An investigation of the Darby and Hornsby Springs sites, Alachua County, Florida (FGS: Special publication 7)

Material Information

An investigation of the Darby and Hornsby Springs sites, Alachua County, Florida (FGS: Special publication 7)
Dolan, Edward M
Allen, Glenn T. ( joint author )
Place of Publication:
Florida Geological Survey
Publication Date:
Copyright Date:
Physical Description:
iv, 124 p. : illus., maps, diagrs. ; 23 cm.


Subjects / Keywords:
Indians of North America -- Florida ( lcsh )
Antiquities -- Alachua County (Fla.) ( lcsh )
Hornsby Spring ( local )
Darby Spring ( local )
Projectiles ( jstor )
Marl ( jstor )
Scrapers ( jstor )
non-fiction ( marcgt )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The author dedicated the work to the public domain by waiving all of his or her rights to the work worldwide under copyright law and all related or neighboring legal rights he or she had in the work, to the extent allowable by law.
Resource Identifier:
AEG1796 ( NOTIS )
000865021 ( AlephBibNum )
01106378 ( OCLC )
a 61009752 ( LCCN )


This item has the following downloads:

Full Text

Florida Agricultural
Experiment Station Library
G.nnes'alh. Foirrda


Ernest Mitts, Director

Robert 0. Vernon, Director




Edward M. Dolan
Glenn T. Allen, Jr.

Florida Geological Survey





Introduction........................................ 1
Geography ....................................... 3
Geology and physiography .......................... 5
Pleistocene vertebrate fauna ....................... 12
Clim ate ....................................... .. 17
The background: Archaeological................... 19
The paleo-Indian tradition ................... 19
The Archaic tradition....................... 20
The Gulf and St. Johns tradition periods ...... 21
The Alachua tradition....................... 21
Discussion of the excavations ..................... 22
Analysis of the materials .................... 27
Conclusions...................................... 39
I. P ottery ............................... .. 41
Typology .......................... 43
II. Lithic materials ........................ 55
Typology .......................... 58
Bibliography ..................................... 112


1 Map of survey area ....................... 2
2 Hornsby Springs map ..................... 4
3 Darby Spring map ................. ....... 6
4 Diagrammatic cross section of a solution pipe
filled by shell marl and noncalcareous quartz
sand containing Pleistocene vertebrates and
worked lithic materials ................... 13
5 Archaeological areas of Florida. .......... 18
6 Areas and periods of culture in Florida .... 23
7 Representative profiles ................... 26
8 Provenience of Darby Spring pottery types 30
9 Provenience of Hornsby Springs pottery types 34
10 Provenience of projectile point types ...... 59



I. Hornsby Springs, Pit # 550 ............... 37
II. Pottery ............................... 49
III. Pottery ............................... 51
IV. Pottery ............................... 53
V. Projectile points ........................ 89
VI. Projectile points ........................ 91
VII. Projectile points........................ 93
VIII. Knives ................................ .. 95
IX Knives ................................ .. 97
X. Knives and scraper...................... 99
XI. Scrapers and knives ..................... 101
XII. Knives and bone fragment.................. 103
XIII. Knives, drill and scrapers ............... 105
XIV. Choppers............................... 107
XV. Choppers............................... 109
XVI. Hammerstones.......................... 111

1 Distribution of lithic artifacts other than
projectile points......................... 73



Edward M. Dolan and Glenn T. Allen, Jr.


In 1951 and 1952 the Florida Geological Survey
sponsored and financed the excavation of two archaeological
sites, Hornsby and Darby springs, in northwestern Alachua
County, Florida. Over this two-year period, William E.
Edwards and James Clarence Simpson worked intermittently
on the project with occasional aid from Allen E. Jordan,
Harry H. Simpson, and Survey personnel. The excavations
yielded fiber-tempered sherds and large amounts of lithic
materials and artifacts.

The quantity and nature of these articles and the
circumstances under which they were found were such that
it was felt that a detailed study, analysis, and report were
warranted. Unfortunately, Mr. Simpson died in the latter
part of March, 1952. The materials were stored by the
Florida Geological Survey until this study began.

It is the purpose of this report to record these archae-
ological materials; to describe those which appear to have
typological significance, and to compare them with other
Florida and eastern United States archaeological materials.
Where possible, areal and temporal relationships will be

Except for brief field trips to resurvey the sites and
to make stratigraphic tests, the field work was done by the
Edwards-Simpson party. Their field notes are relied upon
for descriptions of methodology and conditions under which
the various materials were found. However, the method of
presenting these materials, the comparisons made, and the
conclusions drawn are strictly the authors'.






1i' ,' 1. Map of survey area,

'" :* .



The two major sites discussed in this study Hornsby
and Darby springs lie in extreme northwestern Alachua
County, Florida. Hornsby Springs, located in the SE-NE-
sec. 27, T. 7 S. R. 17 E. is northwest of the town of
High Springs. It is exactly one mile to the east of U.S.
Highway 41- 441 along a county road immediately south of,
and roughly parallel to, the Santa Fe River. The property,
at the time of the excavations, was relatively wild and
wooded. Today it is a cleared park-like area and the site of
a youth camp conducted by the Seventh DayAdventist Church.

The springs emerge from a large underground cavern
at a depth of about 85 feet below the water surface. The
spring run flows between rather low, swampy banks in a
generally westerly to northerly direction for about one mile,
at which place it disappears into a sink. It probably re-
emerges in the Santa Fe River which lies less than one-half
mile north of this sink. In an effort to divert water to Darby
Spring a canal has been dug from the sink to Darby Spring.

Darby Spring is located in the SWINW- of the same
section. This spring is on the south bank of the Santa Fe
and less than one-fourth mile east of the U. S. Highway 41 -
441 bridge crossing the river.

The surrounding country is generally rolling sand hills
with a gradual slope toward the river. Though there are,
at present, many cleared fields in the area, the native flora
is still widely prevalent. In lower areas, especially along the
streams, the thick, jungle-like forests of mixed pine, oak,
gum, Carolina ash, and cypress are found. Many varieties
of wild berries are present; poke weed, persimmons, sabal
or cabbage palm, and probably coontie (Zamia) along with
other plants were employed as foodstuffs by the aborigines.

In addition to these two major sites, the Edwards-
Simpson survey undertook excavations at five other sites in
the surrounding area. These are describedas the Itchtucknee
site, just east of Itchtucknee Springs in Columbia County;
the Butler site, at the forks of the Suwannee and Santa Fe
rivers and located, in this survey, in Gilchrist County; the


Hornsby Il r10-42""" ,
H n TEST SQUAREt 0 -4 nu-A

,-u 0O-700 0 l ---
v -r -

H8HO-82 HO-10 1


0 5o 0 FEET


Figure 2. Hornsby Springs map.


Marchant and Archer sites, along U. S. Highway 41 north of
the town of Archer; and the Meander site on the north bank
of the Hornsby Springs run and about one-fourth mile below
the springs. Materials from these sites do not enter the
discussion inthis report but they're usedin the formulation
of lithic types and for comparative purposes.

Previous to the development of agriculture, game was
abundant in the area. Deer, turkey, bear, and otter are no
longer common in the vicinity but, no doubt, were plentiful
in the past. Raccoon, opossum, turtles and tortoise, bull-
frogs, rattlesnake, and many varieties of edible fish and
snails still occur.


The Florida Peninsula can be divided into three trans-
peninsular zones, the boundaries oriented about perpendic-
ularly to the length of the peninsula. Of these zones,
northern, central and southern, the northern zone extends
southward from the Georgia state line to a line passing ap-
proximately through St. Augustine, Palatka, Hawthorn and
Gainesville, and lying just south of High Springs.

This zone is distinguished by continuouslyhigh ground
that forms a broad upland which extends across Alachua
County eastward to an elongated lowland, here named the
Eastern Valley. Along the northern tier of counties the zone
extends westward continuously into the Western Highland of
Panhandle Florida.

For the most part, the ground surface of the northern
zone lies above the piezometric surface of the artesian water
of Florida and it is characterized by features of dry highland,
or dead zone karst, such as first generation dry sinks,
abandoned springheads, dry stream courses, dry beds of
broad shallow lakes that are now prairies, streams that
rise above their beds and run into sink caverns to appear

1By Robert 0. Vernon, State Geologist, with some remarks on the physiog-
graphy by W. A. White, Professor, North Carolina University.



DA -- __ R". ~ -----------t



_- 0 zx


I ~


FigatQ 3. P b'i-f SSpriiig map.

---Pt7 < *"

oavb- s*w


again lower in the valley. Several large artesian springs
are present along the valleys, including Darby and Hornsby,
but the water in these obviously fill basins and channels to
elevations lower than it once did.

Sellards (1912, p. 34-35) recognized, but didn't name,
three well defined physiographic areas in Alachua County.
These are:

(1) The plateau-like region north of Gaines-
ville, including most of northeastern Alachua
County. This area is underlain by sediments
of Miocene through Pleistocene ages, but the
presence of deeplyburied limestone is evidenc-
ed by steep-walled sinks that break the fairly
level ground surface.

(2) The middle Florida Hammock Belt cros-
sing the center and southeastern part of the
county. Numerous large shallow lakes or
"prairies" such as Paynes Prairie and Alachua
Lake markthis topography. Eocene limestones
with thin covers of Miocene and younger sedi-
ments compose the subsurface.

(3) The hard rock phosphate belt covering the
western part of the county along the Suwannee
and Santa Fe rivers. The land is well drained
and marked by numerous sinks. Eocene and
Oligocene limestones crop out extensively in
the area.

Darby and Hornsby springs are in Sellard's thirdtopo-
graphic division. The land surface is sandy and rolling with
gentle inclinations toward the stream valleys. The well
graded quartz sand soil captures most of the precipitation
and carries the water to the subsurface. Except for slightly
lower elevations along the large streams, the elevations
range between 75 and 100 feet above sea level. The lime-
stone of the Crystal River formation of the Ocala group lies
closetothe surface andis exposedin sinks andalong stream


The top of the limestone has been extensively eroded
and a Karrenfeld of pinnacles, pits and solution pipes or
"natural wells" is generally masked by loose quartz sand,
occasional clay beds, and phosphorite with which the sand
is interbedded.

The Crystal River formation is a very porous and
permeable limestone that is soluble in the slightly acid rain-
water and the water that enters the rock from the numerous
swamps of the area. This limestone is a major part of a
thick section of rock that is filled by ground water under
artesian pressure. Water entering these rocks locally and
in recharge areas such as that in Clay and Union counties
flows freelythrough the limestones and discharges along the
streams and low places inAlachua and other counties. Darby
and Hornsby springs are excellent examples of this artesian

The effects of solution upon the limestone is evident
in numerous sinkholes that are present throughout the High
Springs area. They vary in widthfrom afewfeet to as much
as a hundred feet. The depth may approach 100 feet. The
deeper sinks pass below the piezometric head and the sink
exposes the groundwater as alake or, if one side of the basin
is low enough, the sink may discharge as a spring. The
springheads of Darby Spring and Hornsby Springs have not
been surveyed accurately, but the orifice of Hornsby is about
85 feet deep.

Vertical cavities that cross the bedding planes in the
limestone are called natural wells, solution pipes and "pot-
holes, and these are present wherever the limestone is
exposed. These vary in width from a few inches to several
feet and likewise in depth upto extremes of 40-50 feet. Gen-
erally, however, the broader and deeper cavities represent
several cycles of solution.

The common and well defined solution pipe is fairly
smooth sided, less than 4 feet in width, with the depth being
quite variable. The most extreme dimensions known are
present in the wall of a sink in Levy County where a pipe was


observed that measured 4 inches in diameter and 9 feet
deep the width remaining remarkably constant.

Suggestions of origin of solution pipes range from the
conventional solutional concentration along joint intersections
and along taproot penetrations to the upward movement of
artesian water. Probably all methods are involved, and
individual pipes may form as a result of one or a combination
of methods. Vernon (1951, p. 43) stated:

"Since artesian water is under pressure it
tends to expand upward along joints and more
soluble rockbecause of greater porositytoward
the ground surface. The localization of this
movement vertically may explain the formation
of solution pipes and 'natural wells. '"

Several meandering cavities and irregularities were
encountered inthe limestone of the digs at Darby and Hornsby
springs, and 17 exceptionally well developed solution pipes
or "natural wells" were excavated at the Hornsby Springs
site (pl. I). These solution pipes are vertical and from 8 to
40 inches in diameter. The depth of penetration was not
determined but they appeared to narrow rapidly at about 2
feet. The pipes were lined by a weathered, pasty, white to
gray sandy shell marl, that merged upward along the pipe
walls with a brown sandy organic, pasty, fresh-water shell
marl, containing numerous well preserved specimens of the
gastropod Giniobasis floridensis. The center of the pipe is
brown, fine to medium quartz sand held in a noncalcareous,
organic clay matrix. This core of brown sand contained
remains of Mammut americanum (Kerr), a mastodon, artic-
ulated remains of opossum, turtle, and remnants of Equus
sp. other smaller vertebrates and worked lithic material.
The relationship of the various sediments, fossils and arti-
facts is illustrated in figures 4, 5 and 6, and plate I.

