Lithic Technology and Technological Organization: Results of Lithic Analysis

337Chapter 10: Results of Lithic Analysis

In: Prehistoric Archaeological Investigations in the Applewhite Reservoir Project Area, Bexar County, Texas, 2008,

edited by David L. Carlson, Patricia A. Clabaugh, Rolfe D. Mandel, and Charlotte D. Pevny, pp. 337–376. Reports of

Investigations No. 7. Center for Ecological Archaeology, Texas A&M University, College Station, Texas.

337

10

This chapter details the results of technological, ty-

pological, and functional analyses of the flaked and

groundstone lithic assemblages recovered during

survey and testing of the Applewhite Reservoir

project area (see Appendix C for analytical and de-

scriptive data). The research orientation for the lithic

analysis closely follows the research design for the

survey and testing of sites to be impacted by the

Applewhite Reservoir (Chapter 4).

The lithic analysis for the Applewhite Reservoir

project area was designed to provide useful data

pertinent to the technology, form, and function of

lithic tools recovered. The data are useful for ad-

dressing long-term aspects of temporal change in

land-use systems within the lower Medina River basin

in south Texas (1992:1). Of particular interest is any

evidence for land-use intensification or other changes

from the Late Pleistocene to Holocene periods and

why these changes occurred.

Technological Organization

and Provisioning

Of particular importance in the determination of

land-use strategies is the character of technological

organization and the nature of provisioning (Kuhn

1990 , 1992, 1995). To pursue a living in any given

landscape it is necessary to establish and maintain a

suitable source of raw materials and a manufactur-

ing site for tools (Kuhn 1995:21). One of the pri-

mary objectives of any technological system is to

provide tools and materials at specific times and lo-

cations as necessary (Kuhn 1990:70). Earlier re-

search by Binford (1973, 1979, 1989) stressed that

variables such as mobility, resource distribution, and

tool/task relationships are integrated with tactical and

planning depth. The concept of depth includes plan-

ning artifact manufacture, tool transport and main-

tenance, and tactical strategies to assure that tech-

nologically related needs are met (Kuhn 1990:70,

1995:21). Kuhn (1990:69–76, 1992:188–190,

1995:21–23) recently developed a model of techno-

logical provisioning based on three levels of planning

complexity: activities, individuals, and localities. It is

emphasized here that a particular technological sys-

tem may not exhibit all three levels or may reflect

some degree of each level of planning. At times, the

technological system can be dominated by only one

level of planning.

Activities

Activity-level provisioning means that tools are manu-

factured as they are needed and discarded when

the need ceases. Provisioning of activities requires

the least amount of input into planning and is equiva-

lent to Binford’s (1977, 1979) expedient technology.

LITHIC TECHNOLOGY AND TECHNOLOGICAL ORGANIZATION:

RESULTS OF LITHIC ANALYSIS

John E. Dockall and Charlotte D. Pevny

338 Prehistoric Archaeological Investigations in the Applewhite Reservoir Project Area

The level of technological planning can only occur in

areas where suitable raw material is present in suf-

ficient quantity. The time-scheduling limits of activi-

ties also limits the time that can be devoted to tool

manufacture. The nature of activities often means

that formal tools are not necessary for the success-

ful completion of the tasks.

Individuals

Kuhn’s second type of provisioning is at the indi-

vidual level. Equipping individuals with suitable toolkits

or personal gear (Binford 1977) requires tactical and

planning depth (Binford 1989) well beyond provi-

sioning of activities, since people are limited by what

they can transport. Although transportability is a key

factor in limiting toolkit size and variability, the tech-

nology is kept in a ready state and can be immedi-

ately implemented as needed. Transport and toolkit

size limit the number of backup and extra repair parts

that can be carried. Continual demands on personal

gear often result in higher attrition rates for both

extractive and maintenance portions of the toolkit.

The personal toolkit can be furnished with special-

ized tools, generalized tool forms, or raw material in

the form of cores and tool blanks; the composition

depends upon anticipated needs (Henry 1995:111).

Places

Behaviors associated with provisioning of places are

also related to anticipation of future needs. To suc-

cessfully supply locations with needed material, there

must be some advance information regarding the

future location and scheduling of activities. Strate-

gies that rely on furnishing places with tools and raw

materials decrease the restrictions of transportabil-

ity or time required for manufacture as activity or

individual provisioning. Kuhn (1992:189) predicted

that there should be an emphasis on raw materials

as opposed to tools in the provisioning of places. At

provisioned places, there may be less emphasis on

tool maintenance or resharpening, and it may be just

as easy to retool or replace the tool. This may be

reflected in the archaeological assemblage by the

numbers of minimally worn or damaged tools that

could otherwise be repaired (Kuhn 1990:79–80).

Previous Studies of Lithic Technology

in the Western Gulf Coastal Plains

and Rio Grande Plains Transition

The region encompassed by the Applewhite Reser-

voir project area is dominated by the intersection of

two major physiographic zones: the Western Gulf

Coastal Plains and the Rio Grande Plains. Also

present is the Balcones Escarpment to the north

(Chapter 2). Because the area is dominated by a

broad expanse created by the intersection of these

physiographic provinces, there may be identifiable

region-specific prehistoric adaptations that are char-

acteristic of central or south Texas. Room does not

permit an exhaustive review of all studies of lithic

technology for the area; therefore, only major site-

and region-specific references will be noted, but the

reader is encouraged to consult these references.

Significant studies of lithic technology on the Rio

Grande Plains and Western Gulf Coastal Plain in-

clude the following: Hester (1978) at Chaparrosa

Ranch, Fox et al. (1974) at the Cuero 1 Reservoir in

Dewitt and Gonzales Counties, Hall et al. (1986) and

Thoms et al. (1981) at Choke Canyon, Dockall (1997)

at Laughlin Air Force Base (AFB) in Val Verde

County, Lukowski (1987) within the Leona River

Watershed, Uvalde County, and Taylor and Highley

(1995) at Loma Sandia. Each of these studies pro-

vides detailed discussion of the technological, mor-

phological, and functional aspects of lithic technol-

ogy in these regions as they relate to settlement, sub-

sistence, mobility, and resource use.

Raw Material Procurement:

Sources and Strategies

Sources

The abundance of raw materials is an important as-

pect of the study area that is emphasized in each of

the investigations detailed in this chapter. The main

source of raw material is the Uvalde gravels. The

Uvalde gravels represent an extensive source of

chert gravels for prehistoric inhabitants in south and


southwest Texas (Banks 1990:56–57; Byrd 1971).

Geologists have long known of the presence of these

gravels, and there is still considerable controversy

over which deposits may or may not be Uvalde grav-

els (Banks 1990; Byrd 1971; Caran 1992). The term

Uvalde gravels is considered herein to apply to up-

land relict gravels capping drainage divides of the

Nueces, San Antonio, and Medina rivers (Caran

1992; Dickens 1995:180; Hill 1891:368). Chert gravels

occurring in central Texas alluvial terrace deposits

were erroneously associated with Uvalde gravels

(Banks 1990; Byrd 1971; Dockall 1995). Caran noted

that researchers following Hill (1891) have misap-

plied the term Uvalde gravels to almost every known

exposure of upland gravel and terrace fills. The most

serious problem concerns the use of a formation

name in reference to unstratified lag gravels thought

to be derived from that unit.

