-chief's favorite son?
When Coacoochee was captured October 21, 1837, the chiefs captured with
him included Coi Hajo, his father's second-in-command; Tustunukkee and Chitto
2Niles Register, Dec. 23, 1837, P. 263; National Archives, Department of the Interior,
Indian Office, Seminole (Emigr) File, 1837 (PL 36), Capt. John Page, Tampa Bay,
June 5, 1837, to C. A. Harris.
Yahola, apparently two of Coi Hajo's subordinates; Osceola, the young Georgia
Creek who, not being a chief, recruited his followers wherever he could, prin-
cipally among Mikasuki and Negroes; Micco Potoka, nephew of the late Mikasuki
king; John and Joe Hicks, sons of the late Mikasuki chief John Hicks; John
Cowaya or Cavallo, Negro-Indian emissary from Mikonopi and his associates; and
four or five others unidentified. In other words, outside his own St. John's River
group, of his identifiable companions one was an envoy from the Alachua, one
was a Creek commander of Mikasuki, and three--more than half--were unques-
tionably Mikasuki; of the nine identified, seven were St. John's River Seminole
and Mikasuki. And after Coacoochee's escape from Fort Marion the following
month, he hastened to join the Mikasuki chief Sam Jones--whom his father,
earlier in the year, had "required" to join him. These associations appear rather
more than coincidental.3
The closest voluntary associations of the St. John's River chiefs Philip,
Coacoochee, Coi Hajo, and Tuskiniha appear, then, to have been with such
\ikasuki as Micco Potoka, Sam Jones, and the Hickses, and such a commander
of Mikasuki as Osceola. One may wonder, in the last case, why it was with the
St. John's River chiefs that Osceola chose to associate himself when driven
out of his own domicile of the Cove of the Withlacoochee if not because his
followers were affiliated in some way with the St. John's River people.
The St. John's River chiefs and their people are never, to be sure, definite-
ly referred to as Mikasuki. If they were this roundabout argument would be un-
called for. They are, however, never given any other tribal designation and are
merely identified geographically as being located on the St. John's River. Pos-
sibly they were a group which separated themselves from the main body of Mik-
asuki and their chiefs, while both sides retained a clear knowledge of the con-
nection-just as Cowkeeper and his people, settled on the Alachua savannah
and coming to be know as Seminole--seceders'-- remembered themselves to be
originally Lower Creeks and were themselves so remembered.
"The Mikasuki King"
The Mikasuki king, who unquestionably had a son named Tuskiniha and may
also have been the father of King Philip (Emathla), bore a name which appears
in a greater variety of forms than that of almost any other Seminole chief of
my knowledge. One of his names was Cappichamicco, more properly Kapitca
miko (Lyewater King),4 but he was more commonly known under the name which
3Niles Register, Nov. 4, 1837, p. 146, Nov. 11, 1837, p. 165; Cooper, 1838; Forry,
1928:94; National Archives, War Dep't., AGO, Maj. Gen. T. S. Jesup, St. Augustine,
Oct. 21, 1837, to Brig. Gen. R. Jones (207); MA, vol. 7:846, Aug. 8, 1837.
4This name appears in 1806 as "Copixtsy mico, of Mickacuky" (Swanton, 1922:410).
in many versions alphabetically arranged, appears below. I have been unable to
ascertain its meaning, or even whether the first syllable is the English title
"king," frequently in a corrupt form, or is of Seminole origin. "Hachi" means
water and "hajo," mad, but in many of the versions the final two syllables bear
little resemblance to either of these words. The Mikasuki king was reported in
1818, at the time of the First Seminole War, as 70 or 80 years old (Young, 1934:
84); he probably died not long after the destruction of his village by General
Jackson. The references to the various versions of his name, below constitute
a brief bibliography for the few known facts of his long life:
Kenhaga (MA, vol. 1:704), Kenhagee (MA, vol. 1:700,716; American State
Papers, Foreign Relations, vol. 4:574, 575, 577, 583-585, 589-591, 607, 611),
Kenhajah (MA, vol. 1:716), Ken-ha-jah (Poinsett, 1838:60), Kenhijah (MA, vol.
1:682), Kenhutry (Williams, 1827:191), Kinache (Cooper, 1838:60), Kinghajah
(MA, vol. 1:703), Kinghajoh (Monroe, 1818:83), King Hatchy (MA, vol 1:722-
724, 727), King Heijah (Swanton, 1922:407), King Higo (Brevard, 1924:80),
King Hijah (MA, vol. 1:729), King Hijo (Simmons, 1822:47), Kinhaitsee (Poinsett
1838:120), Kinhaizee (Brevard, 1924:1:21-22, 50; quoting from Long, 1882:41),
Kinhega (Young, 1934:84), Kinhigee (MA, vol. 1:727), Kinhijah (MA, vol. 1:
724,752), Kin-hij-chee (National Archives, War Dep't., Letters Rec'd., Aug. 15,
1850, A135), Kinhijee (American State Papers, Indian Affairs, vol. 1:844).
American State Papers
Foreign Relations, vol. 4. Washington.
Indian Affairs, vol. 1. Washington.
Military Affairs, various volumes, various dates. Washington.
1924. A History of Florida, 2 vols. Deland.
1838. "Letter from the Secretary of War. Seminole Indians--Prisoners
of War," Executive Documents, no. 327, 25th Congress, 2nd
Session, vol. 8. Washington.
1928. "Letters of Samuel Forry, Surgeon, U.S.A., 1837-1838, Part III,"
Florida Historical Quarterly,,vol. 7, pp. 88-105. Gainesville.
Lieutenant of the Left Wing, A
1836. Sketches of the Seminole War. Charleston.
Long, Ellen Call
1882. Florida Breezes or Florida New and Old. Jacksonville.
1818. "Message on the Seminole War," House Documents, no. 14,
15th Congress, 2nd Session, vol. 1. Washington.
Various volumes, various dates. Baltimore.
Officer of the Medical Staff, U. S. Army, An
1840. "A Sketch of the Indian Tribes Known under the Appelation of
Muskogees," The Monthly Magazine of Religion and Literature,
vol. 1, pp. 137-147. Gettysburg, Pa.
Poinsett, Joel R.
1838. "Letter from the Secretary of War. Negroes, etc., captured from
Indians in Florida," Executive Documents, no. 225, 25th Con-
gress, 3rd Session, vol 5. Washington.
Read, William A.
1934. "Florida Place Names of Indian Origin and Seminole Personal
Names," Louisiana State University Studies, no. 11. Baton
Simmons, William Hayne
1822. Notices of East Florida. Charleston.
Sprague, John T.
1848. The Origin, Progress, and Conclusion of the Florida War.
Swanton, John R.
1922. "Early History of the Creek Indians and Their Neighbors,"
Bureau of American Ethnology, Bulletin no. 73, Washington.
Williams, John Lee
1827. A View of West Florida. Philadelphia.
1934. "A Topographical Memoir on East and West Florida, with
Itineraries of General Jackson's Army, 1818," Florida Historical
Quarterly, vol. 13, pp. 16-50, 82-104, 129-164. Tallahassee.
University of Oregon
S. T. WALKER, AN EARLY FLORIDA ARCHAEOLOGIST
Ripley P. Bullen
Archaeologists are always interested in time, the actual or relative date of
artifacts or cultures with which they are dealing. Such obsession is a necessary
prerequisite to any proper understanding of early periods as both the natural
and the cultural environment change with time. Not only is it of interest to date
specimens but it is only by a study of the relative dating of various artifacts
that the student can get an idea of how cultures grow and change.
Only after relative or absolute chronologies are established for various
areas can interconnections between these areas, the transmission of cultural
forms from group to group, the direction of such influences, and the growth of
cultures be properly studied. When the dynamics of archaeological situations
are understood, the data acquired new life and the problems become those of
human beings receiving or inventing new and stimulating tools or artistic ideas,
improving their techniques and way of life, or being oppressed and swamped by
stronger and more vigorous cultures.
For these reasons archaeologists are concerned with time. The establish-
ment of chronologies, especially that of the relative dates of artifacts in dif-
ferent areas, must be done carefully. Obviously, in any case of culture trans-
mission, the earlier must be the giver and the later the receiver. Unless care
is used in dating, influences will seem to go in a direction which is diametri-
cally opposite from that which is actually the case.
Early efforts to establish chronologies for Florida are, therefore, of interest,
both historically and as a check on more recent work. On the east coast the
earliest stratigraphic work is that done by N. C. Nelson at Oak Hill in 1918
(Nelson, 1918). A much earlier start may be claimed for the west coast. I refer,
of course, to the work of S. T. Walker of Clearwater who, in 1883, published
suggestions of chronological changes in styles of pottery based on his investi-
gation of a shell midden at Cedar Keys (Walker, 1883a).
Little is known about Walker except that he lived in Clearwater and is
referred to in Howell's Florida Bird Life as a Colonel (Howell, 1932:19).
Walker's archaeological publications in the Annual Reports of the Smithson-
ian Institution for 1879, 1881, and 1883 cover surveys of the Tampa Bay region
from Shaws Point to Aripeka, the northwest coast from Point Washington to
Pensacola, and around Cedar Keys (Walker, 1880a, 1880b, 1883a, 1885). Walker's
work was considerably ahead of that of his period. His descriptions were ac-
curate, his conclusions sound and his attention always upon problems. Unfor-
tunately, many of his suggestions and observations were not followed by later
workers, particularly Clarence B. Moore.