Descriptions of the occurrences of worked stone,
Pleistocene fossils and sediments in the solution pipes at
Hornsby Springs were prepared by Mr. Clarence Simpson,
and descriptions of the more important occurrences are
reproduced below.


HO 550 Pothole 30 inches in diameter:
(1) Undetermined amount of top material re-
moved by bulldozer, possibly between 1
and 2 feet of sand.
(2) Filling neck of pothole to a depth of 12
inches. Undisturbed fresh-water shell
marl, containing few flint spalls and one
fragment of enamel from tooth of masto-
don. (Sample for C-14 determination. )
(3) Below 1 foot (an additional thickness of 19
inches) black mucky soil containing masto-
don remains and flakes of chert, one rough
artifact. (see plate I. )

HO 551 Pothole 41 feet east of HO 550, 2 feet
in diameter:
(1) 0-12 inches filled with fresh water marl,
slightly hard.
(2) Muck at 12 inches, fragment of possum?
jaw and turtle shell.
(3) 14 inches. Small fragment of rodent's jaw.
(4) 18 inches. Chert scraper in east side of
(5) 21-24 inches. Whole possum skeleton.
Note: "ATALTL" point found in lump of marl
thrown out by shovel from between 0-6

HO 554 Circular pothole 20 inches in dia-
meter. Pothole lined with fresh water
(1) 10-14 inches. Two mastodon teeth badly
crushed by slumping in west edge, partly
imbedded in marl lining of pothole and
partly in black muck filling center of tube.
(2) Betweenthetwo mastodonteethat 1 5 inches
a horse tooth and small chert spall. Six
inches to east of northernmost M. tooth
were two more chert flakes approximately
1 inch across.
(3) Black muck bottomed onfresh-water marl
at 18 inches.


Other important finds were recorded in HO 512 (a
scraper at 48 inches); HO 542 (a triangular blade at 36
inches); HO 521 (a Suwannee point lying on rock at 18-24
inches); HO 600 (unfinished blade at 45 inches, 2 inches
above limestone). In a total of 17 solution pipes excavated,
15 gave evidence of the association of worked lithic materials
and Pleistocene animals.

A sample of the fresh water shell marl (Bed 4) that
covered the solution pipe at location HO 550 was submitted
to the Exploration and Production Research Division of the
Shell Development Company and a carbon-14 date of 9, 880
270 years was determined. 2 This dating is significant
because it establishes the latest date before whichthe worked
lithic material and the vertebrate remains could have been
associated. The brown noncalcareous sandy muck probably
represents a period of time during which the solution pipe
was exposed to the air; the shell marls were formed in
fresh water.

The geologic history can be briefly summarized:
Following the deposition and lithifaction of the limestone of
the Crystal River formation, the limestone was filled by
fresh, slightly acid water, that dissolved the limestone and
enlarged the pores in the rock. Since the geologic history
and stratigraphic sequence is generally known for Florida,
it can be surmised that the Crystal River formation was
filled by artesian water, which, discharging upward in the
High Springs area, dissolved tubes and other openings toward
the ground surface.

Hornsby and Darby springs once filled larger spring-
heads as evidenced by the fresh water marl deposits lying
generally 3-5 feet above the present water levels. In these
larger springheads an organic calcareous paste filled by fresh
water mollusk shells accumulated in the quiet recesses of
the pool head. This sediment would be similar to that which
is accumulating in Lake Panasoffkee, a pasty colloid held in

2Letter from H. A. Bernard, dated March 21, 1960.


partial suspension above the lake bed to depths of several

The drop of artesian water levels that occurred in as-
sociation with decreased sea level during a glacial maximum
would reverse the flow of water through the very porous and
permeable rock and as the water levels declined, the organic
shell paste would be drawn against the rock much the same
as a cake in a filter press thus forming Bed 2 of figure 4.

There was an insufficient quantity of paste to form a
consolidated shell marl that would fill all the surface irreg-
ularities, and many of the solution pipes were left open in
the center, although rimmed by fresh water shell rrarl.
Into these depressions quartz sand from adjacent highlands
and fine grained humus from swamps was washed. Small
animals, remains of larger ones, and any worked lithic
material fell or were carried into these depressions and
preserved by the sediments. This is Bed 3 of figure 4.

During a subsequent recessional of glacial develop-
ment, some ten thousand years ago the ground-water levels
were again elevated and the springheads approached their
former areal spread. In these pools the fresh water shell
marl of Bed 4 of figure 4 accumulated. The surficial quartz
sand generally present above Bed 4 was washed over the
marl from the sand hills that border the springs.


Any discussion of the sequence of events from Pleisto-
cene to historic times finds the fauna playing an important
role. Not only were many of the animals important to the
ancient hunter as a source of food, but tothe modern anthro-
pologist proof of the antiquity of many may hinge upon the
presence of fossilized Pleistocene mammal remains in as-
sociation with archaeological finds. However, it must be
noted, as Neill (1957) pointed out, that in many parts of
Florida, mineralization may proceed at a very rapid rate
and as signing an early dating to remains simply because they
are completely petrified is not justified.


MastWodon *ooth wi/h
scr-aper in matrix


Diagrammatic cross section of a solution pipe
filled by shell marl and noncalcareous quartz sand
containing Pleistocene vertebrates and worked
lithic materials.

Figure 4.


Nevertheless, few problems are of more widespread
concern to the American archaeologist than the faunal turn-
over marking the zoological end of the Pleistocene. Simpson,
in The Extinct Land Mammals of Florida (1929), p. 242-246)
emphasized the extraordinary richness of animal life in
Florida at that time:

"Most, perhaps all, of the recent mammals or
their immediate ancestors were already pres-
ent, but there was a host of other and stranger
animals besides: lions, saber-tooth tigers,
camels, horses, mammoths, mastodons, ground
sloths, giant armadillos, tapirs, dire wolves,

"Even among the less spectacular animals there
were many that no longer inhabit Florida, or
that have entirely vanished fromthe face of the
earth. .

"Flesh-eaters were not lacking toprey on this
abundant life. In addition to the black bear,
there was a short-faced bear (Arctodus flori-
danus) allied to the strange spectacled bear of
South America. There was a dire-wolf (Canis
ayersi), larger than the recent wolf, and a
smaller coyote-like dog (Canis riviveronis)
which is also extinct. Cougars and still larger
lion-like cats (Felis veronis) were widespread.
There was also a saber-tooth tiger (Smilodon
floridanus), with great stabbing canine teeth.

"Ground sloths . were varied, from the
gigantic Megatherium to the relatively small
Megalonyx. An armadillo, like that still living
in Texas but over twice as large, was also
common (Tatu bellus), as wellas a giant arma-
dillo (Chlamytherium septentrionale) with
movable bands of bony armor like the living
form, and glyptodonts (Boreostracon flori-
danus), which resembled the armadillos and
were related to them but were encased in im-
movable armor like the shell of a turtle, from


which their head, legs, and tail emerged.

"The horses, of which there were three sorts
of somewhat different size, all belonged to the
modern one-toed group, Equus. There were
also several types of tapirs, similar to those
still living in South America. Deer were abun-
dant, not only the common white-tail but at
least two extinct types of which, unfortunately,
little is known. Peccaries . were numerous
and varied there appear to have been some
five or six different species. Bisons of a sort
no longer living were common, and at least two
extinct forms of camels added a strange touch
to the landscape.

"Finally, mammoths and mastodons were so
abundant that their teeth are the most commonly
found mammalian remains in the state. The
mastodon was the American true mastodon,
Mastodon americanus, more advanced thanthe
earlier Pliocene mastodons, with tusks in the
upper jaw only. The Imperial Mammoth,
Archidiskodon imperator, most common in
earlier Pleistocene time, arrived fromNorth-
ern India, Southwest Europe, and even South
Africa. The most abundant elephantine species
is the Columbian Mammoth, (Archidiskodon
columbi), which is also abundant in the South
Carolina phosphatebeds and on the plateaus of
Mexico. Thetrue wooly mammoth of the north
never reached Florida. . .

"Even this brief review of the ancient life of
Florida would be incomplete without some men-
tion of the important role of the state of current
investigations regarding the antiquity of man
(that is, of the Indians or their predecessors)
in North America. Did man enter North Amer -
ica for the first time after the final retreat of
the great continental glaciers or was he here
before that time? Was man here associated
with some of the great extinct animals, as he


was in Europe? If so, does this indicate the
great antiquity of man in this continent or the
recent extinction of these mammals? Con-
stantly increasing evidence for the eventual
solution of these interesting problems is being
accumulated, and some of the most important
items of this evidence are being found inFlor-
ida. At several places, especially Vero and
Melbourne, human bones and the products of
human hands have been found in apparent as-
sociation with the extinct animals found above.
If further discoveries confirmthese and if they
can eventually survive the severe scientific ex-
amination and criticism to which they are now
properly being subjected, it will appear that
man has been in Florida for some tens of
thousands of years and that the first arrivals
in this region disputed the ground with the mam-
moths, mastodons, and other great beasts of the
glacial epoch.

"The final chapter in the history of the animal
life of Florida, that of transition from Pleis-
tocene to recent times, is a disastrous one, as
it has been almost everywhere. The present
fauna of the state is onlythe poor and colorless
remnant of that which it once supported. Half,
or perhaps even two-thirds, of the Pleistocene
animals are now extinct, and those of their
companions which survive are not only rela-
tively few in numbers but also generally the
smaller and less striking forms. . It is not
possible to assign definite causetothis deci-
mation, but if present conjectures as to the
antiquity of manhere provetobe correct it will
seem quite probable that the destruction of
animal life by man, still going on, started with
his victory over some of the Pleistocene mam-
mals a victory for which one must now feel
some regret. "



Alachua County, in common with the rest of Florida,
enjoys an unusually pleasant and uniform climate. However,
situated as it is, in an Austroriparian weather zone, winters
are somewhat cooler than Florida's more southerly sub-
tropical to tropical areas. There are about 10 days per
year, onthe average, that temperatures of freezing or below
are reached though the average minimum for January, the
coldest month, is 47 degrees.

Rainfall averages 54 inches per year with the greater
part occurring in midsummer, fortunately when its cooling
influence is most needed. Florida, called "the thunderstorm
capital" of the northern hemisphere, is unique in that it is
almost the only well watered area on the earth's surface
between 24 degrees and 31 degrees north latitude. There
are approximately 90 days every year in which thunder -
storms occur (Baum, 1957).

While south Florida rainfall is truly tropical in char-
acter with a summer rainy season, northern Florida exhibits
a complex rainfall type with late winter and an early spring
secondary maximum. Even in this northern area, however,
the tendency toward the August-September maximum becomes
more and more marked on going eastward along the Gulf

Knowledge of rainfall and climatic changes is important
to archaeological research. This is especially true in an
area bordering on the true tropics, as Florida does. A
slight temperature fluctuation, prolonged over several dec-
ades, could mean the difference between tropical and
temperate flora and fauna. Bartram (1958, p. 74) reported
Royal Palms along the St. Johns River, a region too cold at
present for these plants. Variations in rainfall could well
mean the presence or absence of various shellfishes, de-
pending upon the critical level of salinity in the coastal
waters. It is a known fact, observable at present, that some
shellfish, which were formerly abundant in certain areas,
are either very scarce or are no longer found at all inthe
same places. Streams carrying silt from prolonged periods


I Northwest Gulf Coast

2 Central Gulf Coast

3 Manatee Region

4 Glades Area

5 Kissimmee Region

6 Indian River Region

7 Northern St. Johns Region

8 Central Florida Realon



Figure 5. Archaeological areas of Florida.

(After Goggin, 1949. )

0 25 3- a5 lo l.,.,


of heavy rainfall would certainly influence the growth, if not
the presence, of most of the Mollusca.

Goggin (1948, p. 232) concluded that "the verynature of
the environment on the border between tropical and temperate
zones, plus a long low coastline must have resulted in many
changes since the Pleistocene. "


Since all of the materials to be presented in this report
come from central Florida, a summary of the archaeological
traditions and periods of this area is appropriate. The
archaeologists' central Florida region is a dark spot in the
Florida archaeological picture. Perhaps, as a buffer be-
tween distinctive east and west Florida culture zones it is,
itself, not culturally distinct. However, work has been done
which sheds some little light on this area.

Following an Archaic tradition which cannot, at
present, be given areal definition, Goggin (1949) indicated
Gulf, Glades, and St. Johns as the three traditions which
dominate and characterize Florida archaeology. In central
Florida a short influence of the Gulf tradition follows the
Archaic and is, in turn, followed by a brief St. Johns tradition
influence. Finally, this gives way to a more or less dis-
tinctive Alachua tradition coeval with Weeden Island II of the
Gulf area and St. Johns IIa in the east.