Although these gravels are considered to be

abundant in the project area, there is some sugges-

tion that local differences in their distribution and

character influenced the organization of lithic tech-

nology. The nature of the deposits also seems to have

governed to some extent the methods of core re-

duction and tool manufacture between areas

(Dockall 1997; Fox et al. 1974; Hester 1978; Thoms

et al. 1981).

Despite the confusion associated with Uvalde

gravels, the probable initial source area was described

by Byrd (1971:20). The Ogallala Formation on the

Llano Estacado, and ultimately the southern portion

of the Rocky Mountains, provided the source mate-

rial for Uvalde gravels. The Ogallala Formation rep-

resents the remnants of an alluvial plain mantled with

gravels derived from the eastern ranges of the south-

ern Rocky Mountains in Central New Mexico:

Sangre de Cristo, Sandia, Manzano, Los Pinos,

Gallinas, and Jicarilla. Streams that drained eastward

from these ranges deposited sediment to create the

Llano Estacado of the Southern High Plains. The

exact era to which Uvalde gravels date is also in

question, but the best estimate is somewhere be-

tween the Miocene and Pliocene (Byrd 1971:29).

There are several other geological formations in

the area that may have provided occasionally avail-

able sources of lithic material, including the Queen

City sand, Reklaw Formation, Carrizo sand, Wilcox

Group, and Weches Formation, all of which date to

the Eocene era. These sources probably provided

various sandstones for hearth feature construction

and the manufacture of various types of groundstone

implements. Hematite or red ocher were procured

for a variety of decorative purposes.

Strategies

The abundance of raw material in the study area

suggests that special methods of procurement, such

as quarrying, were either absent or very rare. No

archaeological evidence of formal prehistoric chert

quarrying activity has been documented in the re-

gion. The nature of raw materials negated the need

to employ digging or quarrying. Trade for suitable

material may have occurred but does not seem to

have been a major mode of procurement. Procure-

ment strategies were logistically integrated into the

settlement/subsistence system.

Encounter strategies are often employed by

groups that are quite flexible in terms of residential

location and group size. Groups move freely across

the landscape to take advantage of particular re-

sources that may be widely scattered in a given re-

gion, and materials are procured as needs are iden-

tified (Haurie 1995:26).

Embedded procurement strategies are often as-

sociated with hunter-gatherer groups practicing a

seasonal settlement pattern. Extended periods of time

are spent at specific residential sites. In addition, a

number of specific task groups frequently occupy

smaller special-purpose sites in the area. Embedded

strategies enable a people to schedule lithic resource

procurement within other subsistence-related tasks

such as hunting or plant procurement, and also serve

to minimize labor input (Haurie 1995:26–27).

Embedded and encounter strategies can be col-

lectively considered as expedient methods. Upland

lag gravels and stream gravels were typically ex-

ploited by these methods. There are certain prob-

lems associated with procuring raw material from

these types of sources (Haurie 1995:29), including


poorer quality, variable size, brittleness, and internal

fractures or cavities. These problems were dealt with

by intensive nodule testing.

Previous Models

Previous models of lithic technology have empha-

sized the dichotomy between lithic material procure-

ment and other subsistence-related activities involv-

ing lithic technology. The approach employed by

Hester (1975, 1978) for Chaparrosa Ranch was to

consider lithic technology as only part of the com-

plete adaptive strategies of people living in the area.

Hester’s approach also is related to the linear sys-

tems model described by Collins (1975), in which

stone tool manufacture is a continuum best consid-

ered as a series of stages or phases. Hester modi-

fied Collins’ model of lithic reduction and considered

three phases related to the manufacture of stone

tools. At the time, Hester’s (1978) research on the

Chaparrosa collections focused on providing infor-

mation pertaining to the use-reworking and discard

of stone tools.

The three phases of Hester’s model are pre-

sented here. Phase I includes selection and procure-

ment of raw material. Phase II includes activities

related to workshop tasks associated with initial

manufacture. Phase III represents the final stages

of shaping and trimming of lithic materials into fin-

ished tools. The discrete location of activities related

to each phase is significant in relation to this model.

Phase I activities included procuring, quarrying, and

testing raw materials to obtain suitable stone for tool

manufacture, whether it be in the form of cores or

flakes. Phase I and II activities related to the work-

shop were performed on terraces where abundant

raw material occurred in the form of surface cobbles.

Phase III included shaping, secondary trimming, and

last stages of tool manufacture; these tasks typically

occurred away from terrace sites on the floodplains

of streams in the area.

Each lithic phase or type of site in the model is

characterized by particular types of lithic debris that

can be used to interpret the kinds of lithic activities

which occurred. Cortical flakes are prevalent in

workshop sites, while interior flakes are rare. Inte-

rior flakes are predominant in occupation sites be-

cause the majority of the cortex was removed at the

workshop. Thinning flakes, large percussion flakes

(as tool blanks), and resharpening flakes also are

present at occupation sites (Hester 1978:28).

Associated with workshop activities are three

principal methods of flake production or tool manu-

facture. The first includes all prepared or unprepared

core techniques to produce flakes for tools. Core

types include bifacial or discoid cores, polyhedral

cores, and split cobbles. The second method is a

bifacial strategy and is considered to be a core-tool

or cobble-tool industry (Hester 1978:27). The third

technique is the blade core and blade. Hester notes

occasional blades, blade fragments, and tools manu-

factured from blade blanks, but blade cores have

rarely been recovered. Hester attributes blades and

blade cores to a particular Late Prehistoric adapta-

tion in the area. The nature of raw material distribu-

tion on Chaparrosa Ranch indicates that major stream

valleys are flanked by gravel-mantled terraces

where the raw materials are not present on the allu-

vial floodplains (Hester 1978:24–25). This type of

distribution pattern could account for much of the

observed staging in lithic manufacture and the sepa-

ration between procurement/manufacture and use/

discard.

Thoms et al. (1981:236–240) tested and evalu-

ated Hester’s (1978) model of lithic technology dur-

ing subsequent research at Choke Canyon. Data from

a series of sites with upland terrace and floodplain

settings were employed to test the spatial dichotomy

of workshop activities and those associated with food

processing and consumption. Evidence of both types

of activities was frequently found at different types

of sites, indicating that site function was to some

degree multipurpose. Initial reduction was not con-

fined to terraces, and tool finishing and maintenance

were not limited to floodplain settings (Thoms et al.

1981:237). One of the major reasons for this pat-

terning was the generally good distribution of lithic

raw material over the landscape in both floodplain

and terrace settings. A more uniform distribution of

raw material across the landscape altered the pat-

terns of raw material procurement, manufacture, and

use in a way that is clearly different from the di-

chotomy observed by Hester (1978) at Chaparrosa

Ranch.