Walker was not only interested in archaeology but also sent the Smithsonian
Institution collections of reptiles, fishes, mammals, and birds (Baird, 1883:101-
5). His usual occupation is not evident from his reports but that he was active-
ly engaged is indicated by his writing that the archaeological work was done
"at intervals snatched from business engagements." In November, 1879, he was
connected with the United States Fish Commission at Shaws Point (Walker,
1880b:416). This connection appears to have been a temporary one although
Walker published in 1883 a paper entitled "Fish Mortality in the Gulf of Mexico"
Walker's 1883 archaeological paper covers his investigation of a twelve and
a half foot section of a shell midden at Cedar Keys (Walker, 1883a). In this
paper he started to develop a chronology in terms of changes in styles of pottery
with depths in the midden.
This midden he divided into five zones. The lowest three feet was charac-
terized by thick, heavy pottery, tempered with coarse sand or small pebbles.
vessels were of large size, rudely fashioned, and lacking any attempts at dec-
oration. Rims were plain and not thickened or reinforced to increase their
After an overlapping transitional zone, the next four feet supplied thinner
pottery without coarse tempering. Rims were turned outward, slightly thickened,
and sometimes scalloped like the "edge of a pie." Decoration was made by
punctations and incised straight lines. Various types of check stamping was
also present. At the very end of this period, Walker mentions that curved line
decoration was introduced and a few vessels were supplied with eared rims.
The next higher two feet consisted of soil or dirt which contained very
little pottery. Above this was three more feet of midden deposit with new dec-
orative patterns and new vessel forms.
In this zone all the larger vessels were furnished with ears and some were
made with handles. The finer containers were decorated with zigzag lines,
curves, dots, and, in rare cases, figures of men and of animals. Near the sur-
face was some painted pottery.
The topmost six inches consisted of soil which had formed since abandon-
ment of the midden. Walker pointed out that at least fifty years had elapsed
since Indians lived on the midden. He thus estimated two hundred years for the
two foot thick lower soil zone and a thousand years for the total life of the
While abandonment of the midden probably occurred nearer a hundred and
fifty than fifty years before Walker's visit to the site, his estimate of total
age is close to that generally accepted by archaeologists before revisions
were recently made, partly as a result of new data from Carbon-14 analysis
(compare Goggin, 1948:58 with Goggin, 1950:10).
Walker's generalized stratigraphic section compares reasonably well with
that developed by Willey and Woodbury (1942) for the northwest Florida coast.
The lowest, undecorated zone was not defined by them in northwest Florida
but is similar to the lowest four or five feet of the fifteen foot Terra Ceia mid-
den on the southern side of Tampa Bay, which also contained undecorated pot-
tery (Bullen, 1951).
Walker's second period would appear to equate with Willey and Woodbury's
Deptford with, at the end, Santa Rosa-Swift Creek features. We have no data
for the intermediate two foot thick soil zone. The next zone above the old soil
is clearly Weeden Island developing into or being modified into Fort Walton or
Safety Harbor. The painted pottery, mentioned by Walker as close to the surface,
may have been Mission Red Filmed of the Spanish mission period.
While Walker's work was not definitive in terms of pottery types, the general
sequence he presented is similar to that recognized to-day. Work of a chronolog-
ical nature, equal to Walker's, was not done on the west coast of Florida until
Willey and Woodbury started excavating by stratigraphic methods in 1940, about
sixty years later.
Baird, Spencer F.
1883. "Report of the Secretary," Annual Report of the Smithsonian
Institution for 1881, pp. 1-186. Washington.
Bullen, Ripley P.
1951. "TheTerra Ceia Site, Manatee County, Florida," Publications,
Florida Anthropological Society, no. 3, Gainesville.
Goggin, John M.
1948. "A Revised Temporal Chart of Florida Archeology," The Florida
Anthropologist, vol. 1, pp. 57-60. n.p.
1950. "Florida Archeology -- 1950," The Florida Anthropologist,
vol. 3, pp. 9-20. Gainesville.
Howell, Arthur H.
1932. Florida Bird Life. New York.
Nelson, N. C.
1918. "Chronology in Florida," American Museum of Natural History,
Anthropological Papers, vol. 22, pp. 75-103. New York.
Walker, S. T.
1880a. "Preliminary Explorations among the Indian Mounds in Southern
Florida," Annual Report of the Smithsonian Institution for 1879,
pp. 392-412. Washington.
1880b. "Report on the Shell Heaps of Tampa Bay, Florida," Annual
Report of the Smithsonian Institution for 1879, pp. 413-422.
1883a. "The Aborigines of Florida," Annual Report of the Smithsonian
Institution for 1881, pp. 677-680. Washington.
1883b. "Fish Mortality in the Gulf of Mexico," Proceedings, United
States National Museum, vol. 6, pp. 105-9. Washington.
1885. "Mounds and Shell Heaps on the West Coast of Florida," Annual
Report of the Smithsonian Institution for 1883, pp. 854-868.
Willey, Gordon R. and R. B. Woodbury
1942. "A Chronological Outline for the Northwest Florida Coast,"
American Antiquity, vol. 7, pp. 232-254. Menasha.
Florida Park Service
ARCHEOLOGICAL NOTES ON LOWER FISHEATING CREEK
John M. Goggin
Fisheating Creek rises in the highland area of South Central Florida, wan-
ders southward off the Pleistocene escarpment and then, swinging to the east,
flows into Lake Okeechobee. Carrying water most of the time, the stream is a
concentration point for a rich assortment of animal life attractive to primitive
man. Moreover, it is an easy artery of travel by canoe between the lake and in-
For much of its course the stream passes through reasonably elevated
country, but shortly before entering the lake it opens into a wide marsh, the
northwestern tip of the Everglades. At the point where dry banks widen away
there is on the south a high dry hammock. Now an ideal spot for hunters or
fishermen, it apparently appealed to many early peoples for abundant evidence
may be found here of early Indian occupation as well as remains of an early
U. S. Army post.
PREVIOUS ARCHEOLOGICAL RESEARCH
The establishment here of a small block house, Fort Center, during the
Seminole wars focused the attention of various military men on the site. The
high sand mound (G1 12) attracted most discussion, its elevation in this flat
country being grossly overestimated (Ives, 1856:37). This mound came to the
notice of later archeological writers among them Charles Kenworthy who also
greatly exaggerated the height, reporting it was forty feet (Kenworthy, 1875:307)
and later as fifty feet (Kenworthy, 1883:633). Le Baron (1884:778-779) also notes
the site in his general survey. However, after this time the sites drops into ob-
livion for many years.
An unusual find in the early 1930's again brought attention to the area. A
local resident found near the major mound a large wooden figure of a bird stand-
ing on a post (Anon., 1933:292). This specimen is now in the Florida State Muse-
um in an excellent state of preservation.
The writer first visited the Fort Center area in 1944, making a small sherd
collection for the Yale Peabody Museum. In the summer of 1947 another visit
1This paper is a contribution from the research program of the Department of Sociology
and Anthropology, University of Florida, made possible by a grant from the Wenner-
Gren Foundation for Anthropological Research Inc. (formerly the Viking Fund, Inc.).
Appreciation is expressed to Charles P. Lykes and the Lykes Brothers Inc. and to
Leila Platt for permission to dig on their respective properties. Grateful acknowledg-
ment must be made to Robert Spangenberg for the preparation of the illustrations.
was made with John W. Griffin of the Florida Park Service and material collect-
ed at that time is in both the Yale Peabody Museum and the Florida Park Serv-
ice collection. Later several brief visits were made in 1949, each giving more
insight into the extent of the site and enabling plans to be made for test excava-
tions and a more thorough survey of the whole area.
The sites excavated do not appear to have been mentioned in the literature
unless the Platt site is the large "shell mound" near Lakeport mentioned by
Fewkes (1928:2). It does not appear to be from his reference, but no other com-
parable site is reported in the vicinity by local inhabitants.
THE FORT CENTER AREA
Centering around the site of old Fort Center are a number of distinct arche-
ological remains. These are of varied nature, perhaps of different ages, and
spread over an area less than a mile square. For that reason each has been
given a single number or sub-number, but we will use the term Fort Center site
as =a inclusive term for the whole group. The greatest number of these remains
are located in or adjacent to the oak hammock which borders the south bank of
Fisheating Creek, where the stream flows into the prairie bordering Lake
The hammock, and palmetto-pine complex behind it, are high and dry areas
during rainy seasons and the sites are above water at all normal times. These
sites include a series of small middens along the river edge, all considered as
one site but individually designated by letters (G1 13A, 13B, 13C, 13D). The
possible site of Fort Center (G1 23) itself is indicated by a restricted concen-
tration of early 19th century chinaware. The large sand mound and its associat-
ed platform (G1 12) are the most widely recognized features of the site. However,
even as important but not so casually recognized, is the large circular canal
(G1 22) several thousand feet long. Several small mounds have unknown func-
tions (Gl 18, 19, 20). Another mound or ridge on the floor plain (G1 21) has re-
fuse on the surface.