The Paleo-Indian Tradition

The Paleo-Indian tradition is that of a nomadic or
semi-nomadic group existing on hunting, fishing, and gather-
ing of wild foods. The occurrence of Plainview-like and
fluted points, locally called Suwannee points, near the banks
of streams is an observable phenomenon. These finds are
centered in the central Florida region. While the temporal
positionhere is notwell defined theseforms are early else-
where and, on atypological basis, are given an earlydating.
Griffen placed the beginning of the Florida Paleo-Indian
tradition about 8, 000 B. C. (Griffen, 1952, fig. 205). Bullen


(1958) placed the date at 9,000 B. C. A carbon-14 dating
developed during this study placed the age at 9, 880 270

The Archaic Tradition

The Archaic tradition is that of a semi-sedentary group
obtaining their food by hunting, fishing, and gathering.q The
tradition shows an earlier preceramic horizon later followed
by a fiber-tempered ceramic horizon.

Ceremonial life appears to have been relatively simple;
no burial cult seems to have existed; indeed, the scattering
of human bones in these sites suggests that little interest
was felt for the remains. Wyman pointed out that canni-
balism may have been practised (Wyman, 1875).

"The level of technical skill is fair, the few
artifacts from the preceramic level of this
pattern are not noteworthy, but include much
bone worked intoprojectile points, awls, pins,
and miscellaneous worked fragments, as well
as antler projectile points. (Goggin, 1949. )

The influence of this pattern, in its latter phase, is
expressed by the presence of fiber-tempered sherds. This
latter phase, the Orange period, is present in full develop-
ment in the northern St. Johns and Indian River areas. In
1953 Bullen uncovered a fiber-tempered site along the
Chattahoochee River in northwest Florida. On the basis of
radio-carbon dating he concluded that 2, 000 B. C. is a rea-
sonable date to use for the beginning of the Orange period
(Bullen, 1956). Though no radio-carbondates have yet been
offered for the early Archaic in Florida it would seem from
the above evidence and C14 dates from other eastern sites
that this tradition had its beginnings at least 5, 000 years

3This use of the term Archaic follows that of Goqqin (1949).


The Gulf and St. Johns Tradition Periods

Pre-Cades Pond Period

The pre-Cades Pond period equates with the west
Florida Deptford and Santa Rosa-Swift Creek periods and
east Florida's St. Johns Ia (Goggin, 1949). The pottery of
this period shows a relationship to Franklin Plain and Early
Swift Creek. The burial mound, with primary burials is a
feature of the pre-Cades Pond period. The influence of the
Gulf tradition is evident and continues into the Cades Pond

Cades Pond Period

Central Florida was strongly influenced by the St. Johns
tradition and the Cades Pond period can be considered a
western extension of St. Johns la. Pottery was made of a
chalky paste and tetrapodal bases had some local popularity.
Influences of the Gulf tradition are also reflected in Weeden
Island ceramic traits and types. Burial mounds containing
secondary burials are found.

The Alachua Tradition4

The Alachua tradition centers in the agricultural lands
of Alachua County and seems to have resulted from a south
Georgia influence upon a background of St. Johns and Gulf
tradition influences. This took place in the Hickory Pond
period. The tradition is defined as a sedentary agricultural
complex with extensive villages. Pottery includes gritty and
chalky wares, both plain and checked stamped. Cord-making
and cob-marking decorative techniques are also employed.
While burial mounds were constructed and used for intrusive
burials they are not common. Ceremonialism does not seem
important and no temple mounds have been found. The pat-
tern continued to historic times and its correlation with the
western Timucua Indians is of little doubt.

4Summarized from Goggin, 1949.


The Hickory Pond Period

The Hickory Pond period correlates with Weeden
Island II and St. Johns Ha and ceramics of both of these
periods are found in central Florida. Moreover, distinctive
Hickory Pond types such as Gainesville Linear Punctated
and Prairie Cord-Marked turn up in Gulf coast sites. This
proclivity to trade continues into the later Alachua periods
and the corresponding northwest Florida Fort Walton and
later periods.

Though accurate dating is valuable, it seems almost
regrettable that radio-carbon dates have thrown the pre-
viously neatly ordered southeastern sequences into such
confusion. Admittedly, the comparative charts which appar-
ently tied tradition and culture periods into snug little bundles
were good, educated guesses. An attempt to reconstruct
these charts based upon interpolating the scattered radio-
carbon dates which are available are apt to add to the con-
fusion. Nevertheless, some time reference is necessary to
hang our Florida sequence upon. Therefore, Goggin's (1952)
chart of areas and periods of culture in Florida is reproduced
with the revisions which Bullen's (1958a) radio- carbon dates
would suggest. It is quite noteworthy that from the time
Goggin made his preliminary definition of archaeological
periods in Florida (Goggin, 1947) until Bullen' s radio-carbon
data, the date of the appearance of fiber-tembered pottery
in Florida has been pushed backward from 500 A.D. to
2, 000 B. C.


Because of the similarity and proximity of the sites
and the almost identical circumstances under which they
were discovered and excavated, the general aspects of the
project are discussed as one operation. At both sites, bull-
dozing operations turned up numerous spalls, chert artifacts,
and fossilized mammal bones. The Florida Geological
Survey, through Clarence Simpson, has been interested in
Archaic man for many years and in the winter of 1951 sys-
tematic excavations were started by the Edwards-Simpson
party, under the direction of W. A. Edwards.














+ I t t











II b-c

II a

I b






















.5. 1 a.

Areas and periods of culture in Florida.

(After Goggin, 1952 and Bullen, 1958.)






+ I4










Figure 6.



Unfortunately, over a large area at both sites, the
stratigraphy had been well upset by the bulldozing. It was
necessary either to go outside of the bulldozed zone for a
clear cut stratigraphic sequence or totreat all of the material
as a surface collection unless it came from levels which
were definitely undisturbed. Even in this latter case cultural
sequence above the undisturbed levels would be hopelessly

The seriousness of the loss of these upper strata can
be shown by the fact that, aside from 12 small and widely
scattered sherds at the Darby Spring site, the only post-
Orange ceramic remains were found in a section of the
Hornsby Springs site which had escaped the bulldozer and
which was confined to a relatively small area along the
southern edge of that site. Three possibilities are suggested:
(1) the post-Orange occupation of the sites was confined to
this limited Hornsby Springs area; (2)bulldozing had removed
all signs of post-Orange occupation; (3)the sites, at least in
post-Orange times, were not occupied. The second possibility
can be eliminated because the removed earth was used to
fill the lower areas at Hornsby Springs and several test pits
were sunk throughthis fillwithout turning up ceramic mate-
rials of any type.

Before excavation, tremendous amounts of lithic
materials were evident, even to the casual observer.
Flakes, chips, and spalls littered the surface. Though
several identifiable artifacts are found among these lithic
materials from the surface collection at Darby Spring, only
two identifiable artifacts, a hammerstone and a projectile
point, were found on the Hornsby Springs surface.

Excavations at both sites started with test trenches.
However, the proximity of the limestone to the surface at
times on the surface led to a decision to dig either in the
higher, relatively undisturbed, and therefore stratigraphi-
cally deeper spots, or in places where breaks or solution
"pipes" in the limestone would permit some depth of exca-
vations. Consequently, the excavations really amounted to
digging a large number of test pits, some of which, espe-
cially at the Darby Spring site, could be connected in such
a manner as to form a trench.


The excavations were made using 6-inch levels in
5-foot squares exceptwhere openings in the limestone made
smaller holes necessary. Though the arbitrary stratigraphic
method was used, the natural strata, revealed as distinct
and separate soil strata by the test profiles, were noted in
the field. The field notes were so supplemented that any
significance this natural strata might lend to determining
method of deposition or chronology is not lost. Represent-
ative profiles of the naturally occurring strata are found in
figure 7. These profiles are numbered and the site map of
Hornsby Springs (fig. 2), is divided into distinctive zones so
numbered that the stratum within a zone is, essentially,
similar to that of the representative profile of the same
number. The stratification at Darby Spring closely follows
that of the one Darby Spring pattern shown inthe represent-
ative profiles.

All excavated material was screened through --inch
mesh, though care was exercised in trying to remove dirt
in such small quantities that any archaeological material
might be revealed in situ.

In all, 56 test squares were dug at Hornsby Springs,
36 were dug at Darby Spring, and one was dug about one-
fourth mile down the Hornsby Springs run at a place designated
as the Meander site. Of all of these, only three of the
Hornsby Springs test squares failed to contain any archaeo-
logical remains. None of the squares at the Darby Spring
site were devoid of the signs of human industry, though 13
contained only miscellaneous spalls and chips.

In addition, four other sites in the central Florida
region, the Archer, Butler, Itchtucknee and Marchant sites,
were explored by the Edwards-Simpson survey. The mate-
rials from these sites, while used in the formation of lithic
types and for comparative purposes, do not otherwise enter
into the discussion in this report. It was found that all of the
pits mentioned in the field notes did not appear on the site
maps and now, seven years after the survey, it is unlikely
that they could be relocated accurately. It is fortunate that
pits from which the most significant materials were taken
were not among those omitted from the maps.

LIlHT R ily lN MARL HO 554



------- ---FORMAT-i ON-


Analysis of the Materials

In spite of the wealth of lithic materials at the Hornsby-
Darby site it became apparent that only a small percentage
has spatial or temporal diagnostic value. Though lack of
comparative work could be partially responsible for these
diagnostic shortcomings two other factors seem pertinent:

(1) Most of the utilized flaked and chipped
materials are of widespread, generalized
types which occur over extremely long
reaches of space and time.

(2) Some of the unspecialized forms are quite
likely isolated occurrences which are the
results of local, experimental accident
rather than the result of cultural diffusion.

These materials are significant only in their abundant
distribution in virtually all of the area and strata excavated.
There are, however, some occurrences of lithic artifacts of
such types or in such association that they warrant close

At Darby Spring in Pit No. 1 at 45 inches depth and in
and adjoining Pit No. 0 at 54 inches depth, Suwannee-like
points (Goggin, 1950, p. 46) were found. (Catalogue Nos.
A-638 and A-639, plate VII, h and g respectively. ) In Pit
No. 1 in the 30-36 inch level, the broken base of what was
probably a Bolen point (Bullen, 1958) occurred. (Catalogue
No. A-640, plate V, b. ) The Suwannee point lay deep within
a strata of dark orange-yellow sand; ontop of this stratathe
Bolen point was found. Overlying this orange-yellow soil
there is a gradual shading into an orange-brown strata and
this, in turn, is overlaid by a gray-white sand found from
26 inches in depth to the ground surface. In this gray-white
sand in Pit No. 1 at 21 inches, and in Pit No. 0 in the 18-24
inch level, plain fiber-tempered sherds were found.

This Suwannee point-Bolen point-plain fiber-tempered
sherds sequence is most important inthat it gives clear cut


stratigraphic confirmation to their relationship as recently
postulated by Bullen (1958).

Assuming that our comparative typology is valid and
applying Bullen's radiocarbon-based dates, we find that the
36 inches of sand from the deepest Suwannee point to the
level of the fiber-tempered sherds contains materials which
can be dated thus: 9,000-7,000 B.C. for the Suwannee
points, 7,000-5,000 B. C. for the Bolen points, and 2,000-
1, 500 B.C. for the sherds. To carry our sequence a bit
further we find that, overlying the Plain fiber-tempered
sherds, there are, in Pit No. 0 at the 6-12 and 12-18 inch
levels, sherds of Orange Incised pottery (plate II, a, b).
These may be dated from 1, 500 to 1, 000 B. C.

There are, in the Darby Spring surface collection,
eight sherds of a rather chalky, sherdtempered paste which
appear to be Tchefuncte Plain ware; two grit plain sherds;
and one comparatively fine, hard piece, Lochloosa Punctated.
In Pit No. 3 one untempered sherd of fine sandy texture was
found in the 12-18 inch level. This sherd was possibly in-
cised though the material is so soft and crumbling that posi-
tive determination of surface decoration or lack of it is now
impossible. Two sherds of a dark, gritty ware, fairly hard,
were found just below the surface in Pit No. 507. Inasmuch
as this pit does not appear on the original maps of the site
these two sherds are not added tothe extremely meagre sherd
count for the site.

As we can see, materials which date later than the
Orange period do occur above these fiber-tempered levels,
but it is above these levels that we begin to lose the orderly
stratigraphy necessary for accurate work. At Darby Spring
there are only 14 sherds other thanOrange Plain and Orange
Incised. What lithic materials occur above the fiber -tempered
levels are of such generalized nature that an attempt to place
them temporally would be of questionable value.

With such a small sherd count it seems useless to dis-
cuss the proportionatetype frequencies existing at the site.
Standing by themselves, our pottery types would be deter-
minative of little or nothing. Fortunately, they are types of
rather definite temporal and spatial range and are based
upon extensive and well corroborated evidence of others.


It seems likely that the materials, at least as high
as the upper fiber-tempered levels, are not in inverted se-
quence, and it is highly improbable that any of the materials
below those levels are intrusive. When it is considered that
two practically identical Suwannee points occur within 55
inches horizontally and 6 inches vertically, the logical as-
sumption is that the points are on, or near, what was at one
time the ground surface.