At Cuero I Reservoir, Fox et al. (1974:12) noted

that siliceous gravels cap the lower slopes of upland

terraces. Gravel deposits are exposed at times along

portions of a fossil floodplain. In addition, there are

extensive deposits of these gravels mantling the

Catahoula and Oakville sandstone formations in some

localities. A general similarity was noted among all

excavated sites regarding the lithic technology: abun-

dant quantities of debitage, similar raw material types,

and similar tool forms (Fox et al. 1974:200). The

highest abundance of cores was noted on the el-

evated gravel deposits that seem to have served as

procurement zones. Raw material was obtained from

both river gravel bars and gravel-mantled terraces,

and much of the initial reduction was occurring prior

to the transport of material into the habitation sites

(Fox et al. 1974:202).

The nature of raw material distribution at

Laughlin AFB was considered to be fairly uniform

along all strath terraces on the eastern side of

Sacatosa Creek and only within a few hundred meters

of sites located on top of these terraces and along

the creek. The net result (Dockall 1997) was an

overlap between procurement/manufacture and use/

discard similar to that observed by Thoms et al. (1981)

for Choke Canyon. Even though chert is not abun-

dant along the floodplain of Sacatosa Creek, the dis-

tance to accessible sources is negligible in terms of

the spatial separation of lithic procurement and tool

use activities. The gravel-mantled strath terraces

above Sacatosa Creek served as zones of lithic ma-

terial procurement, initial reduction, and non-lithic

resource procurement as judged by the presence of

scattered fire-cracked rock (FCR).

The areal distribution and size/quality of Uvalde

gravels in south and southwest Texas influenced the

nature of raw material procurement and reduction.

A more even distribution of Uvalde gravels in areas

such as Choke Canyon, Laughlin AFB, and Cuero I

Reservoir resulted in a greater technological and ty-

pological overlap between sites considered as pri-

marily procurement related and those considered as

primarily subsistence related. Chaparrosa Ranch is

considered to have a relatively even resource distri-

bution in the uplands but virtually no raw material

within the floodplain zone. This created a greater

observed technological difference between lithic pro-

curement- and subsistence-related sites.

Uvalde Gravel Procurement Methods

Previous studies indicate that there were specific

regional strategies of lithic resource procurement that

varied with the areal distribution of surface gravels.

Certainly, environmental and non-lithic resource dis-

tribution characteristics also influenced these lithic

strategies to some degree. Based on the character

of lithic assemblages at lithic procurement or quarry

sites situated within access of Uvalde gravel sources,

raw materials were procured in several ways. First,

specific direct procurement trips were made to a

source. Second, procurement of chert gravels was

organized or embedded within other subsistence- or

settlement-related activities. The nature of chert

gravels usually dictates that at least a portion of the

material will be poor in quality, have internal frac-

tures, or be of variable size and shape. All of these

factors influence, to some degree, the methods of

procurement and reduction. Nodule testing was one

of the most common methods of material selection.

Size and shape were other characteristics used to

determine the suitability of raw materials for further

modification. For instance, it may be easier to cre-

ate discoidal or bifacial cores from flat, round, or

elongated cobbles, whereas generalized flake cores

can be produced from a wider variety of cobble

morphologies. An upper and lower size limit also

guided cobble selection. Data on this trend can be

obtained from maximum measurements of tested or

minimally reduced cobbles.

The procurement of Uvalde gravels, or any sur-

face gravel source, implies a low level of energy/

time investment that is distinct from the procure-

ment of ledge cherts and buried, bedded, or gravel

sources. Archaeologists in south and southwest Texas

have misapplied the term “quarry” in reference to

areas of surface gravels with abundant evidence of

raw material procurement. In accord with Haurie

(1995:28), quarrying is used to define the labor-in-

tensive excavation of tunnels and surface pits to pro-

cure raw material. Other quarry-related features and

activities include the use of heat spalling, wedges,

and picks. Quarries are identified by pits, tunnels,


and surface spoil from excavation, in addition to as-

sociated workshop areas, and reflect the intensive

exploitation of desired resources. These features

have not been observed at any sites in south and

southwest Texas associated with Uvalde gravels.

Haurie (1995:29) associated gravel sources with

expedient procurement and use of an encounter strat-

egy of acquisition. Also, raw materials procured from

gravels are typically of lower quality than primary

source materials and require more testing and se-

lection to acquire suitable materials.

Uvalde Gravel Distribution in the Project Area

Unfortunately, the Uvalde gravels have not been

extensively mapped in any portion of south or south-

west Texas. Therefore, our understanding of the in-

fluence of their distribution on procurement and re-

duction strategies remains at a general level (as dis-

cussed above). Previous work within the study area

(McGraw and Hindes 1987) identified several sites

that were considered to be lithic procurement or

workshop areas. These sites are briefly discussed

to provide some idea of their nature in the project

area and the character of Uvalde gravel distribu-

tion. Sites considered include 41BX464, 41BX517,

and 41BX569. A number of other sites were recorded

as lithic workshops or lithic reduction areas charac-

terized by thin to moderate scatters of cortical and

non-cortical debitage and cores, which suggests that

raw material was commonly available throughout the

study area, even if on a sporadic basis.

Site 41BX464 is considered to be a sparsely

occupied lithic workshop situated on an upland ridge

east of Medio Creek. The western slope has deep

gullies with abundant chert cobbles. The site proper

has a significant amount of primary, secondary, and

interior debitage and occasional cores (McGraw and

Hindes 1987:434).

Due to the presence of abundant chert cobbles

and the character of the lithic assemblage, site

41BX517 was characterized as a lithic workshop

and quarry. The lithic debris consisted of several

cores and core fragments, a scatter of cortical and

non-cortical debitage, and an absence of features.

Chert cobbles were noted to occur as a distinct layer

of gravel and cobbles at the base of old alluvial de-

posits. The occurrence is similar to others identified

within the Medina River-San Antonio River drain-

age system (McGraw and Hindes 1987:147).

Site 41BX569 was a small, temporary occupa-

tion site and quarry/workshop locality. The site is

characterized by calcareous gravels and cobbles

exposed to depths of nearly 2 m below the surface.

The chert cobbles are typically larger than 12 cm in

maximum diameter and are medium to fine grained

in quality. Within a bulldozed area of the site, chert

cobbles were exposed to depths of 2.5–3.5 m with

cobble size increasing with depth. The lithic assem-

blage included several tested cobbles and core frag-

ments from within the bulldozed area. Other sur-

face indications included the range of debitage as-

sociated with material procurement and initial re-

duction, but there were no diagnostic or finished tools

recovered during testing and monitoring. The sig-

nificance of this site lies in the nature of the chert

occurrence, which is reported to be distinct from

other localities such as river gravels or Uvalde grav-

els exposed as a thin mantle on eroded hilltops

(McGraw and Hindes 1987:213).