The prairie to the east is often covered with several inches or more of
water and the sites lying in it appear as small cabbage palm islands at such
a time. These individual sites include two pairs of ridges (Gl 24, Gl 25), two
pairs of ridges with terminal mounds (G1 11, G1 15), a small mound of unknown
nature (GI 16), and a long narrow ridge (G1 17) probably a refuse midden which
may have once lain along the stream when a past meander took it closer. An-
other pair of ridges lies across the flood plain on the opposite side of the creek
About three quarters of a mile to the east, on the prairie, are two more sites
(G1 10 and G1 14). The first is a small mound but the second, called the Platt
Site, is a refuse heap forming a large mound. This is bordered in places by the
For several reasons the Fort Center area seemed to be a desirable place for
stratigraphic testing. The site area as a whole is extensive and suggests long
occupation and the concentration of so many mounds, earthworks, middens, and
a canal has few parallels. The Platt site itself is quite large for the region and
appears to be a counterpart of the Belle Glade midden to the southeast of the
lake. Finally, the lack of any excavations in this part of the Glades area made
such work desirable.
With these factors in mind test excavations were made on a weekend in May,
1950, by several students under the direction of the writer. One group worked at
the Platt site (G1 14)2 while another dug at Fisheating Creek midden (G1 12A).3
In conjunction the writer made a general reconnaissance of the Fort Center area
using airphotos to locate sites. Another student based at our camp spent most of
hit time mapping the Ortona Site, some 10 miles to the Southwest.
The results of the two stratigraphic tests will be considered in turn and then
discussed as a whole. Insomuch as most of the artifact types are well known and
have been often discussed in the literature, they will be only briefly considered.
Certain details of Belle Glades Plain pottery are analyzed in another paper by
Rita Porter (1952).
TIE PLATT SITE
The Platt Site (G1 14) is a large, roughly triangular, sand midden with two
distinct and separate rises separated by a swale. This lower central part is
about 3.5 feet above the prairie while the higher points are about 6 feet in ele-
vation. Fisheating Creek briefly touches the mound on the north side.
At the present time the site is occupied by two small houses, several sheds,
and small gardens belonging to the Platt family, occupants of the site since it
was homesteaded many years ago.5 These Florida pioneers found in the site
the same qualities noted by the Indians. It was a dry island surrounded by the
marsh in high water with easy access via the adjacent creek to the interior or
The choice of our excavation spot was due in part to existing conditions.
Abundant surface sherds (1000's) indicated that adequate samples were present
2Donald Kokomoor, Glen Fuguitt, and William Reeves.
3Morton McDonald, William Kyle, and Jay Blowers.
5Grateful appreciation is expressed to Miss Leila Platt for permission to excavate and
for many courtesies offered to us in our work. A small but interesting collection of
material, picked up over the years was given by Miss Platt to the University.
everywhere so a high point was chosen to give the greatest temporal range. The
easternmost elevation was then planted in a Cassava garden so a place was se-
lected on the westernmost ridge.
The test pit was a five by five foot square and excavation was carried out
in arbitrary six-inch levels with all soil being screened, giving a nearly com-
plete collection from the pit. The test was dug to a depth of eighty-four inches
and there stopped when sterile soil was reached. The last cultural material and
refuse was in the 66"-72" level.
Throughout most of the pit the soil was fairly uniform organic darkened sand
with sherds, occasional other artifacts, and a scattering of shell and bone. This
was consistent to a depth of fifty inches where a ten-inch layer of much blacker
sand rested on a sterile white sand. However, pockets of dark material ran down
into the white sand below sixty-six inches. The pockets when examined in the
wall profiles were very suggestive of postholes. One was seen to extend from the
dark material (above fifty inches) into the black soil, others extended from dark
material through the black material into the white, while others ran from the
black downward into the white. No conclusive interpretation can be made of
them. The black soil is suggestive of an occupation layer but few sherds are
in it; they actually begin to occur in quantity just above it. It may represent
the original land surface.
Small fragments of animal bones were found throughout. They have not been
fully identified but casual examination indicates their general nature. There was
apparently little change in the faunal picture. Turtle bones were most numerous
and present in every level. Deer were common and found in all but two levels,
with bird and fish bones almost as numerous. Alligator bones were present as
well as a number of randomly scattered unworked shark teeth. Various rodent and
other small mammal bones have not been identified but may include squirrel and
Shells were also found in all parts of the test. Most common were freshwater
mussels, with a few freshwater snails. Planorbis sp., Paludina sp., and Am-
pullaria sp. Marine shells occurred in surprising frequency -- Busycon sp. being
present in six levels, Venus sp. in four levels and Strombus gigas in one level.
These large conchs may be fragments of artifacts or raw material for manufacture.
Artifactual remains occurred in varying quantities. Table 2 shows the distri-
bution of 1892 potsherds. The ceramic picture is quite simple; the bulk of the
material being Belle Glade Plain, with some Glades Plain and a scattering of
other forms. Examination of the table shows little change in pottery types through-
out the occupation. The only significant variations are the appearance of St.
Johns Check Stamped in the 18"-24" level and its continuation to the surface
and the percentile decrease in the frequency of Belle Glade Plain in the lowest
level. The presence of abberant Glades Plain forms between 54 and 66 inches
TABLE 2. NUMERICAL FREQUENCY OF POTSHERDS, PLATT SITE TEST
Belle Glades Glades St. Johns St. Johns Belle Unclassi-
G Pde Plain Plain, Plain Check Glade fied
Level Plain aberrant Stamped Incised smooth Total
is perhaps of ultimate significance but as yet is unexplainable.
A further breakdown of the ceramic picture by the analysis of Belle Glade
Plain rim forms (Porter, 1952) indicates some differences which correlate with
other ceramic variations. If material from the surface level (0"-6") is ignored
because of possible movement through modern plowing a pattern may be seen in
the stratigraphic distribution of types 2, 3, 4, 5, 11, and 13 (Porter, 1952, Table
4). These all occur in or below the 18"-24" level.
The general ceramic picture for the site can be expanded somewhat with a
consideration of surface material.6 It includes more of the rare Belle Glade In-
cised (4 specimens) including an incised and punctated sherd perhaps from a
boat-shaped vessel. Other types include Okeechobee Plain; an unclassified,
punctated, gritty ware sherd; and a Spanish olive jar sherd. An intriguing object
of chalky ware is apparently the foot of a human effigy figure.
In addition to potsherds seventeen fragmentary or whole artifacts and six
worked bone fragments came from the excavation. These include (proveniences
in parenthesis): 2 sherd hones (0"-6", 18"-24"), 1 small tringular flint point
(6"-12"), 3 possible Busycon pick fragments (6"-12",12"-18",24"-30"), 1
Busycon tool A (48"-52"), an antler adz socket (24"-30"), 5 bone points (18"-
24", 24"-30", 36"-42", 42"-48", 54"-60"), 3 bone pin fragments (18"-24",
36"-42", 48"-54"), 1 possible bone spatula (18"-24"), and 6 worked bone
fragments (6"-12", 12"-18", 18"-24", 24"-30"). These specimens are in too
small quantity to indicate significant temporal differences; none contradict ex-
isting ideas. The triangular point is worthy of special comment inasmuch as the
type is rare in the region. However, its late position here equates with its
known late occurrence in Northern Florida.
Surface collections give a much richer artifact picture. This material in-
cludes a Busycon pick A; Strombus celt, Busycon gouge (?); Busycon celt;
basal notched, triangular, flint point; stemmed, triangular, flint point; bone
point; single and double grooved columella shell pendants; limestone plummet
pendants; bone pin fragment; sherd hone; and both thick and thin grinding slabs.
The lack of distinctive time markers in the pottery from here makes it dif-
ficult to place the sequence in the detailed Glades picture (Goggin, 1947, 1950;
Willey, 1949). However, the sudden appearance of St. Johns Check Stamped can
be equated with the beginning of Glades III. Correlating with this (a one level
overlap) is the disappearance of certain vessel forms of Belle Glade Plain
(types 2, 3, 4, 5, 11, and 13) until the surface level, If the examples from the
0"-6" level are dismissed as being late intrusives because of modern plowing
or other movement of surface material, we can assume that the Belle Glade
Plain rim types in question are pre-Glades III most likely Glades II. Further-
more they do not appear to be too early, since except for one example of type
2 (48"-54") they range no deeper than 42".
The earliest temporal range of the site is unknown; the relative unimportance
of Belle Glade Plain at the expense of Glades Plain forms may correlate with
Glades I, but this is little more than a speculation. On the other end of the time
scale, the presence of a Spanish olive jar sherd suggests historic occupation,
probably in Glades IIIc times.
6This includes a random sample of 794 sherds made previous to the excavation, a se-
lected sample of 44 sherds made during excavation, and 3 sherds from a collection
presented by Leila Platt (all at University of Florida).
FISHEATING CREEK MIDDEN
In the elevated hammock area bordering the south bank of Fisheating Creek
there are a series. of small midden deposits. Scattered refuse is sparsely found
the whole length of high ground but certain rises, with a greater concentration
of refuse, have been distinguished by alphabetical designations after the general
site number (G1 13). The most central of these (Gl 13A), lying between the two
entrances of the circular aboriginal canal, is also the highest midden and was
chosen for a test excavation. At the time of our work the top of this midden was
10 feet above the unusually low creek surface.