Of course, it would be inconclusive evidence to base
relative antiquity on a provenience chart of only 44 projectile
points. We do find, however, that what evidence we do have
tends to confirm previous temporal hypotheses in the area.
The Florida pottery sequence has been fairly well established
and absolute dating of the various pottery styles is but a little
less certain. Onthe other hand, dating of lithic materials is
a more difficult problem and a problem that promises no
easy solution. Only a tremendous amount of comparative
work can allow us to assign, even tentatively, a temporal
position to particular lithic types.

Suwannee points, on a typological basis alone, are
assigned an early dating (Bullen, 1958), a date whichNeill's
(1958) stratigraphic work at Silver Springs tends to corrob-
orate. The provenience at the Hornsby-Darby sites further
confirms this dating. Bullen's work at Payne's Prairie plus
extensive comparative work (Bullen, 1958a) led to tentative
dating of three projectile point types: the Suwannee and Bolen
points as mentioned above, and an Arrendondo point which
was contemporary with the Bolen point. Our Type 1 and
Type 2 (see Appendix) are differentiated on the basis of
thickness, shape of blade, and the location and shape of the
side notch. Both, if basally smoothed, might be classed as
Bolen points. Actually, two of our Type 1 and two of our
Type 2 points are basally smoothed. Type 3 (plate V, f, g,
and h) is tentatively assigned an Archaic dating. Type 4
(plate VI, a, b) appears most often in an Early Woodland
position. The temporal positions of Types 5 (plate VI, c, d,
and e) and 6 (plate VII, a, b) are indefinte though Type 6
would seem to compare with Late Woodland types. Type 7
(plate VII, c, d) shows similarity to Early Woodland points.
One specimen, described in Appendix II as provisional Type 9
is of interest because of its resemblance to a Gypsum point

Da-0 Da-1 Da-3

0CQ CO 1 03 a) 10 OQ CD I*
Sw0 r-4 r-i C12 w -I rH CQ wD H H Cl
44 I I I I I I I I I I I I
0 w w 0 wO C-2 w 0 '0 N3 0D

o 4

DO Lochloosa Punetate 1 -

Grit Plain 2---------------------

Tohefuncte Plain 8

& 2Orange Incised 2 26

o0 Orange Plain 3 1

Figure 8. Provenience of Darby Spring pottery types.


(Wormington, 1957, p. 157-160, fig. 53), though it seems
that this form could be achieved rather easily by accident.

The Suwannee-like and Bolen points are heavily
patinated and while patination, especially in these highly
mineralized spring areas, is, per se, no criterion of antiq-
uity, desilicification of the deep-lying lithic materials has
proceeded to a marked degree. Upon breaking a worked
chip which came from the level of one of the Suwannee-like
points, it was found that a central chert core was all that
remained of the original material. Two-thirds of the arti-
fact had decomposed into a rather soft, flaky, and unworkable
structureless silica.

It is rather notable that there are remarkably few
skeletal remains of edible animals at either of thetwo sites.
In Pit No. 0 small bits of bone, unidentifiable, occur 2 inches
above the Suwannee point. In the Darby Spring surface mate-
rial several pieces of turtle plastron and three cannon bone
of deer were found. About the distal end of one of the cannon
bone was an engraved line.

Thus, the Darby Spring excavations indicate the
presence of a rather extensive lithic industry whose products,
while quite variegated, are equally conspicuous for their
abundance. Two projectile point types, the Suwannee point
and the Bolen point, indicate a considerable antiquity for the
utilization of the chert seams which occur here within the
Crystal River formation. In contrast tothe lithic materials,
ceramic remains arefew and evidences of burials and mid-
dens are lacking.

While it is fortunate that the Darby Spring site is large
enough to have permitted extensive excavation outside of the
bulldozed area, it is quite possible that a great amount of
significant archaeological material was removed and now
lies in the abutment fill for the highway bridge over the
Santa Fe River.

Hornsby Springs, like most of the large springs in
Florida, issue from underground limestone caverns. Cooke
(1939, p. 89) in his explanation of the formation of this type
of springs said that:


"The cavity through which water ascends to an
artesian spring is generally a former sink in
which the direction of the motion of the water
has been reversedbythe rise of the water table.
If the water table were to fall below the m. ith
of the cavity, the spring would cease to flow
and would revert to the form of a sink . .

"As some of the springs risefrom depths well
below sea level, they presumably corroborate
other evidence that sea level has risen. The
depths to which the caverns extend below sea
level appears to give a minimum measure of
the amount that sea level has risen since the
caverns were formed. "

Perhaps, as ancient limestone sinks, these springs
of the future served as natural traps for the ancient probos-
cidians, or, had the springs already assumed their present
form, their swampy shores may have bogged down the big
beasts. An even more exciting possibility is that ancient
man may have driventhe animals intothese places and there,
with little danger to himself, killed and butchered them.
Clovis and Plainview men may have done something like this
(Wormington, 1957, p. 56, 108) and since ancient Florida
man was contemporary with mammoth, he would have done
the same. The most that can be said with certainty is that
several of the large Florida springs contain some seemingly
old cultural materials together with fos silized remains of the
Columbian Mammoth and Florida Mastodon (Simpson, 1948,
p. 13; Neill, 1958, p. 35; Olsen, 1958).

Inthe springhead and the spring run at Hornsby Springs
fossilized mammoth remains have been and are being found,
some in association with lithic artifacts. This association,
because of fortunate geologic circumstance can now be ac-
cepted as true, since the preceding discussion of the geology
at Hornsby Springs points out clearly that articulated or
partially articulated skeletal remains were found in situ with
lithic artifacts; and, moreover, the geologic dynamics that
occurred at the spring explain how this association came


The ground surface of the basin surrounding Hornsby
Springs was, like that of the Darby Spring site, littered with
the chips and flakes of an extensive lithic workshop, and,
like the Darby Spring site, bulldozing operations had de-
stroyed associations and stratigraphy over a large part of
the area. Only a small section along the southern side of the
site seemed to have escaped reasonably intact; here alone
would there have been a sufficient depth of undisturbed soil
to determine a valid stratigraphic sequence. Sixty-three of
the sixty-seven sherds found at Hornsby Springs came from
this small part of the site. Although seven projectile points
are found in the excavations in this corner of the site they
fit into our types 5, 6, and 7 which, as we have indicated in
our discussion of Darby Spring, are difficult to tie down
temporally. The remaining flakes and chips of chert do not
lend themselves to typological groups nor can they be dated
absolutely or relatively.

In the Hornsby Springs basin the material bulldozed
fromthe surfacewas usedtofill and level the area surround-
ing the spring. Some excavations in the basin were made
through this material without turning up a single sherd though
chips, flakes, and lithic artifacts are numerous.

In most of the basin area the surface was scraped
down to a rather dense, hard, fresh-water marl strata a
discouraging prospect for archaeological excavation. How-
ever, several pits were sunk through this strata. Imbedded
in the marl itself were chert flakes and chips and in Pit
No. 111 a large, well shaped plano-convex scraper or blade
was removed with the enclosing marl matrix. On exposure
to the air the marl became so hardened that it was inadvis-
able to remove the scraper by chipping the marl away from
it. A solution of dilute hydrochloric acid dissolved the marl
completely with no damage to the scraper.

Of the excavations in the basin area roughly Zone III
on the Hornsby Springs map none were deeper than 36
inches. It is not surprising, therefore, that a temporal se-
quence of artifacts is not apparent and, considering the
disturbed condition over much of the site, any attempt to
justify a sequence based on stratigraphy should be viewed
with suspicion.



S0-82 s o-91


1 I

Ho-150 o-610
S f l
-s 1 s .... s

I rrdrl
Loobloosa~~TF - I 2IS Illll

luplo Stped .d
St. o b ,
Cheak St-ped 1
St. John 1 1
Plan --- ------- --- --- -- -- -- -
Tohofunoo 1 o
Plain -
Deptford 1 1 1

D.ptford n
Chkc St.ped 2

--------- --_ --.- P
a lag. I ft LLL I I I I [ I I I I

Figure 9. Provenience of Hornsby Springs pottery types.


s Aa3s






Pits 71, 72, 91, 100 and 150 contain sherds of Loch-
loosa Punctated, Plain Urit, Tcheiuncte Plain, and St. Johns
wares. Pits 610 and 611 contain Orange Plain sherds. These
suggest that the site may have been in use from the Archaic
to, possibly, the proto-historic. Further, the occurrence
of a Suwannee-like point inPit No. 110 at 30 inches suggests
that utilization of the site might be extended backward to the
Paleo-Indian period. The base of this point (specimen No.
A- 1533) has a constricted fish-tail stem. It is basally ground
and one face is fluted. There is some possibility that it
might be classified as an Arrendondo point though it would
be quite aberrant for the type.

In Pit 550 there was a definite physical association of
fossilized mastodon remains and worked lithic material.
Plate I shows the chert flakes (rounded paper arrows point-
ing) and fossilized mastodon teeth (triangular paper arrows
pointing) in situ. 5 These associated materials were found
in a sandy muck pocket within a solution tube in the Crystal
River formation. The muck was completely enclosed within
a deposit of fresh-water marl. It was possible to date the
deposition of the marl by carbon-14 dating at 9, 880 270
and this is the latest possible date at which this lithic-
skeletal physical association could have taken place.

As at Darby Spring, the sherd count from Hornsby
Springs is astonishingly meagre considering the size of the
site, the number of excavations, and the otherwise abundant
signs of use of the area. One reason for the distinction be-
tweenthese and the average, predominantly ceramic-bearing
Florida archaeological sites could be attributed to a pre-
ceramic temporal positionfor the period of major utilization
of the sites. However, what ceramic remains are found
would seemtoindicatethat this maynot be the sole causative
factor. Rather, the great predominance of lithic over ceramic
remains at both of the sites must be indicative of the use to
which the sites were put.

5In plate I-A the top of the photograph is north; in plate I-B north is to the



Hornsby Springs, Pit #550, showing mastodon teeth

and flint flakes in situ.



Only with a considerable strain of credulity can the
ceramic data be "fitted in" to make a case for a major oc-
cupation of either site. Though Phillips, Ford and Griffen
(1951) pointed out that at times strata-cut tests "misrepresent
the history of the site by completely skipping or being defi-
cient inthe material that represents certain spans of time, "
still, not all of the large number of test cuts made could
have skipped or been deficient in ceramic materials if these
artifacts had been there to find.

A major problem in American archaeology in general
and southeastern archaeology inparticular is the determin-
ation of the steps in the development of the early nomadic
hunters into the specialized hunter similar to the Clovis-
Folsum type and the further development from the Clovis-
Folsum type to the less specialized Archaic hunter and col-
lector. The tool kit of pre-Clovis and Folsom man may be
represented byunspecialized lithic implements but, because
of their generalized types, these may well have carried,
accidentally or intentionally, down to historic times. The
very fact that they are such generalized forms makes
temporal categorization seem quite impossible.

The picture of the development of the early specialized
hunter and the gradual emergence of the Florida Archaic
tradition is just beginningto assumeform and at present any
scrap of evidence, however small, helps to fill in this
picture. The tremendous amount of work necessary to de-
rive some bit of valuable evidence is dramatically illustrated
by the fact that of the 93 pits dug in the Hornsby-Darby area,
it was from but two, Pits 0 and 1 at Darby Spring, and Pit
No. 550 at Hornsby Springs, that the most meaningful
stratigraphic dating evidence was derived. Nevertheless,
without the evidence fromthe remaining 91 pits the emphasis
onthe lithic predominance at the sites and their vertical and
areal extent would have been diminished or minimized.

In the deepest strata at the Modoc Rock Shelter in
southern Illinois Fowler (1959) reports side notched points
remarkably similar to our type 1, which are, in turn, similar,
if basally ground, to Bullen's Bolen point. Fowler dates his
points from 8,000 to 6,000 B.C. almost identical with
Bullen's Bolen point dating. Strangely enough, however,


these points were from deeper strata than the fishtail, fluted
forms generally regarded to be the earliest in the east.
Though Fowler sees the Early Hunter materials of the east
attributable to the general Clovis-like horizon he assigns
his deeper lying side-notched points to an Archaic horizon.
He says that "rarely in the eastern United States at the
present time . has a site been found where Early Hunter
artifacts have proven to be stratigraphically earlier than
Archaic materials" (Fowler, 1959, p. 268). But, if we
accept Suwannee points as Clovis-like, then in the Darby
Spring and Hornsby Springs sites we have an eastern local
which does contain Early Hunter materials in a position
stratigraphically and chronologically earlier than Archaic


The Darby-Hornsby springs excavations have revealed
two large lithic workshops in north-central Florida. Typo-
logical comparisons of lithic materials indicate that the chert
deposits which occur here in the exposures of limestone of
the Ocala group were utilized from the time of the Early
Hunter through the proto-historic and, perhaps, into the
historic period.

Though further investigation may reveal a nearby vil-
lage site it would appear that, on the basis of the meagre
ceramic remains, the few skeletal remains of edible animals,
and the total absence of midden deposits, neither site
represents an occupation situation though temporary camps
to quarry or work the chert may have been established. The
development of agriculture in the ceramic period probably
precluded the likelihood that an aboriginal village would be
built on the irregular rocky soil of the immediate area.