These sites illustrate how raw material procure-

ment of Uvalde gravels and other gravel deposits

was associated with direct acquisition from the

Medina and San Antonio rivers, as well as alluvial

deposits containing gravels. No sites were identified

as lithic procurement localities during the TAMU

survey and testing of sites within the Applewhite

Reservoir project area. Site 41BX568 was recom-

mended for testing by TAMU because of its loca-

tion near the confluence of the Medina River and

Medio Creek, the presence of chert gravels, and a

cumlic 2Ab-horizon. Testing failed to identify an in-

tact buried cultural component and no further test-

ing was conducted. No indication was present that

the site was used as a lithic procurement locality, but

the site was heavily disturbed at an earlier time.

Other Raw Materials

The second most important lithic material procured

within the project area was probably sandstone,

which was used to make a variety of tools including

hearthstones, groundstone implements, and the oc-

casional chipped stone tool. The most likely source


for the fine-grained sandstones in the project area is

the Wilcox Group, as was discussed earlier. The sand-

stones of the Wilcox Group are medium to fine

grained and vary in color from yellow-brown to red-

dish brown. There is a significant color change as-

sociated with heating this sandstone. Although there

are numerous surface outcrops of varying size, it is

possible that blocks of this sandstone, used for grinding

implements and hearthstones, may have occasion-

ally been quarried. Surface outcrops of sandstone

were noted at sites 41BX669, 41BX793, and

41BX858. It is possible that specific outcrops of sand-

stone were occasionally used as source material for

a variety of purposes within the study area. On oc-

casion, limestone cobbles and fragments were uti-

lized as cobble tools, expedient groundstone imple-

ments, and hearthstones. Limestone cobbles could

be procured from Uvalde gravel deposits or gravels

within the Medina River and other area drainages.

Artifact Descriptions

Core Reduction and Typology

Patterns of core reduction can be employed to de-

termine the degree of standardization in the reduc-

tion process and the types of flakes that were being

produced for tools. The nature of core reduction is

also related to patterns of raw material procurement

and the type and mode of raw material occurrence.

Previous studies of core reduction from south and

southwest Texas have indicated that there were three

principal reduction strategies employed (Hester

1978:25–27). These same core types were identi-

fied at Laughlin AFB (Dockall 1997:26–29), Choke

Canyon (Hall et al. 1986; Thoms et al. 1981), and

Cuero I Reservoir (Fox et al. 1974). According to

Hester (1978:25), the first type was oriented toward

flake production and included both prepared and

unprepared core types. Cores used to produce flakes

were of several forms: split cobbles to produce coni-

cal or polyhedral cores, multifaceted cores, discoid

cores, and unprepared random cores. Implements

that were manufactured include unmodified flakes

used as tools, a variety of minimally modified flake

tools, scrapers, other unifaces, and unifacial cobble

tools.

A second strategy included the manufacture of

various tools using bifacial reduction of cores and

flakes. Bifacial implements included a variety of pro-

jectile points, knives, large scrapers, and chopping

or cobble tools (Hester 1978:27). Bifacial reduction

of cobbles necessitated the selection of thin, elon-

gate, and tabular cobbles. Bifacial reduction strat-

egy is typically associated with production of vari-

ous types of formal tools. Bifacial implements are

also frequently curated for longer periods of time

than flake tools produced via a flake core strategy.

The third type of core reduction strategy was

identified as a blade core technology (Hester

1978:27–28), which was associated with the Late

Prehistoric and Protohistoric periods. At the time of

his Chaparrosa Ranch study, Hester (1978:27) con-

sidered the blades and blade cores to be primarily a

Late Prehistoric phenomenon. Recent research at

site 41BX831 (Thoms et al. 1996:18) identified a

conical blade core strategy and blades associated

with the upper Perez paleosol indicative of the early

Holocene period at Blocks H and T (ca. 8700 B.P.).

These blocks have also produced what are probably

the largest stratigraphically in situ cultural materials

associated with Angostura points in North America.

A similar type of low conical blade core also was

identified by Ken Brown from in situ deposits adja-

cent to Feature 6 in deposits at Berger Bluff (Ken

Brown, personal communication 1996).

Core Reduction Strategy (N = 50)

A total of 50 cores and core fragments was recov-

ered during TAMU survey and testing phases. The

core assemblage was further subdivided into spe-

cific core types based on flake scar orientation,

method of flaking, preparation, and morphology of

flakes removed from the core surface. Types in-

cluded tested material, partial cobble/pebble cores,

angular cores (expedient), blade cores, discoid cores,

flake or macroflake cores, and core fragments (Fig-

ures 10.1, 10.2). These core types are similar to types

identified at Laughlin AFB, Chaparrosa Ranch,

Choke Canyon, and the Leona River Watershed

(Lukowski 1987). Resemblances between core types

in these areas suggest a broadly similar pattern of

Uvalde gravel procurement and overall strategies of

core reduction. In large part, procurement and re-


Figure 10.1. Selected cores from the project area: (a) cobble/pebble core; (b) discoid core; and (c) generalized core.


Figure 10.2. Selected cores recovered from the project area: (a–b) blade cores; (c) macroflake core; and (d) microflake

core.

duction of Uvalde gravels was oriented toward pro-

duction of generalized tools, including unmodified,

but utilized flakes and a variety of unifacial types.

The manufacture of bifacial artifacts represents a

logistical strategy of tool production well in advance

of use and, by extension, represents tool forms with

an extended use life, unlike simpler flake tools

(Hayden et al. 1996). In part, the use of Uvalde grav-

els was associated with conservation of bifacial ar-

tifacts manufactured from non-local materials

(Dockall 1997:61; Johnson 1986:136). The variabil-

ity observed in core types of Uvalde gravels repre-

sents technological modifications of the generalized

core strategy to meet specific needs such as flakes

of specified shape, size, or thickness for particular

tools or tasks (Dockall 1997:5–8; Hayden et al. 1996;

Thoms et al. 1996:18). Each artifact is represented

by a catalog number (Cat. No.) listed with its corre-

sponding site.

Tested Material. (N = 3) This category repre-

sents selecting, testing, and discarding raw material


during procurement or core preparation efforts. Typi-

cally, these cobbles have only one or two small flake

removals to examine material quality, knappability,

and other material properties. Occasionally, an in-

ternal flaw or other reason for discard can be ob-

served. Sites 41BX347 (Cat. No. 1), 41BX460 (Cat.

No. 1), and 41BX793 (Cat. No. 77) each yielded a

tested cobble. All three specimens are elongated and

slightly tabular cobbles that have not been heat treated.

Cobble/Pebble Cores. (N = 1; Figure 10.1a)

This core category includes only minimally reduced

cobbles and pebbles that retain much of their exte-

rior cortex; it is possible to calculate their original

size and shape. Cobble/pebble cores should not be

confused with tested cores, which often exhibit only

one or two flake scars. One method of distinguish-

ing cobble/pebble cores is the absence of potential

reasons for abandonment like internal flaws, cracks,

etc., usually noted for tested cobble/pebble cores.