The excavation was commenced as a five by ten foot trench in the highest
part of the midden, at right angles to the stream and as close as was practical
to the eroded bank. It was divided into two five-foot squares and these were
carried down separately to a depth of 42 inches. At that point work on section
5' 10' (the one away from the creek) was stopped due to lack of sufficient time
and the excavation was continued into sterile soil at 66 inches in section 0'-5'.
Material was removed in six-inch levels and screened.
The midden deposit was predominately dark organically stained sand.
Scattered throughout were occasional shells, bones, and a few artifacts. At a
depth between 37 and 48 inches there was a layer of consolidated mussel shell
and sand cemented together into a hard whitish breccia. It extended over half
(closest to the stream) of section 0'-5'. Below that the stained sand continued,
gradually becoming lighter below 54 inches, grading from grey into white. At
66 inches the sand became buff in color with many small yellow concretions.
Animal bones were found in some quantity. They have not been precisely
identified, as yet, but an examination indicates a number of forms. Turtle, deer
and fish bones are most numerous, being found in practically every level from
the surface to the bottom. Bird remains are perhaps next in frequency, with many
small mammal remains, some alligator bones, and shark teeth. These last are un-
modified but may have been brought in as future material to be worked. Some
marine shells were present; Macrocallista sp. were most common, but others in-
clude Busycon perverse, Venus sp., Fasciolaria gigantea (?), and Cardium sp.
Freshwater mussels were sporadic throughout, but mainly concentrated in the
In most levels some pebble concretions of sand cemented together by dark
red ochre were present. It is uncertain whether these were formed in situ or
brought in by the previous inhabitants of the site; most probably the former ex-
planation is correct. They are excellent sources of red pigment and were prob-
ably used by the Indians in any case. Other mineral remains include flint chips
from section 0'-5': 1 from 30"-36" level, 9 from 48"-54" level, and 2 from 54"-
Artifacts were found in all levels of the excavation except the deepest. They
include 996 potsherds, 20 other artifacts, and 3 fragments of worked bone. The
distribution of potsherds is shown in Table 3 which combines the material from
both section level by level.7 On examination it can be seen that the great bulk
of the material present is Belle Glade Plain followed in quantity by Glades Plain
with scattering of other forms. Both types are in reasonably similar proportions
TABLE 3. NUMERICAL FREQUENCY OF POTSHERDS,
FISHEATING CREEK MIDDEN TEST
Belle Glades St. Johns St. Johns Dunns St. Johns Unclassi
Glade Plain Plain Check Creek Incised field
Level Plain Stamped Red (?) smooth Total
0"-6" 8 2 10
from 48 inches to 18 inches; from there upwards, Glades Plain is in considera-
bly smaller frequency, although it increases in the uppermost level.
The analysis of the Belle Glade Plaim rim forms (Porter, 1952) shows some
consistency again. Types of the B series, numbers 11, 12, and 13 are the only
forms found in the surface level (Porter, 1952, Table 5). Types 8 and 14 are ab-
It should be remembered that from 42 to 66 inches only one five-foot square was dug.
sent, while the remaining types are found only in or below the 6"-12" level.
St. Johns Plain occurs sporadically throughout much of the test but is in no
significant position except for its earliest occurrence. The late occurrence of
St. Johns Check Stamped (6"-12") is not surprising but the few Dunns Creek
Red sherds cannot be given any diagnostic interpretation, as is also the case
of the unclassified smooth plain gritty ware specimen. The single possible St.
Johns Incised sherd deep in the site is not in an unexpected position.
Surface collections do little to amplify the general picture from the test.
They include both Belle Glade Plain and Glades Plain, as well as Belle Glade
Red, Dunns Creek Red, and a fiber and grit tempered sherd (Florida Park Serv-
Other artifacts from the excavation include: I Busycon tool A (l"-6"), 1
sherd hone (6"-12"), 3 worked sherds (2 from 18"-24", 1 from 30"-36"), 4
perforated shark teeth (1 from 18"-24", 1 from 30"-36", 2 from 36"-42"), 1
notched shark tooth (24"-30"), 2 bone points (24"-30" and 42"-48"), 1 basal
notched flint point (24"-30"), 7 bone pin fragments (1 from 24"-30", 5 from
30"-36", 1 from 36"-42"), and 3 fragments of worked bone (1 from 0"-6", 2
from 24"-30"). A few modern objects in the surface level need no mention.
Nonceramic objects as a whole are in too few numbers to show any significant
Surface material includes only a single weathered green glass fragment.
This probably dates from the early 19th century occupation down stream.
As with the case of the previous site there are not many distinctive markers
which enable this picture to be fitted well into the general Glades picture. Like
at the Platt site, the appearance of St. Johns Check Stamped suggests the be-
ginning of Glades III times, but the sample is so limited that exactly when this
took place is hard to say. The earliest levels in the site are of some interest.
The 48"-54" level lacks both Belle Glade Plain and typical Glades Plain. But
there are present several St. Johns Plain fragments, and a possible St. Johns
Incised sherd. The next deepest level has a single sherd of a ware intermediate
between Glades Plain and Belle Glade Plain. However, there was some question
at the time it was found as to whether it fell in from the side higher up in the
trench. No sherds came from the lowest level. In both of the deepest levels
there were some bone fragments.
In summary we can stratigraphically interpret the test as follows: levels
from 0 to 12 inches are Glades III; levels from 12 to 48 inches certainly date
in part from Glades II, with perhaps the uppermost extending into Glades III
times, while the lower may possibly be equivalent to the late Orange period and
levels 54 to 66 inches may be preceramic. There is a high degree of probability
for the above interpretation with the exception of the placement of the lowest
The pottery found is as a whole rather undistinguished. Most of the forms
are plain wares, adequately made but not artistically attractive. With few ex-
ceptions all are well known types, discussed in detail in other places. There-
fore a mere characterization of their nature will suffice.
Belle Glade Plain. This is the most common pottery of the Fort Center
area and typical of the northern Glades areas. It is a hard, fine textured, well
made ware, with a distinctive dragged surface, often faceted from a smoothing
tool. One sherd with a polished surface came from the 18"-24" level at the Platt
site. The type is described by Willey (1949).
An analysis of vessel and rim forms by Rita Porter (1952) shows some
change in the history of the type. However, further studies in a greater num-
ber of sites will be necessary to demonstrate more clearly the meaning of the
Belle Glade Incised. This is an undescribed type to be described in detail
in a book in preparation (Goggin, n.d.). The paste and surface treatment are
characteristic of Belle Glade Plain, but has in addition simple, crudely incised
or incised and punctated decoration. Five examples were found in the excava-
tion and on the surface of the Platt site. Other examples are illustrated by
Willey (1949, P1. 2, A-B).
One sherd deserves mention. It appears to be part of a small square or rec-
tangular slab or tray. The upper surface is well polished and incised, the lower
is rough but in one corner is a small foot (?) or support.
Belle Glade Red. This undescribed type is a rare form similar to Belle
Glade Plain but with one or both vessel surfaces painted red. A single example
came from the surface of Fisheating Creek Midden.
Glades Plain. This is the second most abundant form of pottery in the area,
but more common to the south. It is a sand-tempered ware, usually quite gritty
to the touch, and made in a variety of vessel shapes and degrees of surface
finish. Detailed descriptions are available in a number of sources; especially
see Willey (1949), and Goggin and Sommer (1949).
Glades Plain, aberrant form. Pottery of this type came from the lower levels
of the test excavation at the Platt site. It differs in having more temper, and is
softer and more friable. In some respects it is like Goodland Plain.
Okeechobee Plain. An uncommon form first described by Willey (1949:29-
30) it was found here in surface collections at the Platt site.
St. Johns Plain. Formerly this type was called Biscayne Plain in the
Glades Area but the two types are now recognized as being similar and this
name most appropriate. It is a soft, chalky ware pottery found at both sites.
Detailed descriptions are in Goggin and Sommer (1949:44-45) and Willey (1949:
St. Johns check Stamped. Similar to the proceeding form, this type is
characterized by a check-stamped surface. It serves as the period marker of
Glades III. Examples are illustrated in Goggin (1950a, Fig. 78, H), Goggin and
Sommer (1949, P1. 3, I), and Willey (1949, P1. 4, A-F).
Dunns Creek Red. Formerly known as Biscayne Red (Willey, 1949:98) this
is a red painted form of St. Johns Plain. It occurs on the surface and in the
30"-36" level of Fisheating Creek Midden.
St. Johns Incised (?). A probable example of this type came from deep in
the midden at Fisheating Creek. -The type is discussed by Ferguson (1951: 26-
Unclassified punctated gritty ware. A small sherd from the surface of the
Platt site may possibly be Weeden Island Punctated.
Unclassified smooth plain gritty ware. Several sherds from both sites do not
fall within the range of local forms. They are hard, dark in color, and well
smoothed. They are suggestive of Weeden Island Plain.
Unclassified fiber and grit tempered ware. A single sherd (Florida Park
Service collection) came from the eroding face of Fisheating Creek Midden. This
is characterized by fiber and grit tempering. While not identical with Orange
Plain, it is undoubtedly related.