The frequent discovery of fossilized remains of the
extinct proboscideans in the survey area has lead to the
discovery of a bona fide contemporaneous association of
lithic artifacts and some Pleistocene skeletal material at
Hornsby Springs site. The contemporaneousness of these
materials seems well supported by the geological evidence
presented in this report and the earliest date assignable


to this physical association is 9,880 270 years as revealed
by radiocarbon dating. Darby Spring site amply supports
the assignment of this early date, since with only the tools
of stratigraphy and typology the earliest date for Darby
Spring could be placed at 9, 000 years. It is further con-
cluded based on the evidence that both Hornsby and Darby
springs sites were coeval.

Though a tremendous amount of work was done and
great quantities of lithic material recovered, what seems to
be of major importance is the existence of a Paleo-Indian-
Preceremic- Early Archaic sequence in Florida with a begin-
ning date of 9, 880 270 years as revealed in the Suwannee
point-Bolen point-Fiber tempered pottery sequence excavated
at Darby Spring and Hornsby Springs and the Pleistocene
vertebrate material in association with lithic artifacts be-
neath sediment dated by radiocarbon measures.






The Hornsby-Darby sites are not essentially, ceramic
sites. Of the recorded total of 92 test squares dug, only 12
contained ceramic remains and, of these 12 ceramic-bearing
squares, only four contained sherds below the 24-inch level.
These four, as explained in the discussion of the sites, are
grouped along the southern edge of the Hornsby Springs site.

All of the materials lithic and ceramic with the
exception of most of the flakes and spalls, were first sorted
according to the pit and the level from which they came.
They were then laid out in this order upon the floor of a large
room. This permitted a simultaneous, overall view of all
of the artifacts frombothof the sites and inthe order closely
approximating that in which they lay in the ground before

This section concerns the typology of the ceramic re-
mains from the Hornsby-Darby sites. The source of the
type description and a brief synoptic sketch of the types is
presented in the order of their temporal occurrence.


Orange Series:

Orange Plain (Griffen, 1945, p. 220; pl. I c, III b) .
This type is considered the basic fiber-tempered ware of
east and central Florida. Ideally, fiber is the only temper-
ing material which, on firing, was burned out leaving the
characteristic vesicular appearance. Later specimens may
have quartz sand added to the fiber as temper. While the
surface is often well smoothed, there is usually a granular
appearance due to the burned out fiber holes. The rim is
usually simple and straight-sided with no differentiation
between a rim fragment and a section of the side wall. The
lip is narrow and rounded or slightly flattened. The base is
normally flat and unmarked.


This pottery is most commonly found to the east of the
Alachua area but it is also present in central Florida. The
few fiber-tempered sherds found occasionally in most other
parts of Florida attest to the influence of this pattern.

Florida is one of the major centers of the fiber-
tempered ceramic tradition of the southeast. Bullen, in his
latest radiocarbon dating (Bullen, 1958a) sets the beginnings
of early Orange Plain pottery at 2000 B. C.

Orange Incised (Griffen, 1945, p. 219, pl. II a, b).
Except for the incised decoration, the ware form, vessel
form, and surface finish are similar to Orange Plain. The
rectilinear designs are made by incising with sharp or
rounded tool on the wet clay and consist of extensive hatching
arranged in a number of designs.

The areal range of this type is identical with Orange
Plain. The temporal position suggested by radiocarbon dating
(Bullen, 1958a) is from a beginning date of 1500 B.C. to
slightly later than 1000 B. C.

The St. Johns Series:

St. Johns Plain (Griffen, 1945, p. 220, pl. IV e). This
plain, chalky ware gives every indication of having developed
or evolved from the earlier Orange Incised. Some of the
earlier examples contain fiber temper and, though the typical
St. Johns paste is temperless, later types may merge into
a grit-tempered series. The early types are quite soft and
can be scratched by a fingernail; the later types are much
harder. Early examples, from the Orange period, frequently
are rather thick and crudely finished.

This chalky ware seems to be a local development which
spreadto other parts of Florida and it is found in muchof the
peninsula. It ranges in time from the late Orange period
well into historic times.

St. Johns Check Stamped (Goggin, 1952, p. 103-104,
pl. IV a). This is one of the most important and widely distri-
buted types found in Florida. The paste is similar to that


of St. Johns Plain but the later, hard forms, seem to be
found along the east coast. The check stamping was usually
over the entire vessel surface and in its execution ranged
from neat to careless, overlapped imprints. Check size,
too, is not consistent and there is indication here of tem-
poral and geographical differences.

St. Johns Check Stamped does not occur earlier than
the St. Johns II period and is characteristic of that period.
It continues, in diminishing quantity, into historic times.

St. Johns Simple Stamped (Griffen and Smith, 1949,
p. 346-347). The paste is similar to that of St. Johns Plain.
The type is characterized by parallel, or slightly crossed,
simple stamping. Though this type is widely distributed it
does not seemtobe common at any of the sites where it does

It is dated in St. Johns II B or II C times.

The Deptford Series:

Deptford Linear Check Stamped (Willey, 1949, p. 354-
356), pl. III d, IVd). This ware was the earliest appearing
check stamped pottery in the southeast. It was manufactured
of compacted, fine sand tempered paste by a coiling technique.
The designwas impressed onthe vesselwith a carvedwooden
or bone instrument and is distinguished by its large, pro-
nounced, parallel lands of one direction and the smaller
transverse lands of the other. There is considerable over-
lapping of the stamping.

This type is found throughout northern Florida and
northward into South Carolina. Some sherds have been found
in east Tennessee. It represents an early ceramic sequence
wherever it occurs, usually shortly after complexes of plain
and decorated fiber-tempered pottery.

Deptford Bold Check Stamped (Willey, 1949, p. 357,
pl. III C). This type is similar to Deptford Linear Check
Stamped except that the individual rectangles of the check
stamped units are square and the lands in both directions
are of approximately the same size. Squares are deep and
the lands bold and wide. Individual vessels may have been


slightly thicker than those of the Linear Check Stamped
de sign.

The geographical range is the same as Deptford Linear
Check Stamped. The time range is slightly later.

Deptford Simple Stamped (Willey, 1949, p. 357-358,
pl. III e). The temper used in the paste of this ware seems
slightly coarser than that of the two preceding Deptford
series types. Decoration consists of stamped, linear,
parallel grooves.

Willey does not mention the occurrence of this type
in central Florida. It is found from northern and western
Florida along the Atlantic Coast into South Carolina. Its
chronological position is not clear cut in Florida. It seems
to be coexistent with Deptford Linear Check Stamped but it
continues much longer and is found in association with Weeden
Island types.

Tchefuncte Series:

Tchefuncte Plain (Jennings, 1952, p. 259-260, pl. II e,
IV c). The Tchefuncte culture can best be understood as a
late Archaic period or, better, a transition from the gener-
alized Archaic to the more specialized base which ensued.

Pottery was manufactured by a coiling and smoothing
technique. The paste was soft and chalky and was tempered
with angular pellets of clay and fine sand. The vessels were
cylindrical with nearly vertical sides, some with tetrapods
and occasionally with notched lips.

Similar types are found in Louisiana; from Stallings
Island, Georgia; Pickwick Basin, Alabama; Florida; and
north Mississippi.

Alachua Series:

Lochloosa Punctated (Bullen, 1953, p. 57, pl. IV b).
This type of pottery has not been well described in the liter-
ature. Bullen's brief description is as follows:


"It refers to gritty pottery decorated by ir-
regularly spaced punctations or "jabs" made
oy a pointed tool. While the temporal range
of Lochloosa Punctated is not well defined at
present, Goggin would place the time of its
maximum popularity as late Weeden Island or
possibly later. "



a Orange Incised Pottery. Darby Spring, Pit # 0, 6-12
inch level.

b Orange Incised. Darby Spring, Pit # 0, 12-18 inch

c Orange Plain. Hornsby Springs, Pit # 610, 18-24
inch level.

d Grit Plain. Hornsby Springs, Pit # 72, 30-36 inch

e Tchefuncte Plain. Hornsby Springs, Pit # 71, 12-18
inch level.


c d



a Semifiber-tempered ware. Hornsby Springs, Pit #41,
0-6 inch level.

b Orange Plain. Hornsby Springs, Pit # 611, 30-36 inch

c Deptford Check Stamped. Hornsby Springs, Pit # 72,
18-24 inch level.

d Deptford Linear Stamped. Hornsby Springs, Pit # 150,
36-42 inch level.

e Deptford Simple Stamped. Hornsby Springs, Pit # 72,
36-42 inch level.



1PITT 31IT l II T7 i /V /
I. 'I 21 1 34 1 6



a St. Johns Check Stamped. Hornsby Springs, Pit # 71,
12-18 inch level.

b Lochloosa Punctate. Hornsby Springs, Pit # 71, 18-24
inch level.

c Tchefuncte Plain. Hornsby Springs, Pit # 91, 6-12
inch level.

d Deptford Linear Stamped. Hornsby Springs, Pit # 100,
0-6 inch level.

e St. Johns Plain. Hornsby Springs, Pit # 72, 18-24
inch level.


~: ~

7) C




Lithic Materials



Lithic Materials

In organizing the lithic materials for classification,
Krieger's (1944) typological method of analysis was employ-
ed but with the description form following that agreed upon
at the 1951 Southeastern Archaeological Conference. It
would seem advisable in using this form to refrain from
placing all artifacts intypes, and that types should be named
only when they might show definite temporal and/or areal
significance. This would mean that types should not be
named from a single site in most cases, but should be based
on comparative study in both time and space.

Two methods are used to present these materials:
(1) the descriptive form mentioned above, and (2) photo-
graphs of individual artifacts of the various types.

Artifacts and sherds are pictured in plates II-XIV.
The Darby Spring material in the Simpson private collection
at High Springs were photographed and those of apparent
importance were carefully examined and measured for com-
parison of types.

While there are many projectile points from among the
artifacts at the two sites, they are far outweighed numerically
by the blades, scrapers, choppers, and other lithic artifacts
recovered. Chips and spalls were countless. In the few pits
where artifacts or spalls did not occur to any depth it is
usually only because that particular pit "bottomed out" on
limestone of the Ocala group at a relatively shallow depth.

In classifying the projectile points the following
descriptive terms are used:


A point is long (or short) if its length is greater (or
less) than six centimeters. It is narrow (or broad) if the
length-breadth ratio is greater (or less) than 2:1.

Description of the shapes, parts, stems, notches, and
other features, techniques, and methods of manufacture fol-
lows that of Martin, Quimby, and Collier (1947, p. 34-39).

In figure 10,provenience of projectile points, the fol-
lowing abbreviations are used:

Ho Hornsby Springs site
Da Darby Spring site
Bu Butler site
Me Meander site

Four projectile points from Goodacres Midden are included
in the figure. This site was located on the east bank of the
St. Johns River about five miles north of Sanford, Volusia
County, Florida. This site is abbreviated GM.


Projectile Points:

Provisional Type 1 (pl. V a, b); 5 specimens

Outline: Triangular; short and broad
Cross section: Bi-convex
Edges: Straight to convex, tapering to tip
Base: Concave to slightly convex. Average breadth
of base 25 mm. Occasionally show basal
Stem: A continuation of the blade but separated by
the side notches.
Notches: Very prominent side notches
Barbs: Prominent to lacking
Blade: Roughly isosceles


1 2 3 4

Do 13
2 specimen


0"- 6"

6"- 12"

12"- 18"

18"-2 4'





Ho 81


Ho 71

6 7 8

Ho 51

Do= Z
Ho 150

Ho 91

Ho 71

Dol HoiO HoO2 HollO
30"-36" D'o02 Bu3

36 42

Do6 Do102 Bu3 Do I
42"- 48" Ho512 HoSI1

Do o
48"-54 2

Figure 10. Provenience of projectile point types.

Do 101
Do 105




Length: 39-58 mm. Average 52 mm.
Breadth: 22-30 mm. Average 25 mm.
Length-breadth index: 1.5/1
Thickness: Average 6 mm.
Weight: Average 4.5 gms.

Technique: Fine pressure flaking, about four chips per
cm.; one specimen "spinner" chipped opposite edges
beveled on opposite sides of point.

Material: Chert
Color: Yellow-brown, tan
Texture: Smooth
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably arrow

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson
archaeological survey is shown in figure 10.

Other occurrences of points which appear to be typo-
logically similar follow:

(1) Jefferson County, Florida (Allen, 1954, p. 134,
provisional type 15, pl. XVII H).
(2) Alachua County, Florida (Bullen 1958b, p. 14,
39, pl. I G-I).
(3) Badin Focus of the Carolina piedmont (Coe 1952,
p. 306-307, fig. 163, F,G).
(4) Kleine site near Jonesboro, Alabama (Josselyn,
n.d., 199 F, 200 M).
(5) Candy Creek, eastern Tennessee (Rowe 1952,
p. 199-201, fig. 103 G).
(6) Western Tennessee (Lewis and Kneberg 1947,
p. 18, pl. 1, side notched type).
(7) Parrish Village site, Hopkins County, Kentucky
(Webb 1951a, p. 428, fig. 5).
(8) Gulpha Creek, near Hot Springs, Arkansas (Orr
1952, p. 240-241, fig. 130 C).
(9) Brewerton Focus of western and central New York
(MacNeish 1952, p. 47-48, fig. 14 A-D).
(10) Faulkner site, Massac County, Illinois (MacNeish
1948, p. 236-237, fig. 47, no. 13-15).