A single specimen was recovered from site

41BX528 (Cat. No. 72). Cortex is present on both

surfaces; two large percussion flakes were removed

from one surface. These flake scars served as a

striking platform for the removal of five large per-

cussion flakes before the core was discarded. The

original cobble shape was ovate to subspherical.

Partial Cobble/Pebble Cores. (N = 21) This

core category is quite broad in definition and includes

any core that retains sufficient cortex and sufficient

size to identify that it was originally a cobble or pebble,

even though the size cannot be determined. The

cores still retain much of the exterior cortex, typi-

cally on multiple surfaces. These cores are a stage

in the continuum from cobble/pebble to decorticated

cores. Included in this category are discoid cores

and generalized cores, which are discussed sepa-

rately below.

Discoid Cores (N = 10; Figure 10.1b) Discoid

cores vary considerably in size, dependent either upon

initial nodule size or the degree of reduction of indi-

vidual cores. Technologically, discoid cores can be

perceived as one of two distinct surfaces, either (1)

a convex striking surface covered with cortex or

created by flake removal, or (2) a flat flake removal

surface from which flakes are removed in a radial

pattern around the core. The resulting flakes are quite

flat and thin with multidirectional flake scars on the

dorsal surface from previous flake removals. The

periphery of the convex striking surface also has

smaller percussion flake removals representing

preparation of the striking platform. Discoid cores

were also identified at site 41BX831 (Dockall and

Pevny 2007; Thoms et al. 1996:18) in deposits dated

to the early Holocene (8700–7000 B.P.) and at

Laughlin AFB (Dockall 1997:29), from an undated

context along with various Late Paleoindian/Early

Archaic projectile point forms. Flakes removed from

discoid cores were typically useful in mass-produc-

ing expedient flake tool blanks for a variety of tasks

requiring a thin cross section and acute lateral edges.

The reduction of discoid cores also involves an ef-

fort to maintain core symmetry and general shape

during the flaking process and is the result of the

need for particular flake morphologies and relative

dimensions.

Recovered discoid cores include one each from

sites 41BX526 (Cat. No. 19), 41BX554 (Cat. No.

15), 41BX669 (Cat. No. 187), 41BX793 (Cat. No.

70), and 41BX833 (Cat. No. 2), two from site

41BX544 (Cat. Nos. 9 and 13), and three from site

41BX539 (Cat. Nos. 3, 6, and 35).

Generalized Cores (N = 11; Figure 10.1c) This

category includes all cores having some cortex on

one surface and randomly oriented flake scars that

were not removed from a specific or specially pre-

pared platform. The initial size of the cobble or pebble

cannot be determined. Often these cores are angu-

lar and have a variable number of striking platforms.

The appearance of these cores can occasionally be

confused with tested nodules because some do not

appear to be completely exhausted and may only

have a few flakes removed. Discoid cores are usu-

ally more completely reduced than generalized cores.

Flakes produced via generalized cores at site

41BX831 (Dockall and Pevny 2007; Thoms et al.

1996:18) are thicker than flakes removed from dis-

coid cores and have more dorsal cortex and fewer

dorsal flake scars. Striking platforms are typically

cortical or single-faceted, and flake cross sections

range from triangular or irregular to wedge-shaped.

The lateral edges of these flakes often exhibit a range


of edge angles from obtuse to acute and were em-

ployed in a variety of tasks requiring edge angles

suitable for cutting/scraping and chopping/adzing.

Generalized core flakes were selected to create a

variety of unifacially flaked implements such as

notched flakes, denticulates, and beaked tools. A simi-

lar array of tool forms was identified within the vari-

ous Applewhite Reservoir project area lithic assem-

blages recovered during TAMU survey and testing.

Recovered generalized cores include one each

from sites 41BX526 (Cat. No. 20), 41BX544 (Cat.

No. 14), 41BX663 (Cat. No. 1), 41BX863 (Cat. No.

25), 41BX865 (Cat. No. 3), and 41BX916 (Cat. No.

13), two from site 41BX793 (Cat. Nos. 55 and 142),

and three from site 41BX872 (Cat. Nos. 16, 17 and

481).

Split Cobble Cores. (N = 5) The split cobble

core technique involves halving or quartering a suit-

ably large cobble from which the pieces are used as

individual cores. Some of the discoid cores were ini-

tially portions of split cobbles. Hester (1978:26) il-

lustrates a split cobble stage in the reduction of coni-

cal or unidirectional cores from Chaparrosa Ranch

in Zavala County. Hester also notes that the most

common core form was the simple prepared core

created by cobble splitting (1975:215). In most cases,

the flat facet created by splitting the cobble served

as the striking platform during flake production. De-

pending upon the shape or size of flakes to be re-

moved, the cobble could be split lengthwise or across

the width. The split cobble technique was probably

a fairly common technique in core preparation, but

further reduction eliminates much of the evidence

for recognition. All split cobble cores from the

Applewhite Reservoir project area survey retain

cortex on the exterior of the cobble half. Sites

41BX460 (Cat. No. 2), 41BX528 (Cat. No. 70), and

41BX793 (Cat. No. 159) each yielded one, and two

were found at site 41BX544 (Cat. Nos. 10 and 12).

Decorticated Cores. (N = 8) These cores are

small, angular, generalized cores that have no sur-

face cortex. There are multiple unprepared platforms

and no preferred flaking directions. This category

represents exhausted cores that are inclusive of a

wide variety of morphological types. Recovered

decorticated cores include one each from sites

41BX539 (Cat. No. 9), 41BX568 (Cat. No. 9), and

41BX865 (Cat. No. 2), two from site 41BX670 (Cat.

Nos. 64 and 227), and three from site 41BX916 (Cat.

Nos. 14, 15, and 16).

Blade Cores. (N = 4; Figure 10.2a–b) Blade

cores represent a specific type of prepared core strat-

egy in which elongated flakes with relatively straight

sides are produced for use as cutting and scraping

tools. A single Toyah phase blade core was recov-

ered from site 41BX528 (Cat. No. 71). This core

has a single-facet platform created by the trunca-

tion of one end of the cobble. Elongated flakes were

removed along one portion of the cobble surface.

Two smaller blade cores were recovered from site

41BX534 (Cat. Nos. 6 and 106), which may or may

not be associated with a Late Prehistoric occupa-

tion. Cat. No. 6 is an opposed platform blade core,

while Cat. No. 106 has a large single-facet platform

from which only two primary bladelike flakes were

removed. A final blade core was recovered from

site 41BX793 (Cat. No. 73). It has a multi-faceted

platform and unidirectional flaking.