Spanish olive jar. A single sherd of this coarse Spainsh utility ware came
from the surface of the Platt Site. It has a green glazed interior and a white
slipped exterior. (See Boyd, Smith, and Griffin, 1951, P1 X, 10 for a nearly
These artifacts will be only briefly noted since most are described in more
detail in available literature. Reference is made in many cases to these sources.
Busycon shell tools. Tools made from whole Busycon perverse shells are
characteristic of the Glades Area (Goggin, 1950b, Fig. 21, A-B; Goggin and
Sommer, 1949, P1. 5, A, C; Willey, 1949, P1. 11, H-J, P1. 15, A). A single whole
specimen of Busycon pick A came from the surface of the Platt site. (for an
example of the type see Goggin, 1950a, Fig. 79, X).
Other tools from the same site, too fragmentary to determine their exact
type, include four picks (surface, 12"-18", and 2 from 24"-30") and undeter-
mined picks or hammers of type A from the 6"-12" and 48"-54" levels. At
Fisheating Creek Midden a Busycon type A tool came from the 0"-6" level.
Shell celts. Celts from the surface of the Platt site include a small ex-
ample (1 7/8 by 3 inches) of the Strombus type and a Busycon form. Another
specimen, from the same site may be a Busycon celt or a gouge.
Perforated shark tooth. Perforated shark teeth are widely found in the
Glades area, being used as knives or set in clubs. Examples from the Fort
Center area are typical, exhibiting no modifications except a neat perforation.
Four came from Fisheating Creek midden in levels 18"-24" (Fig. 9, I), 30"-36"
(Fig. 9, J), and 36"-42" (Fig. 9, H).
Notched Shark tooth. A single shark tooth slightly notched at the base,
presumably for hafting came from Fisheating Creek Midden (24"-30", Fig. 9,
K). These are not uncommon in the Glades area (Goggin, 1951, Fig. 7, C).
Antler adz socket. A fragmentary piece of worked antler appears to be the
blade end of an antler adz socket. With a raised triangular facet at the end it
most closely resembles examples from Key Marco (Willey, 1949, P1 15, B; also
see P1. 8, A-B).
Grinding stones. Fragments of both thin and thick grinding stones (Willey,
1949) came from the surface of the Platt site. The latter is a piece of limestone
5/8 inch thick with one side worn smooth through use. The other specimen, of
hard, sandy, limestone (?), measures 1 5/8 inches thick and is worked on both
flat surfaces and unbroken edge.
Sherd hones. Potsherds are found with deep grooves worn in the surface
which are believed to be the result of wear from the sharpening or polishing of
bone artifacts. Two hones made from Belle Glade Plain sherds came from the
surface of the Platt site and from Fisheating Creek Midden (6"-12"). Two
others, made from St. Johns Plain ware were found in Platt site (0"-6", 18"-
Bone spatula (?). A heavy splinter of deer bone from the Platt site (18"-
24") shows wear on a rounded end. It may have served as a spatulate tool al-
though it is not well finished.
Bone points. The most typical projectile points of the Glades area are made
Fig. 9. Artifacts from the Fort Center Area; A, bone pin, GI 13A; B. long bone
point, G1 14; C, long bone point, G1 13A; D, long bone point, GI 14; E, small bone
point, GI 13A; F, small bone point, GI 14; G, double grooved columella pendant,
G1 14; H, perforated shark tooth, Gl 13A; I, perforated shark tooth, GI 13A; J, per-
forated shark tooth, Gi 13A; K, notched shark tooth, GI 13A; L, limestone pendant, G1
14; M, flint point, G1 14; N, flint point, GI 14; 0, flint point GI 13A; P, limestone
pendant, GI 14. Scale 4/5.
of bone in several styles. Eight examples came from the Fort Center area.
These are bipointed objects usually made of a dense bone, most often deer
Two specimens of the "small bone point" came from the Platt site (24"-
30", Fig. 9, F) and Fisheating Creek Midden (42"-48", Fig. 9). Another broken
specimen from the Platt site (18"-24") is probably of this type but slightly
larger, measuring 2 inches.
The other five specimens fall within the group called "long bone points."
Two are rather small and slender; one from Fisheating Creek Midden (24"-30",
Fig. 9, C) and the other from the Platt site (42"-48", Fig. 9, D). Another
couple from the Platt site (surface; 36"-42") fall within this group but have
one end more appreciably thinned than the other. The last from here is larger
and sturdier (54"-60", Fig. 9, B). Comparative examples may be seen in Willey
(1949, Pls. 7, 16), Goggin and Sommer (1949, P1. 5, C-D), and Goggin (1950a,
Flint points. The presence of five flint projectile points in our collection
is somewhat unusual in the Glades area where native flint is not present. A
small triangular specimen from Fisheating Creek Midden (6"-12", Fig. 9, M)
is the usual late form from Northern Florida, but the first example from the
Glades Area. It appears to date from Glades III times. Two notched base speci-
mens from the surface of the Platt site (Fig. 9, N) and Fisheating Creek Midden
(24"-30"; Fig. 9, 0) are also North Florida types which sporadically occur in
the Glades area (Willey, 1949, P1. 16, I). The Platt site specimen is note-
worthy because of the excessive wear on the point, as if from use as a drill.
Two other fragmentary points came from the surface of the Platt site. One
is the tip of a large triangular specimen while the other is a stemmed triangu-
lar example. It is equilateral in form and measures 1V/ inches long (less the
Stone pendant; plummet form. Two limestone pendants, variations of the
plummet form, come from the surface of the Platt site. One is of usual shape
and workmanship (Fig. 9, L) while the other is very well made (Fig. 9. P).
Shell pendant; columella form. The central columella of the Fasciolaria
gigantea shell is ground down to form pendants. They occur throughout Florida
in several forms. Surface specimens from the Platt site include a single grooved
specimen and a double grooved variety (Fig. 9, G).
Bone pins. Polished bone pins are the most characteristic ornaments of the
Glades Area. One large bipointed specimen (Fig. 9, A) made of a thin split bone
(perhaps a bird femur) came from Fisheating Creek Midden (24"-30"). It meas-
ured more than 102 inches long when complete.
Other bone pins are represented by only points or central shaft pieces.
Five small fragments came from level 30"-36" at Fisheating Creek Midden and
another from the 36"-42" level at the same site. Similar small pieces also
came from the Platt site on the surface, two in the 18"-24", and one each in
the 36"-42", and 48"-54" levels.
Pottery effigy fragment. What appears to be the foot and lower limb of a
human effigy came from the surface of the Platt site. It is well made from hard
Worked sherds. Belle Glade Plain sherds worn on the edges for some un-
determined purpose came from Fisheating Creek Midden (2 from 18"-24"; 1
from 30"-36" levels). They may have been smoothers.
Worked bone. Nine small pieces of bone show worked surfaces but it is
impossible to determine what they were intended to be. They were found at
Fisheating Creek Midden (0"-6", four from 24"-30") and the Platt site (6"-12",
12"-18", 18"-24", 24"-30").
DISCUSSION AND CONCLUSIONS
Our test excavations in two Fort Center Area middens yielded stratigraphic
samples from depths of 5/2 to 6 feet. There was little differentiation in the
ceramic picture throughout the history of either site. Belle Glade Plain and
Glades Plain were the dominant forms but varying somewhat in frequency. In
both sites the most significant note in the ceramic picture was the late appear-
ance of St. Johns Check Stamped pottery which can be correlated with Glades
III times. The presence of a Spanish sherd on the Platt site surface indicates
an occupation as late as Glades IIIc times.
Of equal importance but not so well defined are the indications of extremely
early occupation. The presence of a probable St. Johns Incised sherd in the
lower levels of Fisheating Creek Midden, together with a fiber and grit tempered
sherd found here previously, suggests a possible correlation with the Orange
Period, characterized by such wares, to the north. Then too, there are sugges-
tions of a preceramic horizon at the same site in the lower two levels (54 to 66
inches). No pottery was found in the lowest level and only a single sherd in the
next highest which may have fallen in from above. Evidence is not conclusive
The major occupation at both sites is neither Glades III nor very early.
Therefore, it must fall within what we recognize as Glades I and II times. Much
of it is undoubtedly in the latter period, perhaps also in Glades I.
The two sites excavated are only part of the large complex in the Fort
Center area, and unfortunately our work has not been extensive enough to indi-
cate the relationship between the periods recognized in our stratigraphic tests
and the building and occupation of the various earthworks. Small sherd collec-
tions were made on the surface of some other sites but they are not distinctive.
Belle Glades Plain is present in all, along with Glades Plain. St. Johns Plain
occurs at a single site. None of these forms are diagnostic and none came from
either the large sand mound or the canal.
Comparing our material with that from elsewhere in South Florida, it seems
to fall within the general pattern of the northern Glades area. Here we find, in
general, only plain wares and few of the distinctive incised types characteris-
tic further south.
Parallels between the Platt site and the Belle Glade midden are very strik-
ing. Both are similar in size and nature. The stratigraphic picture is also simi-
lar, although the much more extensive excavation at Belle Glade yielded a
larger sample containing a few incised types. At both sites the bulk of the pot-
tery was of Belle Glade Plain and Glades Plain types. The former was always
the majority type but Glades Plain was in its highest frequency early in the
sites history. Finally St. Johns (Biscayne) Check Stamped appears late in the
Belle Glades history (Willey, 1949) as it does in the Fort Center area.