(11) Titterington Focus of the southern Illinois River
valley (Wray 1952, p. 158, fig. 67 E, F).
(12) Raymond and Dillinger Foci of southern Illinois
(Maxwell 1952, p. 186-188, fig. 100, no. 4).
(13) Black Sand Focus of central Illinois (Wray 1952,
p. 158, fig. 69 A-C).
(14) Osceola site, Potosi, Wisconsin (Wormington 1957,
p. 153, fig. 50).

Temporal Range: Typologically similar points seem to
be found most often in early Archaic associations
though three late Woodland occurrences are noted.
Bullen's work at Paynes Prairie (Bullen 1958)has re-
sulted in the definition of a Bolen point, similar to
Type 1 in form, but whose basal edges are ground.
Actually, two specimens from Type 1, A-1544 and
A-640, show basal smoothing. Bullen dates the Bolen
point from 7000 B. C. to 5000 B. C. Judging by the
levels at which Type 1 points are found and the asso-
ciated artifacts at the Darby Spring site, an early
Archaic dating is suggested for this point inthis area.

Remarks: The salient features of this type are the promi-
nent side notches; shoulders or barbs are obscured by
the wide base. These points show good to excellent
workmanship and extreme symmetry of form.

Provisional Type 2 (pl. V c, d, e); 4 specimens
Outline: Triangular; short and broad
Cross section: Bi-convex or maybear rough similarity
to a parallelogram or hexagon
Edges: Straight to slightly convex
Base: Slightly convex or slightly concave; may show
basal smoothing.
Stem: Expanding
Notches: Corner notched
Barbs: Prominent
Blade: Isosceles, 3 specimens serrated
Length: 29-40 mm. Average 33. 5 mm.
Breadth: 20-30 mm. Average 24. 5 mm.
Length-breadth index: 1.4/1
Thickness: 8-9 mm. Average 8.3 mm.
Weight: 3. 1-6. 5 gms. Average 5. 1 gms.


Technique: Fine pressure flaking, 3-4 chips per cm.,
either biface of "spinner" form.

Material: Chert
Color: Gray-brown, dark red, tan
Texture: Smooth
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably arrow

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards Simpson
archaeological survey is shown in figure 10.

Other occurrences of points which appear to be typo-
logically similar follow:
(1) West Florida Gulf coast (Willey 1949, pl. 42 E).
(2) St. Johns Region, Florida (Goggin 1952, pl. 71).
(3) Macon Plateau, Georgia (Kelly 1938, fig. 3, no. 11).
(4) Specimens from near Lovic, Bessemer, and Bir-
mingham, Alabama (Josselyn, n. d. Ill. C 1, D2,
and 0 respectively).
(5) Addicks Basin near Houston, Texas (Wheat 1953,
pl. 34 T, U, V). Wheat's Eddy Stemmed Point
(Wheat 1953, p. 202) seems to be quite similar
though smaller than this provisional type.
(6) Eastern Tennessee (Lewis and Kneberg, n. d.,
illustration on page 113, 2 specimens at top).
(7) Western Tennessee (Lewis and Kneberg 1947, pl. 2,
3rd from top, right hand column).
(8) Faulkner site, Massac County, Illinois (MacNeish
1948, fig. 47, no. 16).
(9) Site 14 PH4, the Woodruff Ossuary near Woodruff,
Kansas (Kivett 1953, pl. 23 a, no. 3, 2 specimens).

Temporal Range: Undetermined. Typologically similar
points occur from the Archaic to the late Woodland
time elsewhere. The occurrence of this type at the
Hornsby-Darby sites below any ceramic levels would
suggest an Archaic temporal position for this type at
these sites.

Remarks: This type is characterized by expanding to flar-
ing stem formed by the diagonal notches at the basal


corners. This produces a barbed shoulder which is
partially obscured by the flaring base. Three of the
four specimens are "spinner" chipped opposite edges
beveled on opposite sides of the point.

Provisonal Type 3 (pl. V f, g, h); 6 specimens
Outline: Triangular, long and broad
Cross section: Bi-convex
Edges: Slightly concave to convex
Base: Usually concave due to basal thinning; one
specimen has a rounded base; may be basally
Stem: Contracting
Notches: Corner notched
Barbs: Not present
Shoulders: Prominent to sloping
Blade: Roughly isosceles; sides straight about two-
thirds of the length of the blade, then incurving to
Length: 64-75 mm. Average 70. 3 mm.
Breadth: 37-41 mm. Average 39. 0 mm.
Length-breadth index: 1. 8/1
Thickness: 8-15 mm. Average 11.5 mm.
Weight: Average 22 gms.

Technique: Percussion flaked; edges and base pressure
flaked 3-5 chips per cm.

Material: Chert
Color: Gray, brown
Texture: Smooth
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably spear point

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson
archaeological survey is shown in figure 10.

Other occurrences of points which appear to be typo-
logically similar:
(1) Jefferson County, Florida (Allen 1954, p. 123-
124, Provisional Type 5, pl. 13 G, I, p. 158).
(2) Silver Springs, Florida (Neill 1958, pl. 1 G).


(3) St. Johns Region, Florida (Goggin 1952, pl. 7 P).
(4) Stalling's Island, Georgia (Fairbanks 1942, fig. 23,
no. 4, 5).
(5) Western Tennessee (Lewis and Kneberg 1947,
p. 18, 7th from top of page, left side).
(6) Indian Knoll Focus, Kentucky (Morgan 1952, p. 84,
fig. 29 G).
(7) Marcy Creek, Virginia (Manson 1948, pl. 23 A,
third from left, lower row).
(8) Gulpha Creek, Arkansas (Orr 1952, p.. 240-241,
pl. 130 B).

Temporal Range: Both typologically and because of its
associational relationship in the Darby-Hornsby area
this type is assigned to the Archaic.

Remarks: These points appear rather crudely made though
fine pressure flaking frequently occurs along the edges.
The off-center stem makes the point asymmetrical.

Provisional Type 4 (pl. VI a, b); 8 specimens
Outline: Oval, long and broad
Cross section: Bi-convex
Edges: Convex
Base: Rounded and thinned
Stem: Contracting
Notches: Corner notched
Shoulders: Occasionally prominent, but generally they
slope gradually into the body of the blade
Barbs: Two specimens show slight barbs on one edge
of blade
Blade: Roughly isosceles though sides are convex
Length: 60-72 mm. Average 67.0 mm.
Breadth: 35-43 mm. Average 39. 6 mm.
Length-breadth index: 1. 7/1
Thickness: 10-11 mm. Average 10.5 mm.
Weight: 22. 0-27.2 gms. Average 24.6 gms.

Technique: Percussion flaked, pressure flaking along
edges about 4 chips per cm.

Material: Chert
Color: White, gray, pink
Texture: Smooth, one specimen granular


Stratigraphical identification: Crystal River formation

Function: Projectile point, probably spear point

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson
survey is shown in figure 10.

Other occurrences of points which appear to be typo-
logically similar:
(1) Jefferson County, Florida (Allen 1954, pl. 11 G,
I, p. 156; pl. 13 B, C, p. 158; pl. 15 H, p. 160).
(2) Site Su2, Suwannee County, Florida(Goggin 1950,
p. 48, fig. 21 T).
(3) Bluffton, Volusia County, Florida (Bullen 1955,
p. 10, fig. 3 D). In strata with decorated, fiber-
tempered sherds.
(4) Silver Springs, Florida (Neill 1958, p. 38, pl. 1 EQ.
(5) Russell Cave, Alabama (Miller 1956, p. 548,
illustration at top, right).

Temporal Range: On the basis of typologically similar
points occurring at other sites in Florida, an Archaic
through early Woodland position is indicated for this
type in the Hornsby-Darby area.

Remarks: This type is characterized by its oval outline
and its contracting stem with a rounded base. It does
not seem to have a widespread geographical range.

Provisional Type 5 (pl. VI c, d, e); 9 specimens
Outline: Triangular, long and broad
Cross section: Bi-convex; one specimen concave-
Edges: Straight, incurving toward tip
Base: Rounded
Stem: Expanding or contracting
Notches: Corner notched
Shoulders: Square to barbed
Barbs: Neither sharp nor prominent
Blade: Isosceles
Length: 52-75 mm. Average 64.7 mm.
Breadth: 33-49 mm. Average 41. 1 mm.
Length-breadth index: 1.6/1


Thickness: 6-10 mm. Average 8.2 mm.
Weight: 11.7-29.2 gms. Average 18.9 gms.

Technique: Percussionflaked with edges pressure chip-
ed, 4-6 chips per cm.

Material: Chert
Color: Purple, brown, gray, white
Texture: Smooth
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably spear point

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson sur-
vey is shown in figure 10.

Other occurrences of points which appear to be typo-
logically similar:
(1) Florida West Coast (Willey 1949, p. 599, pl. 54 Bg).
(2) Silver Glen Springs, Florida (Goggin 1952, pl. 7 H).
(3) Russell Cave, Alabama (Miller 1956, Archaic
group, right).
(4) Northern Alabama (DeJarnette 1952, p. 274,
fig. 147, no. 20).
(5) Parrish Village site, Kentucky (Webb 1951a,
fig. 5B, center group, left).
(6) Ohio Hopewell (Maxwell 1952, p. 184-186, fig. 32 G).
Points of this general type appear frequently in
Hopewellian sites in the Ohio valley.
(7) Raymond Focus, Southern Illinois (Maxwell 1952,
p. 186-187, fig. 100, group 4, second from left,
top row).
Temporal Range: Indefinite. Points of this type seem to
have been in use from Ar chaic to late pre-historic time.
Remarks: Good workmanship, symmetry, and prominent
to barbed shoulders with corner notches are charac-
teristics of this type. A cross section of a broken
specimen of this type shows a high degree of patination;
one-third to one-half of the chert has been decom-
posed. On one specimen a longitudinal flake has been
removed from one face of the blade, resulting in a
fluted appearance.


Provisional Type 6 (pl. VII a, b); 4 specimens
Outline: Triangular, long and narrow
Cross sections: Bi-convex
Edges: Straight
Base: Square or concave
Stem: Contracting or absent
Notches: Corner notched
Shoulders: A gradual concave curve from base to
Barbs: Not barbed
Blade: Isosceles
Length: 59-90 mm. Average 77. 0 mm.
Breadth: 27-40 mm. Average 34.0 mm.
Length-breadth index: 2. 3/1
Thickness: Average 8 mm.
Weight: Average 12. 4 gms.

Technique: Percussion flaked with fine pressure flake
finish. 3 chips per cm.

Material: Chert
Color: Gray, brown
Texture: Smooth to porous
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably spear point

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson sur-
vey is shown in figure 10.

Other occurrences of points which appear to be typo-
logically similar:
(1) Northwest coast of Florida(Willey 1949, p. 598,
pl. 42 C, pl. 54 Bc and pl. 56f). Willey's stem-
less point may be a reworked broken point of this
(2) Battery Point, Hernando County, Florida (Bullen
and Bullen 1953, p. 87, fig. 1).
(3) Bayou Goula, Iberville Parish, Louisiana (Quimby
1942, pl. 16, no. 6).
(4) Addick's Basin, near Houston, Texas (Wheat 1953,
pl. 36 H shows some similarity).


(5) Western Tennessee (Lewis and Kneberg 1947,
p. 18, straight stem type, top left and lower right.
(6) Raymond and Dillinger Foci of southern Illinois
(Maxwell 1952, p. 186-188, fig. 100, 3 specimens
from group 4, right, fig. 101, group 22, second
from left, top).

Temporal Range: Indefinite, but this type is apparently
rather late over most of the area in which it occurs.

Remarks: Characteristic of this type are the very straight
sides terminating in a sharp tip. The form is sym-
metrical and the workmanship is good to excellent.

Provisional Type 7 (pl. VII c, d); 3 specimens
Outline: Triangular, short and broad
Cross section: Bi-convex
Edges: Straight to slightly convex
Base: Rounded or slightly concave
Stem: Contracting
Notches: Corner notched
Shoulders: Prominent to tapering
Blade: Roughly equilateral
Length: 43-49 mm. Average 45. 7 mm.
Breadth: 27-37 mm. Average 31. 7 mm.
Length-breadth index: 1.4/1
Thickness: 8-11 mm. Average 9.0 mm.
Weight: 6. 9-14. 3 gms. Average 9. 6 gms.

Technique: Percussion flaked with secondary pressure
flaking. Pine chipping along edges, 4-5 chips per cm.

Material: Chert
Color: White, tan, pink
Texture: Smooth
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably arrow

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson sur-
vey is shown in figure 10.