Blade core technology in the Applewhite Res-

ervoir project area is substantiated by the presence

of edge-modified tools recovered from several sites:

eleven from 41BX528, two from 41BX544, one each

from 41BX793 and 41BX863, and three from

41BX872. The majority of these blade tools were

identified as hafted convex endscrapers, which link

them technologically to a probable Late Prehistoric

date as surmised by Hester (1978) for the majority

of blade technology in south Texas. Certainly, the

blade cores recovered from site 41BX831, which

date to the early Holocene, are different technologi-

cally and yielded smaller, more refined blades than

those selected as convex end-scraper blanks de-

scribed above. This suggests that blade technology

changed from the early Holocene to mid/late Ho-

locene as subsistence patterns changed to empha-

size bison procurement in the Late Prehistoric pe-

riod. Blade technology seems to have been more

important during the Late Prehistoric period. End-

scrapers on blades and other blade implements were

also identified from other sites in the region (Black

and McGraw 1985; Hall et al. 1986:312–315;

Lukowski 1987:252).


Macroflake/Flake Cores. (N = 2; Figure

10.2c) Both examples of this core type were recov-

ered from site 41BX793 (Cat. Nos. 56 and 137) and

probably represent special-purpose cores or recycled

implements. Cat. No. 56 is a large percussion flake

with a striking platform prepared around the edge

on the ventral surface; the dorsal side retains exte-

rior cortex. Several small flakes were removed from

the ventral surface. The second specimen (Cat. No.

137) is a heavily patinated percussion flake with ex-

terior dorsal cortex. A series of small percussion

flakes was removed from the dorsal surface. It is

possible that this specimen was a small cobble tool,

but no wear was observed on the retouched edge.

This core type is identified as a large flake that was

removed from a cobble and used as a core for the

production of smaller flakes. The reduction of these

cores is similar to that of partial cobble/pebble cores

except that a flake served as the core.

Microblade/Microflake Cores. (N = 2; Fig-

ure 10.2d) Both specimens are small angular frag-

ments with percussion flake scars on one or more

surfaces. A single specimen recovered from site

41BX534 (Cat. No. 5) has cortex on one surface.

The second specimen from site 41BX540 (Cat. No.

7) has small percussion flake scars on both surfaces

and cortex along one end. The size of percussion

flakes removed from both cores suggests that they

were intended for a special purpose, perhaps for in-

sertion into composite cutting implements or as ex-

pedient projectile points. The size of the flakes in

relation to the core does not suggest that they repre-

sent exhausted cores.

Core Fragments. (N = 4) Core fragments are

identified as angular remnants of cores. They ex-

hibit remnants of previous flake scars and negative

bulbs of percussion, but could not be identified as a

specific type of core. The majority probably repre-

sent fragments of partial cobble/pebble cores that

were broken during reduction or attempts to salvage

a ruined striking platform. A total of four cortical

core fragments was recovered from four sites: one

each from 41BX544 (Cat. No. 10), 41BX554 (Cat.

No. 10), 41BX669 (Cat. No. 5), and 41BX833 (Cat.

No. 4).

Bifacial Tools (N = 99)

A total of 99 bifaces was recovered during sur-

vey and testing in the Applewhite Reservoir project

area. This category includes all specimens that were

identified as biface fragments or whole bifaces of

Stage I through Stage III technology, but does not

include projectile points, which are discussed as a

separate category. The stages employed in this study

closely follow those established by previous research-

ers (Callahan 1979, Collins 1975, Whittaker 1994).

Stage I Bifaces. (N = 3; Figure 10.3) This

biface category is representative of the initial stage

of manufacture. Artifact shape was not defined and

the original shape and form of the blank can still be

determined. Retouch and modification at this stage

is largely limited to the lateral edges of the biface

blank, and cortex may be present on one or more

surfaces. Flake scars are typically deep and exhibit

a random orientation. The goal of this stage of biface

manufacture is usually to remove cortex and pre-

pare the biface for initial thinning.

It is interesting that a greater abundance of early

stage bifaces was not identified in the lithic assem-

blages from the Applewhite Reservoir project area

survey and testing. Two complete specimens were

recovered from sites 41BX872 (Cat. No. 9) and

41BX526 (Cat. No. 9) and a proximal fragment from

site 41BX540 (Cat. No. 4). The proximal fragment

was transversely fractured. All three specimens are

manufactured from thick chert percussion flakes

with some dorsal cortex. One has a cortical plat-

form. The technology of these early stage bifaces is

not the same as that evidenced by later stage frag-

ments or the whole specimens described below. The

early stage bifaces have rough percussion retouch

along the periphery of the flake and appear to repre-

sent failures in thinning. The specimen from site

41BX540 also has an oblique transverse fracture.

Stage II Bifaces. (N = 39; Figure 10.4a–b)

Stage II bifaces represent a more advanced reduc-

tion of the tool blank through thinning and shaping.

Flake scar patterns are still randomly oriented, but

the initial shape of the final bifacial product has be-

gun to emerge. Edges are typically sinuous, but not


as pronounced as in Stage I bifaces. Stage II bifaces

also were manufactured from chert. Four specimens

exhibit probable heat-treatment luster and color

changes. Only two of the bifaces characteristic of

this stage were complete; 10 were proximal-medial

fragments, 22 were distal or distal-medial fragments,

and 5 were small undetermined fragments. Reasons

for abandonment or manufacturing failure include

bend/break fractures and one transverse break.

The majority of Stage II bifaces were recov-

ered from site 41BX863, which yielded nine speci-

mens (Cat. Nos. 3, 4, 5, 6, 7, 8, 10, 12, and 13). Five

specimens each were collected from sites 41BX528

(Cat. Nos. 5, 7, 10, 12, and 17), 41BX554 (Cat. Nos.

1, 2, 4, 7, and 8), and 41BX872 (Cat. Nos. 8, 10, 12,

19, and 20). Three Stage II bifaces were found at

site 41BX534 (Cat. Nos. 7, 13, and 38). Sites

41BX538 (Cat. Nos. 221 and 426) and 41BX539

(Cat. Nos. 1 and 2) produced two bifaces each. A

single Stage II biface was recovered from each of

the following sites: 41BX537 (Cat. No. 99), 41BX568

(Cat. No. 10), 41BX664 (Cat. No. 308), 41BX669

(Cat. No. 9999), 41BX670 (Cat. No. 51), 41BX793

(Cat. No. 135), 41BX837 (Cat. No. 1), and 41BX865

(Cat. No. 35).

Stage III Bifaces. (N = 34; Figures 10.4c, 10.5)

The major objectives of this reduction/manufacture

stage are continued shaping and thinning. Both hard-

and soft-hammer flaking techniques can be applied,

but the use of soft-hammer percussion is dependent

upon the type of biface being manufactured and the

nature of the initial biface blank. Flake scar orienta-

tion has now become more uniform and patterned,

and the shape of the biface has been refined to ap-

proach the final form. All were manufactured from

chert, and three display evidence of heat treatment.

Only three of the Stage III bifaces were complete;

17 were proximal-medial fragments, 10 were distal

or distal-medial fragments, and 4 were small unde-

termined fragments.