In summary we can state that stratigraphic tests in the Platt site (Gl 14)
and Fisheating Creek Mlidden (GI 13A) yielded a ceramic sequence remarkably
like that from the Belle Glade Midden (PB40). The upper levels of both sites
can be placed in Glades III times, the Platt site being occupied as late as
The majority of occupation at both sites probably falls within Glades II
times and perhaps also in Glades I. There are suggestions of early occupations
at Fisheating Creek Midden which may equate with the Orange Period (early
Glades I ?) and with preceramic horizons.
1933. "Totem Pole from Florida," Scientific American, vol. 148, pp.
292-3. New York.
Boyd, Mark, Hale G. Smith and John W. Griffin
1951. Here They Once Stood. Gainesville.
Ferguson, Vera Masius
1951. "Chronology at South Indian Field, Florida," Yale University
Publications in Anthropology, no. 45, New Haven.
Fewkes, Jesse W.
1928. "Aboriginal Wooden Objects from Southern Florida," Smithson-
ian Miscellaneous Collections, vol. 80, no. 9. "Washington.
Goggin, John M.
1947. "A Preliminary Definition of Archaeological Areas and Periods
in Florida," American Antiquity, vol. 13, pp. 114-127. Menasha.
1950a. "Stratigraphic Tests in the Everglades National Park,"American
Antiquity, vol. 15, pp. 228-46. Menasha.
1950b. "Cultural Occupation at Goodland Point, Florida," The Florida
Anthropologist, vol. 2, pp. 65-91. Gainesville.
1951. "The Snapper Creek Site," The Florida Anthropologist, vol. 3,
pp. 50-64. Gainesville.
n.d. Archeology of the Glades Area, Southern Florida. Manuscript in
Goggin, John M. And Frank H. Sommer III
1949. "Excavations on Upper Matecumbe Key, Florida," Yale Univer-
sity Publications in Anthropology, no. 41, New Haven.
Ives, J. C. (compiler)
1856. Memoir to Accompany a Military Map of the Peninsula of Florida.
Kenworthy, Charles J.
1875. "Indian Mounds and Canals," in Camp Life in Florida, edited*
by Charles Hallock, pp. 313-21. New York.
1883. "Ancient Canals in Florida," Annual Report of the Smithsonian
Institution for 1881, pp. 631-35. Washington.
Le Baron, J. Francis
1884. "Prehistoric Remains in Florida," Annual Report of the Smith-
sonian Institution for 1882, pp. 771-790. Washington.
Porter, Rita Krestenson
1952. "An Analysis of Belle Glades Plain Rim Sherds from Two Fish-
eating Creek Sites," The Florida Anthropologist, vol. 4, pp. 67-
Willey, Gordon R.
1949. "Excavations in Southeast Florida," Yale University Publica-
tions in Anthropology, no. 42. New Haven.
University of Florida
AN ANALYSIS OF BELLE GLADE PLAIN RIM SHERDS FROM
TWO FISHEATING CREEK SITES
Rita Krestensen Porter
Differentiation in pottery types is one of the chief methods used today by
the archeologist to determine the time sequence of a site but its use must be
limited, generally, to those sites exhibiting a diversity of decorated and plain
ceramics. Occasionally a site is excavated which yields little or no decorated
pottery although its plain pottery probably represents many years of occupation
and perhaps several culture periods. To determine a time sequence on the basis
of this plain ware requires another approach to the problem of pottery variability.
With this in mind a study was made of two related sites in the Glades area of
southern Florida having sequences of predominantly plain wares.
These two sites are found on the banks of Fisheating Creek on high, dry
hammock land surrounded by marshes at the northwestern tip of the Everglades,
a short distance from the point where the creek empties into Lake Okeechobee.
Test excavations of a limited nature were conducted at the sites on a week-end
in May, 1950. Since a comprehensive discussion of the sites and the excavation
procedure has already been written (Goggin, 1952), they shall be described only
briefly and with emphasis on information relevant to this study.
The more profitable site, in terms of material collected, is the Platt Site
(GI 14), a large sand midden with abundant and widely distributed surface mate-
rial. The only test pit for this site, a five by five foot square, yielded a great
quantity of specimens when it was excavated in arbitrary six inch levels, each
carefully screened. Sterile soil was reached around eighty-four inches, while
cultural material was not found below the 66"-72" level.
The other site under consideration is the Fisheating Creek Midden (GI 13A)
which lies on the southern bank of the creek in an elevated hammock. Among a
series of small midden mounds there is one which rises above the rest and is
somewhat in the middle of the area. This prominence reaches an elevation of
around ten feet above the creek surface. It was therefore selected as the point
for a test excavation. A five by ten foot trench divided into two five foot squares
was excavated to a depth of 42" when lack of time demanded cessation of work
on the half that was farthest from the creek. Digging of the other section con-
tinued to 66" before sterile soil was encountered.
This is a contribution from the research program of the Department of Sociology and
Anthropology, University of Florida, aided by a grant from the Wenner-Gren Foundation
for Anthropological Research, Inc. This paper was originally prepared as a student
research project under the supervision of Dr. John M. Goggin, to whom sincere thanks
are due for access to his manuscript (Goggin, 1952) on the area.
The ceramic material from both sites shows a preponderance of potsherds
of the type known as Belle Glade Plain. This pottery is intermediate between
the St. Johns (Biscayne) and Glades series in temper and hardness. It is of
medium to coarse quality, coiled, compact, sand-tempered, and while the paste
is hard and brittle the sherds approach the chalky appearance of St. Johns
(Biscayne) ware (Goggin and Sommer, 1949:46). There is a complete lack of
decoration, although the exterior is often roughly faceted through the use of a
smoothing implement. The inner and outer surfaces range in color from gray to
pinkish cream with a light gray to black core and the hardness ranges from 2.5
to 3.5 (Moh's scale). Thickness of the body is generally between 4 and 8 mm.
while the rims are sometimes thicker. It is characteristic of most of the sites
in the Okeechobee area and is reported to extend from Upper Matecumbe Key
north to the Indian River Area (Rouse, 1951:222; Ferguson, 1951:30-2).
A major problem of this pottery is its inadequacy as a time marker because
it probably occurs in all three of the major Glades periods in great abundance.
Usually it is more common in Glades II and III than in the earlier period but the
time boundaries of a site cannot be set on this basis. In order to possibly utilize
Belle Glade Plain for chronological purposes, therefore, it was determined to
disregard surface modifications and to concentrate on the form.
Pottery is well adapted to the diversities of artistic expression and experi-
mentation which do not always involve some design or texture but may instead
result in a variety of vessel forms. It seems reasonable that, just as a certain
type of surface decoration was in vogue in an area at a given time, a distinct
rim treatment might also have been characteristic of that period. If these fluc-
tuations in rim style occur regularly and correlate with known decoration changes
or other time markers from period to period, one might assume the rim type is of
some value in delimiting the time span of a site. This method might not be as
satisfactory as others due to its greater dependence on personal interpretation
and classification with a resulting susceptibility to differences of opinion, but
in cases where there seems to be no other alternative it might prove invaluable.
The primary steps in the procedure of investigation of the two sites described
previously were to separate the rim sherds from the Belle Glade Plain body
sherds, clean them in a ten-percent hydrochloric acid solution to remove calcium
carbonate deposits, and to mark each sherd with its site number and level. The
profile of each rim from the surface collection at the Platt Site was then drawn
on separate cards. Since the surface collection contained the greater number of
sherds (196) and since it was possible that some of the styles persisted from
lower levels to the surface, it was assumed that the categories derived from
shuffling and dividing the cards into groups would comprise a significant sample
of types occurring in the site. After discerning the differentiating characteristics
of each group and standardizing two series and fourteen types, the remaining
rims were classified accordingly.
It was found that these types persisted through the lower levels, and the
classification was discriminating enough to form the basis for the arrangement
of a stratigraphic pattern of types. In several instances a few similar rims of a
lower level did not belong to an existing category. A new group was then formed
if their similarity was great and their number warranted it. The same type clas-
sifications were effectively used for the rim sherds from Fisheating Creek Mid-
den because of the close relationship of the two sites.
In the following discussion lip is defined as the margin of a vessel where
the inner and outer surfaces are joined through an arc or plane or modification
of the two. The term rim refers to the lip and the uppermost part of the wall of a
Two major series are evident in the rim types: Series A is composed of the
rims having basically a flat lip which may be horizontal, beveled inward or out-
ward, or vertical. Series B contains the rims that exhibit rounded lips of varying
degrees of curvature. In each series, types are defined according to the width
of walls, degree of curvature of the rim, thickening of rim or bulges below the
lip, slant of lip, width of the lip in comparison to the wall width, and with some
consideration of the probable form of the total vessel (according to the classi-
fication of basic Florida vessel forms given by Willey, (1949b:496-506). In all
there are fifteen types grouped as follows:
Series A (Flat lip).
Type 1. Lips are thicker than the side walls with profiles generally slant-
ing outward. 141 examples. (Fig. 10, A-D).
Type 2. Lips are the same width as the walls, which are very thin, between
3 and 4 mm. 26 examples. (Fig. 10, E-F).