Other occurrences of points which appear typologically
(1) Jefferson County, Florida (Allen 1954, p. 120-
121, pl. 12 H-K; Provisional Type 3).
(2) St. Johns Region, Florida (Goggin 1952, pl. 7 J).
(3) Silver Springs, Florida (Neill 1958, pl. 1 F).
(4) Stalling's Island, Georgia (Fairbanks 1942, p. 226,
fig. 23, no. 9).
(5) Near Birmingham, Alabama (Josselyn n. d.,
p. 201, fig. 8).

Temporal Range: On the basis of associated materials
and typological similarities an Archaic to Early Wood-
land temporal range is suggested.

Remarks: This type is somewhat similar to Provisional
Type 5. It differs in that it is considerably shorter,
has less prominent shoulders, and is rather asym-

Provisional Type 8, Suwannee Points (pl. VII e-h); 5 speci-
Outline: Triangular; may be either long and narrow or
short and broad.
Cross section: Bi-convex; bi-concave near base
Edges: Partially straight but incurving toward the
tip. Edges near base may be ground.
Base: Concave; corners rounded or "eared. Base
may be smoothed. No stem.
Blade: Roughly isosceles
Length: 52-80 mm. (estimated. Average 68. 5 mm.
Breadth: 28-39 mm. Average 34. 5 mm.
Length-breadth index: 2. 0/1
Thickness: 6-7 mm. Average 6. 8 mm.
Weight: 7. 8-12. 5 gms. (2 smaller specimens only. )
Average 10. 1 gms.
Technique: Fine chipping, 3-5 chips per cm. in semi-
uniface or biface form. Base andbasal edges ground.

Material: Chert
Color: Brown
Texture: Smooth. Surface shows some tendency to


Stratigraphical identification: Crystal River formation

Function: Projectile point, probably spear point

Geographical Range: The distribution of this type of point
in the area covered by the 1951-1952 Edwards Simpson
archaeological survey is shown in figure 10.

Other occurrences of points which appear typologically
(1) Hillsborough County, Florida (Simpson 1948, p. 12,
fig. 3 b).
(2) Silver Springs, Florida (Neill 1958, p. 43, pl. 3G).
(3) Site Su2, Suwannee County, Florida (Goggin 1950,
p. 48, fig. 21 N).
(4) Russell Cave, Alabama (Miller 1956, p. 548, lower
(5) Eva Site, Benton County, Tennessee (Lewis and
Kneberg n. d. p. 125).
(6) Shoop Site, near Harrisburg, Pennsylvania(Witthoft
1952, p. 469, pl. 1, no. 13).
(?) Meserve Site, Hall County, Nebraska (Wormington
1957, p. 114, fig. 37),
(8) Clovis, New Mexico (Wormington 1957, p. 45,
fig. 12, left).

Temporal Range: Because of its typological affiliations
and its stratigraphic position at the Hornsby-Darby
site, the temporal range a this type seems to rnm from
Palee- Inftdian to Earlly Archaic..

Retmars Points of this type are tharacteriaed bTyparal-
letl srmo(Ithed basal edges and a coameawe,, sSMtSed,
thinned base.. Form is sy metrical ard workaasip
is good ..

Speci4mnt A 638 ($.. VIAW g) shows a skiht sismn&ariity
to points attributed to the Diten (Mst) )ategry
found at rahalf Caw., Mssoiaarii ((Worriirtmi S19-57,
p. 64, 1 S, I14) 4na teyio dated %W,, t) 9,,'T) .. P..--
t o.4 nt 43 1.. C..)..

Tf'e point of sditlarity of speionts A 5% %mt A 639


to points from the Red Smoke site near Cambridge,
Nebraska (Wormington 1957, p. 117) is also noted.
The Red Smoke site points, found in association with
fragmentary bison bones, were regarded as Plainview
points reworked in such a way as to give the appear-
ance of Meserve points. It was suggested that they
were resharpened without being removed from the shaft
in which they were hated. Specimens A638 and A639
also suggest this in that, when the points are viewed
withthe tip upward, the left edge of each face is steep-
ened far more than the right edge.

In spite of the similarities noted above, the overall
relationship of the two specimens mentioned is far
closer to Florida's Suwannee point.

The specimens are heavily patinated. Upon breaking
a worked flint chip from the same pit and level as
specimen A 639, it was found that only a central core
of the original chert remained. About two-thirds of
the material had desilicified and was soft.

Provisional Type 9; 1 specimen
Outline: Triangular, long and broad
Cross section: Bi-convex
Edges: Slightly convex
Base: Rounded
Stem: The stem is simply a slight rounded protuber-
ance of the blade proper.
Notches: Indistinct corner notching
Shoulders: Prominent but not barbed
Blade: Isosceles with incurving toward tip
Length: 71 mm,
Breadth: 41 nmn,
Length-breadth index: 1, 7/1
Thickness: 15 nm,
Weight 27/2 gms..

Technimjue: Rather crudely percssion chipped; stem alnd
edge show soene pressure flaking,

Material: Chert
CoSlorS Gray


Texture: Rough
Stratigraphical identification: Crystal River formation

Function: Projectile point, probably spear point

Geographical Range: This specimen was found in the 30"-
36" level of Pit No. 7Z at Hornsby Springs.

A point found in the Chattahoochee River Basin, Florida
(Bullen 1950, p. 123, fig. 11 O0) shows a slight similar-

At Gypsum Cave near Las Vegas, Nevada, points re-
markably similar in shape, size and technique were
found (Wormington 1957, p. 157-160, fig. 53).

The areal range of this type is unknown at present.
It is possible that this particular specimen is an ab-
errant form inthis area and does not constitute a local
type at all.

Temporal Range: There is nothing to indicate any great
degree of antiquity for this type at the Hornsby Springs
site. Sand tempered sherds occur above, below, and
in the same level as this point.


Table 1. Distribution of Lithic Artifacts,
Projectile Points, at the Darby
Springs Sites.

Hornsby Springs pits are indicated
at Darby Spring by Da.

Other than
and Hornsby

by Ho, those

Knife Types

Type 1, Ovoid Biface Knife

36-42 inches


6 inches

Type 2, Ovoid Uniface Knife

17 specimens from Simpson Darby Spring collection

33-36 inches

Ho-109 24-30 inches

Type 3, Large Stemmed Knife

Two specimens fromSimpson Darby Spring collection

Ho-82 12-18 inches
105 0-12

Type 4, Small Biface Knife

Da-Surface collection
Ho-43 24-30 inches
105 24-30

Ho-511 18-24 inches
610 12-18

Ho-521 18-24 inches
521 42-48

Type 5, Flake Knife

One specimen from Simpson Darby Spring collection




Da-Surface collection, 2
1 12-18 inches
1 18-24
1 18-24
Ho-41 0-6
71 0-6
72 30-36
72 54-60
81 7






Type 1


0-6 inches

Da-150 12-18 inches


Type 1, Circular or Ovoid Turtleback Scrapers

Six specimens from Simpson Darby Spring collection

Da-Surface collection, 2 specimens
X 30-36 inches Da-6 6-12 inches
Y 18-24 2 specimens ExOl1 30-36inches
Y 36-42 Ho-71 6-12
Y 42-48 2 specimens 71 53
4 30-36 2 specimens 81 7
5 48-54 511 60-66

Type 2, Uniface End Scrapers

Da- Surface collection,
X 24-30 inches
X 42-48 inches
0 6-12
1 12-18
2 30-36
ExlOl 30-36
ExIOl 36-42
Ho-43 6-12

one specimen





Type 1, Adze-like Choppers

One specimen from Simpson Darby Spring collection

Da- 0

0-6 inches

Ho-71 18-24
Ho-105 0-12

Ho-806 20 inches
806 24

Type 2, Large Irregular Choppers


30-36 inches

Type 3, Hammerstones

Da-Surface collection, 1
X 42-48 inches
O 24-30
O 37
Ho-Surface collection, 1
103 24 inches
107 18-24

Da-4 18-24
5 42-48



6-12 inches

600 24-30


Knife Types

Type 1, Ovoid Biface Knife (pl. VIII a-d); 10 specimens
Outline: Oval
Cross-section: Bi-convex
Edges: Convex
Length: 68-95 mm. Average 77. 2 mm.
Breadth: 40-61 mm. Average 51. 3 mm.
Length-breadth index: 1. 5/1
Thickness: 14-24 mm. Average 20.7 mm.
Weight: 39. 9-105. 5 gms. Average 73. 6 gms.

Technique: Percussion flaked; edges may be pressure

Material: Chert
Color: Light to dark brown
Texture: Smooth to granular
Stratigraphical identification: Crystal River forma-

Function: Probably a multi-purpose tool. This type
couldwell have been used as a hand axe or chopper
as well as a knife.

Geographical Range: The occurrence of this type in the
area covered by the 1951-1952 Edwards-Simpson
archaeological survey is shown on table 1. How-
ever, these range so widely in space and time that
they are consequently of little value in tracing
temporal and areal relationships.

Remarks: The workmanship on this type varies but is
generally quite good. The better examples are well
balanced and fairly symmetrical.

Type 2, Ovoid Uniface Knife (pl. IX a-d); 19 specimens
Outline: Oval or irregular
Cross section: Plano-convex
Edges: Generally convex, steepened
Length: 55-115 mm. Average 95 mm.
Breadth: 35-72 mm. Average 53 mm.


Length-breadth index: 1. 8/1
Thickness: 18-29 mm. Average 24. 0 mm.
Weight: 107. 6-176.9 gins. Average 131.5 gms.

Technique: Percussion flaked, edges may be pressure
flaked; about 3 chips per cm. Appears as unifaced
form but edges of both sides may show pressure

Material: Chert
Color: Light to granular
Texture: Smooth to granular
Stratigraphical identification: Crystal River forrna-

Function: Probably a multi-purpose tool. The larger
specimens of this type could have served as a chop-
per as well as a stemless knife while the smaller
specimens could have been used as scrapers.

C-., raphic ai Range: The range in time and space is so
wide that a definition of temporal and geographical
range is difficult.

Remarks: Workmanship is fairly good; most of the
specimens are well balanced and quite:' syrammnetricaL

Type 3, Large Stemmed Knife (pl. X a-c); 6 specimens
Outline: Triangular, long and narrow
Cross section: Bi-convex
Edges: Straight to convex
Base: Rounded or concave
Stem: Contracting
Notches: Corner notched
Shoulders: Prominent to tapering
Ell ;ditl: Isosceles
Length: 76-163 mmnr. Average 123, (5 m.
Breadth: 36-56 mm. Average 49, 3 mm,
Length-breadth index: 2, 5/1
Thickness: 9-15 mm, Average 12.. 8 mm,
Weight: No atteilmpt was made to take thweight as
allow of the specimens were broken. Some of the


measurements above were estimated in one or
two dimensions.

Technique: Percussion flaked; edges may be pressure
flaked. One specimen has very fine pressure re-
touch along the edges with approximately 8 chips
per cm.

Material: Chert
Color: Gray, brown
Texture: Smooth
Stratigraphical identification: Crystal River forma-

Function: Presumably hafted knives. It is entirely
possible that they may have served as very large
projectile points though a consideration of the length-
breadth index and the relative thinness of the blade
indicate the obvious weakness of such large blade
as a projectile point.

Geographical Range: The occurrence of this type in the
area covered by the 1951-1952 Edwards-Simpson
survey is shown on table 1.

Other occurrences of blades which seemtobetypo-
logically similar:
(1) Addick's Basin, near Houston, Texas (Wheat,
1953, p. 220, Type la, pl. 40a).
(2) The Boone Focus of Central Missouri (Chapman,
1952, p. 140-141, fig. 60 M).
(3) Western Tennessee (Lewis and Kneberg, 1947,
p. 20, pl. 3, large stemmed blades).

Temporal Range: Indefinite. One specimen of this type
occurs immediately above a fibre tempered pottery
level. Probably an Early or Middle Woodland
temporal position.

Remarks: These blades vary from an extremely crude,
assymmetrical form to a rather well marked, well
shaped form.


Type 4, Small Biface Knife (pl. XI d-g); 5 specimens
Outline: Roughly triangular, long and narrow
Cross section: Bi-convex
Edges: Convex
Base: Convex; one specimen as a concave base
Length: 46-84 mm. Average 61.4 mm.
Breadth: 28-34 mm. Average 29. 8 mm.
Length-breadth index: 2. 1/1
Thickness: 7-12 mm.
Weight: 8. 3-14. 0 gms.

Technique: Percussion flaked; edges pressure chipped,
3 to 5 chips per cm.

Material: Chert
Color: Cream, brown
Texture: Smooth
Stratigraphical identification: Crystal River forma-

Function: Probably a knife; may have been hafted

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson
survey is shown in table 1.

Other occurrences of blades which seem typologi-
cally similar:
(1) Western Tennessee (Lewis and Kneberg, 1947,
p. 20, pl. 3; Stemmed Knife Blade and Rectan-
guloid Knife Blade).
(2) Addick's Basin, near Houston, Texas (Wheat,
1953, p. 222, Type 3; pl. 40 h and i).
(3) The Baumer Focus, Pope and Massac counties,
Illinois (Maxwell, 1952, p. 180-184, fig. 96,
no. 9, 3 specimens).