Site 41BX528 produced 13 Stage III bifaces

(Cat. Nos. 2, 3, 4, 6, 8, 9, 11, 13, 14, 15, 16, 18, and

54). A total of five bifaces from this stage was re-

covered from site 41BX863 (Cat. Nos. 2, 9, 11, 15,

and 16). Sites 41BX554 (Cat. Nos. 3, 6, and 9) and

Figure 10.3. Example of a Stage I biface.


Figure 10.4. Selected bifaces from the project area: (a–b) Stage II bifaces; (c) Stage III bifaces; and (d–e) knives.


41BX526 (Cat. Nos. 5, 8, and 10) yielded three Stage

III bifaces each. Two Stage III bifaces each were

collected from sites 41BX544 (Cat. Nos. 2 and 264),

41BX837 (Cat. Nos. 1 and 2), and 41BX872 (Cat.

Nos. 1 and 11). A single Stage III biface was recov-

ered from each of the following sites: 41BX347 (Cat.

No. 8), 41BX540 (Cat. No. 2), 41BX669 (Cat. No.

250), and 41BX858 (Cat. No. 1).

Preforms. (N = 3) A total of three chert pre-

forms was collected from sites 41BX528 (Cat. Nos.

19 and 20) and 41BX544 (Cat. No. 226). All three

are thin (<1 cm) and plano-convex in cross section.

Cat. Nos. 19 and 226 are complete or nearly com-

plete specimens; Cat. No. 20 is a proximal fragment.

Cat. No. 19 has a triangular planview; Cat. No. 226

is more ovoid in shape. None of the three preforms

have haft elements, nor are they beveled. Percus-

sion flaking of Cat. No. 20 probably resulted in its

fracture. Evidence of pressure flaking can be ob-

served on the margins of Cat. Nos. 19 and 226.

Knives. (N = 6; Figure 10.4d–e) Bifacial imple-

ments used as knives were recovered from sites

41BX526 (Cat. Nos. 3, 4, 6, and 7), 41BX537 (Cat.

No. 141), and 41BX916 (Cat. No. 11 ). Only one of

these was not manufactured from chert; Cat. No. 7

was composed of quartzite. The surface of Cat. No.

6 shows crazing and is potlidded. Two specimens,

Cat. Nos. 6 and 7, have haftwear along the proximal

edges. Two knives are complete. The other artifacts

are represented by a proximal fragment, a medial

fragment, a distal fragment, and a lateral edge. All

six implements are bi-convex in cross section. Two

specimens are lanceolate in shape, and one is trian-

gular.

With the exception of Cat. No. 141, these knives

can be characterized as light-duty cutting tools used

to cut relatively soft materials for restricted periods

of time. Use-wear consists of rounding, smoothing,

and blunting that is primarily from light intensity. Cat.

No. 141 can be classified as a heavy-duty cutting

tool with more moderate to pronounced use-wear

and resharpening.

Distally Beveled Bifaces. (N = 11; Figures

10.6a–b, 10.7) This category includes all bifaces and

biface fragments that have a unifacially beveled dis-

tal end or fragments that technologically resemble

known beveled artifact types such as Clearfork tools

(Figure 10.6). All examples are made from chert.

Evidence of incidental heat exposure, not related to

manufacture, was identified on two specimens.

Three specimens can be considered preforms

of Clearfork bifaces (site 41BX526, Cat. No. 2; site

41BX863, Cat. No. 14; and site 41BX916, Cat. No.

12). Although the general shape and dimensions of

these preforms compare well to those of complete

Clearfork bifaces, the bits were not completed, indi-

cating that bit manufacture was probably one of the

last steps of the manufacturing process. The main

reason for abandonment in manufacture seems to

have been a failure to successfully thin the distal

end and achieve bit shape. One preform is a thick

side-struck secondary percussion flake.

Three Clearfork bifaces are complete and rep-

resent specimens abandoned at the end of their use-

life and exhibit signs of heavy use (site 41BX526,Figure 10.5. Example of a Stage III biface.


Figure 10.6 (a–b). Examples of distally beveled bifaces.


Cat. No. 1; site 41BX552, Cat. No. 1; and site

41BX863, Cat. No. 1). Two of these are triangular

in shape with convex proximal ends and flat to con-

cave bit shapes, while the third is ovate, broader dis-

tally, and has a convex bit. Only one has dorsal cor-

tex.

The Clearfork sample also includes one medial

fragment (site 41BX540, Cat. No. 3) and one medio-

distal fragment (site 41BX539, Cat. No. 1). Both

have transverse fractures indicative of haft break-

age during use, and the proximo-medial fragment is

heavily step-fractured at the bit. Another plano-con-

vex lanceolate biface has a convex beveled end with

bifacial step-terminated macroscars suggestive of

use in a chopping motion. This specimen was recov-

ered from site 41BX528 (Cat. No. 1).

Two beveled bifaces from site 41BX872 (Cat.

Nos. 6 and 7) resemble Nueces bifaces (Hester et

al. 1969:130–165), although one is slightly more tri-

angular than the type description. The beveled edge

is the widest portion of both bifaces, and one retains

dorsal cortex, suggesting that it was manufactured

from a larger flake. Cat. No. 6 has remnants of a

single burin spall scar along one lateral edge, after

which it was reworked, removing the proximal end

of the spall scar (Figure 10.7). The presence of a

burin spall scar is similar to those frequently observed

on Olmos bifaces (Turner and Hester 1993:269). The

bits of both tools have light usewear and step frac-

turing and there is no evidence of haft wear.

Clearfork and Guadalupe tools were recovered

previously from several sites in the Applewhite Res-

ervoir project area. Guadalupe tools were encoun-

tered during the test excavation of a burned rock

feature at site 41BX274 in direct association with

Martindale and Bell dart points (McGraw and Hindes

1987:121), and others were recovered from the sur-

face of this site. A number of Clearfork tools, both

bifacial and unifacial, were also recovered from the

surface of site 41BX274. In addition, a single

Guadalupe tool was recovered during work done by

SMU at site 41BX526 (McGraw and Hindes

1987:191). Testing by SMU at site 41BX540 yielded

a Clearfork tool and a Bell projectile point (McGraw

and Hindes 1987:160).

Excavations at site 41BX831, the Richard Beene

site, yielded an extensive collection of complete and

fragmentary Clearfork tools from excavated con-

texts (Dockall and Pevny 2007). The Late

Paleoindian component (ca. 9700–8700 B.P.) pro-

duced bifacial Clearfork tools in addition to burins,

points, and gravers. Other Clearfork tools were re-

covered from an Early Archaic surface and the up-

per portion of the Medina pedocomplex, which was

dated to the middle Holocene (Thoms et al. 1996:24).

Clearfork tools were found widely throughout cen-

tral and south Texas and into adjacent areas, with

some forms being associated with Paleoindian and

Early Archaic contexts, but continuing into the Middle

Archaic (Turner and Hester 1993:246).