Type 3. The lips are slightly thicker than, or equal to, the width of the
walls. The profiles indicate they are probably rims of shallow bowls
or dishes. 11 examples. (Fig. 10, G-H).
Type 4. Lips are slightly thicker or equal in width to the side walls with
the upper edge of the exterior wall shaped so as to form an angle
with the lip. In profile it gives the appearance of a pointed lip
while the rim slants outward. 13 examples. (Fig. 10, I-J).
Type 5. Lips are thinner than or sometimes equal to the walls with a slight
bulge in thickness just before the exterior wall joins the lip. 20
examples. (Fig. 10, K-L).
Type 6. The lips are of the same width as the side walls and profiles slant
outward or curve inward very slightly.102 examples. (Fig. 10,M-N).
c D f E F G H
\\ % \
1 K L M N
Fig. 10. Belle Glade Plain rim profiles. (Interiors to the right.) A-D, type 1; E-F,
type 2; G-H, type 3; I-J, type 4; K-L, type 5; M-N, type 6; O-P, type 7; Q-R, type 8;
S-T, type 10; U-W, type 11; X-Y, type 12; Z-BB, type 13; CC, type 14; DD, type 15.
Type 7. The width of the lip is usually equal to or a little greater than the
walls. A unique characteristic is a lip flattened vertically and a
profile that curves inward suggesting the vessel was a flattened
globular bowl. 21 examples. (Fig. 10, O-P).
Type 8. Two distinct planes on the interior of the rim combine with the
horizontal plane of the lip to form an unusual angled projection on
the profile, which may slant inward or outward. 5 examples. (Fig.
Type 9. Aberrant rims. 12 examples.
Series B (Rounded lip).
Type 10. The lip is curved and pointed, sometimes with a slight bulge in the
wall thickness below the lip. 12 examples. (Fig. 10, ST).
Type 11. The lip is an arc, the chord of which is greater than the width of
the side walls. The profile shows the rim curved inward and implies
the vessel was a flattened globular bowl. 22 examples. (Fig. 10,
Type 12. A bulge of the interior of the wall occurs below and extending up to
the lip, which sometimes may curve out over the exterior of the
wall. 19 examples. (Fig. 10, X-Y).
Type 13. The chord of the lip's arc is usually the same width as the side
walls and the arc is generally very slight. The walls are vertical
or curve inward to a small degree. 32 examples. (Fig. 10, Z-BB).
Type 14. The lip is part of a curve that forms almost a complete circle to-
ward the interior of the vessel. 10 examples. (Fig. 10, CC).
Type 15. The thickness of the wall and lip are undifferentiated and thin, with
the lip slightly pointed. 5 examples. (Fig. 10, DD).
The frequency distributions (see tables 4 and 5) show that rim types 1 and
6 are the more abundant types, together comprising 63.1% of all rims studied
from Fisheating Creek Midden and 52.1% of all rims from the Platt Site. At the
Platt Site type 1 has a percentage of 35.3 and type 6 accounts for 16.8%, while
the majority is reversed at the second site with percentages of 10.5 and 52.6
respectively. The occurrence of all series B types is similar at the two sites,
representing 17.1% of all Belle Glade Plain rims at the Fisheating Creek Mid-
den and 23.7% at the Platt Site. Rims of the B type, however, were found main-
ly on the surface with a few in two of the lower levels of each site. In the case
of the Platt Site they were found from 18 to 30 inches deep and at the other site
they were in the 24"-30" and 36"-42" levels.
After reviewing references to rim and lip forms in other publications, it is
apparent that the dominant form at other related sites in Southeast Florida, Up-
per Matecumbe Key, and at some sites in the Indian River area is also a flat,
untapered lip that is wedge-shaped or of the width of the vessel walls and
slightly curved in profile (Rouse, 1951:222; Willey, 1949a:25-6; Goggin and
TABLE 4. PERCENTILE DISTRIBUTION OF BELLE GLADE PLAIN RIM
TYPES AT THE PLATT SITE
Surface 49 7.3 2.4 4 8.1 11.4 6.5 4 6.5 122 62.2
0"-6" 57.9 5.2 5.2 5.2 5.2 21 19 79
6"-12" 55 29 16 27 100
12"-18" 60 40 15 83.3
18"-24" 36 11.5 5.5 8 5.5 19 8 5.5 36 95
24"-30" 47 5 5 2.6 26.3 10 2.6 38 95
30"-36" 14.3 28.5 14.3 14.3 28.5 14 87.5
36"-42" 22 8 8 54 8 13 100
42"-48" 33.3 66.6 3 100
48"-54" 100 1 100
54"-60" 100 1 100
Series B Total Per cent
Level Type: B Rim of total
10 11 12 13 14 15 sherds rims
Surface 13.5 25.6 17.5 35 8.1 74 37.8
0"-6" 20 20 20 20 20 5 21
12"-18" 100 3 16.6
18"-24" 100 2 5
24"-30" 100 1 5
30"-36" 100 2 2
Sommer, 1949:46). Specimens from South Indian Field and the Manatee region
are reported to have predominantly round or flattened-round lips of Series B
(Ferguson, 1951:30-2; Willey, 1949b:365). Incurving, straight, and outslanting
rims from large, open-mouthed sub-globular bodies are represented at all of the
When considering the distribution of rim types, several patterns suggestive
of real meaning become apparent. It is difficult, nevertheless, to draw any def-
inite conclusions of the temporal change in rim types due to the limited quantity
of sherds studied. Of the 451 sherds from both sites, 196 were a surface col-
lection at the Platt Site, while the remaining 255 were distributed through some
TABLE 5. PERCENTILE DISTRIBUTION OF BELLE GLADE PLAIN RIM
TYPES AT THE FISHEATING CREEK MIDDEN
Series A Total Per cent
Level Type: A rim of total
1 2 3 4 5 6 7 8 9 sherds rims
0"-6" 0 0
6"-12" 33.3 22.2 11.1 33.3 9 90
12"-18" 20 20 40 20 5 100
18"-24" 25 25 50 4 100
24"-30" 18 18 9 54.5 11 78.6
30"-36" 11.7 11.7 76.6 17 100
36"-42" 7.1 85.7 7.1 14 93.3
42"-48" 100 2 100
Series B Total Per cent
Level Type: .B rim of total
10 11 12 13 114 15 sherds rims
48 to 60 inches at both sites.
Due to the fact that the land on which the Platt Site is located has been
under cultivation for a long period of time, numerous plowings are thought to
have contaminated the upper six inches of soil with material from earlier oc-
cupations. This material was probably removed from its lower, peripheral areas
and spread over the higher areas of more recent deposition. With this considera-
tion in mind, therefore, the results from the top level of that site have virtually
According to Goggin (1952) the presence of St. Johns Check Stamped pottery
at the Platt Site in the 24" level and at the Fisheating Creek Midden at 12"
probably equates closely with the beginning of Glades III period at the sites, al-
though it could quite possibly have begun several levels earlier. The relative
abundance of material in the 18"-30" levels of the former site and in the 24"-
36" levels of the latter, followed by a decline in specimens might represent the
Glades II period and the transition to Glades III at both sites, although this
would result in an earlier dating of Fisheating Creek Midden.
Some significance should be assigned also to the grouping of Series B types
from 18"-36" at the Platt Site and from 24"-42" at the other site, with a com-
plete absence of these types in the other levels until the top level is reached,
since their occurrence correlates with the levels of heavier concentration of
sherds and greater variety of types.
Considering the concentration of sherds an indication of the zenith of the
Glades II period, the supposition is that this culture reached its greatest de-
velopment through the 18"-36" levels at the Platt Site and 24"-42" at the other
site. This period was followed by a diminution of the characteristic traits and the
gradual transition to the later Glades III period, which probably attained some
importance only in the last level of occupation at both sites.
Although this interpretation of the sites should be considered nothing more
than a useful suggestion of the temporal sequence, it may demonstrate the im-
portance of rim changes and the possibility of further refinement of the method
until more definite results can be obtained by its use.
A study was made of the Belle Glade Plain rim sherds from two sites in the
Glades area of Southern Florida. The Platt site and the Fisheating Creek Midden
site were chosen because of their plain pottery type sequences which made the
setting of chronologies extremely difficult. Working under the hypothesis that
although the pottery surfaces may be unmodified through several periods, the rim
forms may fluctuate through the periods to denote temporal changes and thus be
of aid in delimiting the time boundaries, two major series of types were defined.
Series A consists of rims having flat lips and is composed of nine types, while
series B contains rims with curved lips and has six types.
After classifying and counting the rim sherds from both sites percentage dis-
tribution charts were compiled. Upon comparison several correlated patterns sug-
gested that the major part of both sites was of the Glades II horizon while the
later occupation, as evidenced in the top levels, was probably Glades Ill. The
transition period between the two main culture periods was marked by a decline
in material and types. Thus it seems that rim types are of some value in deter-
mining the time sequences of the two sites, although the results cannot be con-
Ferguson, Vera Masius
1951. "Chronology at South Indian Field, Florida," Yale University
Publications in Anthropology, no. 45. New Haven.
Griffin, John W. and Hale G. Smith
1948. "The Goodnow Mound, Highlands County, Florida," Florida
Park Service, Contributions to the Archeology of Florida, no. 1.