The few locations listed above are inconclusive as
to areal distribution of this type. These blades are
probably sowidespread that thetype lacks any areal
diagnostic quality.

Temporal Range: Temporal distribution is unknown.


Remarks: While workmanship varies it is generally
good. The percussion flaking is good on all of these
blades and three of the five specimens show rather
fine pressure shaping and sharpening.

Type 5, Flake Knife (pl. XII a-d); 19 specimens
Outline: Irregular, roughly triangular or rectangular
Cross section: Plano-convex
Edges: Dependent upon the shape of the blade;may
be steepened.
Length: 46-84 mm. Average 60. 2 mm.
Breadth: 25-49 mm. Average 32. 9 mm.
Length-breadth index: 1. 8/1
Thickness: 5-17 mm. Average 8.0 mm.
Weight: 6.0-49. 1 gms. Average 13. 8 gms.

Technique: Percussion flaked, the bulb of percussion
may be present; edges pressure chipped on some

Material: Chert
Color: Gray, pink, brown, tan
Texture: Smooth
Stratigraphical identification: Crystal River forma-

Function: A cutting tool

Geographical Range: Boththe geographical and temporal
range of this type is so widespread asto make it of
little diagnostic value.

Remarks: These blades seem to have been utilized
percussion flakes which may or may not have had
the edges pressure flaked. The shapes, dependent
as they are uponthepercussion flake, aretherefore
highly irregular.



Type 1 (pl. XIII c); 2 broken specimens
Outline: Long and narrow
Cross section: Oval to round
Edges: Straight
Length: (estimated) 39-45 mm. Average 42. 0 mm.
Breadth: (base) 17-22 mm. Average 19. 5 mm.
Length-breadth index: 2. 2/1
Thickness: 4-5 mm. Average 4. 5 mm.
Weight: 1. 5-3. 5 gms. Average 2. 5 gms.

Technique: A pressure retouched percussion flake;
3 chips per cm.

Material: Chert
Color: Light to dark brown
Texture: Smooth
Stratigraphical identification: Crystal River forma-

Function: Presumably a hafted drill

Geographical Range: Both of these specimens occur at
the Hornsby Springs site. One was found in Pit 82
in the 0-6 inch level; the other was found in Pit 150
in the 12-18 inch level.

Some other occurrences of typologically similar
drills are found at:
(1) Stalling's Island, Georgia (Fairbanks, 1942,
p. 226, fig. 23, nos. 1-7).
(2) Poverty Point, Louisiana (Webb, 1948, p. 229,
fig. 44, nos. 8-11).
(3) Addick's Basin, near Houston, Texas (Wheat,
1953, pl. 44 C and F).
(4) Western Tennessee (Lewis and Kneberg, 1947,
p. 22, pl. 5).
(5) The Faulkner site, Massac County, Illinois
(MacNeish, 1948, p. 237, fig. 47, nos. 24-27).
(6) Lamoka Lake, New York, and fromthe Brewer-
ton Focus of New York (MacNeish, 1952, p. 47,


fig. 12 D and F, and 14 J and K).

Remarks: Once introduced, thistype continues to occur
until historic times. Therefore, on a typological
basis they are not diagnostic of a specific temporal
position. Their areal distribution seems to have
been wide in the East.


Type 1, Circular or Ovoid Turtleback Scrapers (pl. XIII g,
h); 23 specimens

Outline: Circular or oval
Cross Section: Plano-convex, "turtleback"
Edges: Curved and steepened
Length: 37-73 mm. Average 49. 1 mm.
Breadth: 25-54 mm. Average 36. 9 mm.
Length-breadth index: 1. 3/1
Thickness: 10-29 mm. Average 15.4 mm.
Weight: 9. 7-59. 9 gms. Average 24. 6 gins.

Technique: Percussion flaked; edges frequently show
fine, secondary pressure flaking with 3-4 chips per
cm. This is generally a unifaceform but one speci-
men has percussion flaked edges on both faces.

Material: Chert
Color: Brown, tan
Texture: Smooth
Stratigraphical identification: Crystal River forma-

Function: Probably a scraper but may have also been
used as a knife

Geographical Range: Both the spatial and temporal range
of this type is so widespread that its typological
significance is of doubtful importance.

Remarks: A very common and numerous type; the work-
manship is, on the average, fairly good and the
proportion and balance is good.


Type Z, Uniface End Scrapers (pl. XIII d-f); 16 specimens
Outline: Irregularly oval
Cross section: Plano-convex
Edges: Curved
Length: 32-47 mm. Average 38. 6 mm.
Breadth: 21-39 mm. Average 30. 0 mm.
Length-breadth index: 1. 3/1
Thickness: 4-10 mm. Average 6. 8 mm.
Weight: 3. 1-9. 6 gms. Average 6. 7 gms.

Technique: Percussionflaked with edges of some speci-
mens showing secondary pressure flaking, up to 6
chips per cm.

Material: Chert
Color: Light to dark brown
Texture: Smooth
Stratigraphical identification: Crystal River forma-

Function: Probably a small scraper or blade

Geographical Range: Table 1 shows the distribution of
this type in the area covered by the 1951-1952
Edwards-Simpson archaeological survey.

Some other apparently typologically similar mate-
rials occur in:
(1) Jefferson County, Florida (Allen, 1954; Scrapers
Type 6, pl. XXI A-C).
(2) Site Su2, Suwannee County, Florida (Goggin,
1950, p. 48, fig. 21).
(3) Battery Point Site, Hernando County, Florida
(Bullen and Bullen, 1953, p. 87, fig. 1 C and
(4) Macon Plateau, Georgia (Kelly, 1938, p. 6,
fig. 4, nos. 11-16).
(5) Western Tennessee (Lewis and Kneberg, 1947,
p. 21, pl. 4, flake end scrapers).
(6) Parrish Village Site, Hopkins County, Kentucky
(Webb, 1951a, p. 441, fig. 13 C).
(7) Shoop Site, near Harrisburg, Pennsylvania
(Witthoft, 1952, p. 477, pl. 3, nos. 20-29;
p. 480, pl. 4, no. 29).


(8) The Bull Brook Site, Ipswich, Massachusetts
(Wormington, 1957, p. 77, fig. 25, top three
(9) Claypool Site, Otis, Colorado (Wormington,
1957, p. 131, fig. 42, center row, first 3 speci-

(10) Lindenmeir Site, near Fort Collins, Colorado
(Wormington, 1957, p. 36, fig. 11, top right).

Temporal Range: Although this type appears to be quite
early, occasional occurrence in late horizons makes
dating on a typological basis alone uncertain. The
stratigraphic position at the Hornsby-Darby sites
where it underlies plain fiber-tempered pottery
levels would suggest an Archaic dating at the latest.

Remarks: These flake scrapers are so numerous at
these sites that only 16 of the specimens were meas-
ured for typing. The flakes were thrown off from
a core by percussion and shaped into an ovoid form.
The thick endwas then made snubnosed by pressure
fracture. The pressure flaking is generally only
at one end and the rest of the flake is left unworked.
Occasionally the bulb of percussion remains. Some
of the specimens are quite thin and might well be
classed as "Thumbnail" scrapers. There is no
evidence of hafting.


Type 1, Adze-like Choppers (pl. XIV a, b; XV a-d); 6 speci-

Outline: Rectangular to oval
Cross section: Plano-convex
Edges: Curved and usually steepened
Length: 69-124 mm. Average 88. 0 mm.
Breadth: 37-74 mm. Average 49. 3 mm.
Length-breadth index: 1. 8/1
Thickness: 15-42 mm. Average 24.2 mm.
Weight: 41.4 297. 6 grns. Average 106. 7 gms.


Technique: Percussion flaked; one specimen shows
pressure flaking along the edges about 3 chips per

Material: Chert
Color: Gray, brown
Texture: Granular to smooth
Stratigraphical identification: Crystal River forma-

Function: Probably a hand axe or celt, may possibly
have been hafted.

Geographical Range: Table 1 shows the distribution of
this type in the area covered by the 1951-1952
Edwards-Simpson archaeological survey.

Some other occurrences noted are at:
(1) The Battery Point Site, Hernando County, Flor-
ida (Bullen and Bullen, 1953, p. 87, fig. 1L-N.
Coates, 1955, p. 28, fig. 1 K).
(2) Silver Springs, Florida (Neill, 1958, p. 41,
pl. 2 M).
(3) Macon Plateau, Georgia (Kelly, 1938, p. 5,
fig. 3, no. 7).

Temporal Range: There is no indication of any definite
temporal position for this type. Neither strati-
graphy, associated artifacts, nor typologically
similar materials enable us to tie this type down

Remarks: There are large pieces of generally good
percussion workmanship.

Type 2, Large Irregular Choppers (pl. XIV c); 3 specimens
Outline: Oval to irregular
Cross section: Plano-convex
Edges: Curved and steepened
Length: 121-140 mm. Average 129.0 rmm.
Breadth: 99-118 mm. Average 105.3 mm.
Length-breadth index: 1. 2/1
Thickness: 51-40 mm. Average 46. 7 mm.
Weight: 449. 5-515.4 gms. Average 474.5 gms.


Technique: Percussion flaking

Material: Chert
Color: Gray, brown
Texture: Smooth to granular
Stratigraphical identification: Crystal River forma-

Function: Probably a large hand chopper

Geographical Range: The distribution of this type in the
area covered by the 1951-1952 Edwards-Simpson
survey is shown on table 1.

Areal and temporal distribution of material of this
type is so widespread that, as a type, it is of little
diagnostic value.

Remarks: These specimens seem to be nothing more
thanlarge pieces of chert which are roughly battered
into shape and given an edge to serve as a heavy,
crude chopper.

Type 3, Hammerstones (pl. XVI); 11 specimens
Form: (1) Roughly spherical or ovoid
(2) Probably a battered core
Length: 62-97 mm. Average 77.4 mm.
Breadth: 48-74 mm. Average 63.4 mm.
Thickness: 43-60 mm. Average 50.6 mm.
Length-breadth-thickness ratio: 1. 5 : 1. 2 : 1
Weight: 162. 1 375. 6 gms. Average 223. 8 gms.

Technique: Percussion flaked and pecked

Material: Chert
Color: Cream, tan, brown, dark gray
Texture: Smooth to porous
Stratigraphical identification: Crystal River forma-

Function: Presumably a hammerstone

Geographical Range: The distribution of this type in the


area covered by the 1951-1952 Edwards-Simpson
survey is shown on table 1.

Hammerstones are material traits of all cultures
making lithic artifacts and, as such, are of little
temporal or areal diagnostic value.

Remarks: These hammerstones do not show any delib-
erate shaping but are characterized by roughening
and abrasion marks on the striking surface. Some
seem to be remnants of cores from which flakes
had been struck for other artifacts.



a Projectile Points, Type 1. Bullen (1958) calls this a
Bolen point. Hornsby Springs, Pit # 511, 6-12 inch

b Projectile Point, Type 1. Darby Spring, Pit # 1, 30-
36 inch level.

c Projectile Point, Type 2. Also could be classed as a
Bolen point. Hornsby Springs, Pit # 110, 30-36 inch

d Projectile Point, Type 2. Could be classed as a Bolen
point. Provenience unknown.

e Projectile Point, Type 2. Hornsby Springs, Pit #512,
42 inches.

f Projectile Point, Type 3. Butler site (see area map,
p. 2), 12-18 inch level.

g Projectile Point, Type 3. Butler site, 32 inches.

h Projectile Point, Type 3. Butler site, 42-48 inches.


b c




- Projectile Point,

- Projectile Point,

- Projectile Point,
12-18 inch level.

- Projectile Point,
0-6 inch level.

- Projectile Point,
24-36 inch level.


Type 4.

Type 4.

Type 5.

Type 5.

Type 5.


Meander site, 12-18

Meander site, 18-24

Hornsby Springs, Pit

Hornsby Springs, Pit #

Hornsby Springs, Pit #



# 81,





. \.::-:.- *^1.
* ;~. -1



a Projectile Point, Type 6. Butler site, 36-42 inch

b Projectile Point, Type 6. Hornsby Springs, Pit # 91,
23 inches.

c Projectile Point, Type 7. Hornsby Springs, Pit # 51,

d Projectile Point, Type 7. Hornsby Springs, Pit # 71,
12-18 inch level.

e Projectile Point, Type 8. Similar to Suwannee point.
Hornsby Springs, Pit # 110, 18 inches.

f Projectile Point, Type 8. Similar to Suwannee point.
Butler site, 54 inches in situ.

g Projectile Point, Type 8. Similar to Meserve point.
(Wormington, 1957, p. 114.) Darby Spring, Pit# 1,
45 inches.

h Projectile Point, Type 8. Darby Spring, Pit # 0, 54





a Knife, Type 1. HornsbySprings, Pit#600, 45 inches.

b Knife, Type 1. Hornsby Springs, Pit # 806, 12-18

c Knife, Type 1. Hornsby Springs, Pit # 557, 6 inches.

d Knife, Type 1. Hornsby Springs, Pit # 806, 36-42
inch level.