Goggin, John M.
1952. "Archeological Notes on Lower Fisheating Creek," The Florida
Anthropologist, vol. 4, pp. 50-66. Gainesville.
Goggin, John M. and Frank H. Sommer III
1949. "Excavations on Upper Matecumbe Key, Florida," Yale Univer-
sity Publications in Anthropology, no. 41. New Haven.
"A Survey of Indian River Archeology, Florida," Yale Univer-
sity Publications in Anthropology, no. 44. New Haven.
Willey, Gordon R.
1949a. "Excavations in Southeast Florida," Yale University Publica-
tions in Anthropology, no. 42. New Haven.
1949b. "Archeology of the Florida Gulf Coast," Smithsonian Miscel-
laneous Collections, vol. 113. Washington.
University of Florida
Excavations at Kolomoki, Season II 1950 by William fl. Sears, University of
Georgia Series in Anthropology, No. 3, 35 pp., 14 pls., 3 figs., map. Athens,
This report covers partial excavation of Mound E at the Kolomoki Mounds
State Park in southeastern Georgia. The unexcavated portion of the mound and
some of the features uncovered have been left as part of the permanent exhibits
at the park.
Mound E was found to consist of a nearly sterile, basket-loaded, secondary
mound covering a primary mound built of sandy loam and rocks. The primary
mound capped a centrally located burial pit on the floor of which were the re-
mains of a cremation, 307 shell beads, and two copper ear ornaments. Near the
top of the pit were two primary extended burials and two isolated skulls while a
third extended burial was nearby and another isolated skull, with a copper-
covered wooden ornament, was in the top of the primary mound. Shell beads ac-
companied the extended burials. A small cache of sherds plus another copper
ear ornament was located a few feet east of the primary mound under a small ad-
dition to it. A second and rather large pottery cache, including many sherds and
over fifty restorable vessels, was found further to the east, near the edge of the
Sears is of the opinion Mound E. while not built in one day, was completed
over a very short period of time. This is reasonable, as the copper-covered orna-
ment from the first cache is similar to those with the cremation and there were no
significant differences between pottery types in the two caches.
The most important result of the excavation of Mound E is the pottery cache
in which various types of pottery were in direct association with each other.
This pottery includes Weeden Island Plain, Weeden Island Incised (not zoned but
of a variant Sears refers to as "free" incised), vessels with cut out portions,
KolomokiComplicated Stamped, and Mercier Red on Buff, a new type. All ves-
sels exhibited "kill" holes some of which were made before firing.
These type names hardly give a real idea of the ceramics. Examination of the
illustrations reveal typical Weeden Island Plain vessels, Weeden Island decora-
tive details such as cut work on Middle Mississippian vessel forms, Kolomoki
Stamped containers with Weeden Island type rims, Mercier Red on Bluff with
Weeden Island cut work on Middle Mississipian shapes with either Middle Mis-
sissippi or Weeden Island type rims, Mercier Red on Buff on otherwise typical
Weeden Island Plain vessels, incised containers with cut work, and two vessels
with incision which stylistically resembles Ft. Walton Incised
Handles are absent but there can be no question but that this collection
represents late Weeden Island funeral ceramics modified by Middle Mississippian
influences. In terms of Florida archaeology, it represents the logical first step
in the ceramic shift from Weeden Island to Fort Walton. Sears points out the
Middle Mississippian traits, similar pottery associations found by Moore in
some burial mounds near the Florida gulf coast, and that it may be necessary
to rearrange our concept of what is early and what is late in Weeden Island
It may also be desirable to review our ideas regarding Moore's material
from Crystal River in the light of this new data. This reviewer professes to see
some similarities between certain Crystal River Negative Painted vessels and
some of the Mercier Red on Bluff while some of the Crystal River Incised may
be the same as Sears "free" variant of Weeden Island Incised. Extremely close
similarity between ear ornaments from Crystal River and those from Mound E
at Kolomoki has been noted by Sears.
Excavations at Kolomoki, Season II 1950 has produced provocative data.
We look forward the report on Season Ill.
Ripley P. Bullen
Florida Park Service
A Survey of Indian River Archeology, Florida by Irving Rouse. 296 pp., 8 pls.,
15 figs. Yale University Publications in Anthropology, no. 44, New Haven, 1951.
Chronology at South Indian Field, Florida by Vera Masius Ferguson. 62 pp., 4
pls., 10 figs. Yale University Publications in Anthropology, no. 45, New Haven,
1951. (The two are bound in one cover, price $4.00)
These significant papers represent a major contribution to Florida archeology
based on research carried out in the 1940's under Yale University's Caribbean
program. They are a necessity in any minimum library of Florida archeology.
The monograph by Mrs. Ferguson reports on two stratigraphic tests made by
the Yale-University of Michigan expedition at South Indian Field, a large mid-
den at the headwaters of the St. Johns River, west of Malabar, Florida. This
site, property of a society member, Mr. A. T. Anderson, is unique in many re-
spects but only the details of the particular project are considered in this paper.
Ably presenting the results of her excavations, Mrs. Ferguson illustrates
a ceramic sequence of chalky and gritty ware pottery overlying fiber-tempered
pottery (Orange Period). The pottery and other artifacts found are described in
detail. Of special interest are the numerous fragments of soapstone vessels,
unique because of the absence of this material in the state.
An appendix "Animal Remains from South Indian Field" by Margaret Van
Winkle Iouck is the first detailed stratigraphic study of faunal remains in
Florida and well illustrates the value of such an approach. The extensive
occurrence of broken dog bones in the early levels suggests their use as food
during the Orange Period. Other anomalies in fauna are noted such as the pres-
ence of sturdier deer and muskrat remains than are now typical.
This study of South Indian Field will be of value for years to come. It is
the first, and to date the only, detailed report on a site having a major occupa-
tion in the Orange Period, and it represents a major contribution in terms of its
presentation and details.
Dr. Rouse's book grew out of original plans for a limited site survey in the
vicinity of South Indian Field. However, the unusual features of South Indian
Field and the large amount of unreported work by A. T. Anderson captured his
interest. When the site surveys were later begun it was found that the original
scope was too limited. The area finally covered by Rouse, essentially the In-
dian River and bordering regions, represents a good geographic unit and a cul-
ture area as well defined as any in Florida.
The scope and detail of Dr. Rouse's work are so great that it is possible
to touch only on some of the high points. The first 71 pages comprise introduc-
tory material including data on the environment, geology, ethnology, and history.
This section alone justifies the work. Special consideration given to the ge-
ology has resulted in a well integrated summary of many conflicting reports on
Pleistocene and Recent events. The chapter on ethnology, primarily concerned
with the Ais Indians, represents the most thorough ethnographic summary for
any of the early Florida tribes, while his summary of history is more detailed
for the Spanish period than that for any comparable area in Florida.
South Indian Field is the subject of the next section of 35 pages. Work
originally done by A. T. Anderson is presented and his early recognition of the
essential stratigraphic features is acknowledged. Details of the site are fas-
cinating in their dynamic picture. Originally built on a dry prairie the first in-
habitants (Orange Period) dug deep wells for water. Later a rising water table
(correlated with rising sea level) inundated the prairie leaving the midden site
an island. The race between man's rising midden and nature's water table must
have been a close one, because by modern times (before drainage) only the
very top of the midden was above high water.
The third and most extensive section, the site survey, comprises 112 pages
and discusses in greater or less detail 114 sites out of a total number of 202 in
the area. These represent many new sites never reported before, but equally im-
portant is the analysis of sites excavated or reported by earlier workers, es-
pecially Clarence B. Moore.
The discussion of the "Early Man" sites at Vero and Melbourne represents
a sincere attempt to evaluate all of the geological evidence and to interpret
the archeological remains. The result is a great clarification of this long vexing
problem. All in all, the survey section sets up the archeological validity of the
culture area studied and makes possible an understanding of South Indian Field's
Brief characterizations of artifact types comprise section four. Most of the
significant as well as the unique forms are illustrated in the figures and plates.
The final section comprises detailed conclusions concerning the geology
and archeology and a discussion of their implications. They include summaries
for each archeological period.
To round out this study are two appendices, "The Derrotero of Alvaro
Mexia, 1605" by Charles D. Higgs and "Human Skeletal Remains from South
Indian Field" by Charles W. Goff. Mr. Higgs' study has been awaited for a
long time by scholars of this area as well as those interested northwards on
the Florida coast. His translation and annotation make this early account of
the Indians especially useful. With our scant knowledge of physical anthropology
in the state, Dr. Goff's appendix is a small but welcome contribution.
John M. Goggin
CONTRIBUTORS TO THIS ISSUE
Kenneth W. Porter. Dr. Porter, now visiting Professor in American History
at the University of Oregon, is known to all students of Seminole history. His
many articles on this subject have been based on extensive research in the
National Archieves as well as among the Seminoles in Oklahoma and Mexico.
Ripley P. Bullen. Mr. Bullen, Assistant Archeologist, Florida Park Service,
is a familiar contributor to our journal. The present paper is an outgrowth of
archeological research on the Florida West Coast.
John M. Goggin. Dr. Goggin, is Associate Professor of Anthropology at the
University of Florida.
Rita K. Porter. Mrs. Porter is a student in Anthropology at the University