Drills/Perforators. (N = 1; Figure 10.8) Ac-

cording to Turner and Hester (1993) these artifacts

are primarily characterized by a long, tapering bit

that is often diamond-shaped in cross section. The

bit and base are often retouched by bifacial pres-

sure flaking and/or light percussion. The distal ends

of the drill/perforator bits can be lightly to moder-

ately rounded through use or can be snapped off in a

shear fracture. Paleoindian and Archaic drills are

typically identified as reworked projectile points, while

Late Prehistoric drills and perforators are typically

smaller and often manufactured from small flakes.

Not all drills or perforators are bifacial. Unifacial

forms also exist, as do unmodified flakes of an ap-

propriate shape that exhibit typical wear for these

tool types.

Figure 10.7. Nueces biface from site 41BX872.


A single complete drill, or perforator, was re-

covered from site 41BX537 (Cat. No. 29). It was

made from chert and shows no evidence of burning

or heat treatment. Pressure flaking along the edges

has produced a uniform bifacial edge shape. The

shaft is missing due to a bending fracture; it is pos-

sible that it was broken during use due to torsional

stress at the base of the shaft.

Wedges/Chisels. (N = 2) A total of two chert

wedges was recovered from sites 41BX554 (Cat.

No. 5) and 41BX793 (Cat. No. 108). Both pieces

retain evidence of their original bifacial manufac-

ture with subsequent alteration by bipolar percus-

sion, use, and resharpening, resulting in an irregu-

larly shaped edge. The wedge/chisel from site

41BX793 remains complete; the one from site

41BX554 is very fragmentary.

Projectile Points (N = 116)

The assemblage of projectile points was subdivided

into the types established for Texas following Turner

and Hester (1993). The assemblage is broadly

grouped by the Quaternary subdivisions of Pleis-

tocene and Holocene. The reasoning for this broad

temporal classification is the lack of chronological

control over the sample of projectile points recov-

ered during survey and testing. The majority repre-

sents isolated surface finds at various sites, or they

were recovered in excavated contexts with an ab-

sence of associated dateable material. Most of the

represented point styles also have a substantial tem-

poral span or were never adequately dated.

Late Pleistocene/Early Holocene (Lan-

ceolate Forms). Angostura (N = 1; Figure 10.9)

This lanceolate specimen (Cat. No. 132) was manu-

factured from chert, is heavily patinated, and was

recovered from site 41BX793. Reworking is present

on the distal end, and the blade edges are slightly

convex. Prominent shoulders are present near the

midpoint of the specimen, and the greatest width

occurs at these shoulders. The base is slightly con-

cave and is only lightly smoothed, as are the proxi-

mal-lateral edges below the shoulders. Two Angos-

tura specimens were recovered from the Panther

Springs site (41BX228), but were from a later con-

text (Black and McGraw 1985:124). In addition, two

reworked and fragmentary specimens were recov-

ered from the surface of site 41BX652 (McGraw

and Hindes 1987:319).

Research at the Richard Beene site (41BX831)

conducted by TAMU revealed what must be con-

sidered as the best example of Angostura points from

a distinct stratigraphic context (Dockall and Pevny

2007). Multiple radiocarbon dates on bulk sediment

samples containing Angostura points range from 9660

± 100 to 10,130 ± 130 B.P. (Thoms and Mandel

1992). Thoms (1993) further illustrates a very lucid

picture of the great degree of variability among An-

gostura points as a stylistic grouping, and this senti-

ment is echoed by Thurmond (1990:24) for Late

Paleoindian formal bifaces. The variety of artifacts

found with the Richard Beene site specimens strongly

argues for an Early Archaic subsistence mode rather

than the traditional view of Paleoindian big game

hunting (Thoms et al. 2004 [Author: Not on refer-

ence list.]).

Early Holocene (Early Archaic Unstemmed

Lanceolate and Triangular Forms). Early Tri-

angular (N = 2; Figure 10.10) One specimen (Cat.

No. 10) was manufactured from chert and was re-

covered from site 41BX539. Lateral blade edges are

straight to mildly convex and finely pressure flaked.

Right blade edges are alternately beveled, and there

is no specific haft element or haft wear observed.

Figure 10.8. Example of a bifacial drill/perforator.


Microscopic use-wear suggests that this biface was

employed as an unhafted cutting implement for at

least the last portion of use-life. A second specimen

(Cat. No. 83) was recovered from site 41BX670.

This specimen has straight blade edges and a straight

basal edge. The edges are alternately beveled. The

lateral edges have microwear suggesting cutting, and

the distal tip is well smoothed with microscopic step

fractures at the edges indicative of a boring or per-

forating task.

Early Holocene (Early Archaic Notched-

Shouldered Forms). Uvalde (N = 1; Figure 10.11a)

The point was manufactured of chert and has a tri-

angular blade with convex edges and shoulders that

are broken. The stem is expanding and the basal

edge is concave. The distal end of the blade is re-

worked. This specimen (Cat. No. 1) was recovered

from the surface of site 41BX532.

La Jita (n = 1; Figure 10.11b) La Jita is the site

where this point was first encountered, as described

by Hester (1971:74). Researchers note that this type

grades into the Nolan type and indicate that it actu-

ally may represent unfinished Nolan points (Black

and McGraw 1985:118–119; Turner and Hester

1993:140). This specimen was a surface find (Cat.

No. 1) at site 41BX849. The raw material is chert,

and the blade is triangular with convex lateral edges.

The stem is slightly expanding and has been thinned

by bifacial removal of longitudinal flakes.

Bulverde (N = 3; Figure 10.11c) All specimens

are chert and have triangular blades with straight to

slightly concave edges. Shoulders are highly promi-

nent with small barbs, and the stems are contracting

with straight to convex edges and a wedge-shaped

basal edge. All are reworked distally and one has

alternately beveled lateral edges. Three Bulverde

points were recovered from sites 41BX526 (Cat.

No. 21), 41BX554 (Cat. No. 1), and 41BX854 (Cat.

No. 1). The Bulverde type is assigned to the Early

Archaic period (Turner and Hester 1993:82). Suhm

and Jelks (1962:169) assigned an age range of 3000

B.C. to A.D. 500 or 1000, and Weir (1976:51) at-

tributes them to the Clear Fork and Round Rock

phases, which correspond to Local Periods 6 and 7

at Panther Springs (Black and McGraw 1985:116).

Local Period 6 (Early Archaic) was radiocarbon

dated to 2920 ± 130 B.C. and Period 7 to 800 ± 60

B.C. at the Panther Springs site (Black and McGraw

1985:277, 299).

Martindale (N = 1; Figure 10.11d) The speci-

men is a surface find from site 41BX528 (Cat. No.

75), is considerably reworked on the distal end, and

has a remnant of a lateral macrofracture from pro-

jectile impact. The raw material is chert and the blade

edges are slightly recurved. The stem is expanding

and the basal edge is recurved.

Middle Holocene (Transitional-Middle Ar-

chaic Unstemmed and Triangular Forms).

Pandora (N = 1; Figure 10.12a) The Pandora point

(Cat. No. 2) recovered from site 41BX552 has a

lanceolate blade, a straight basal edge, and was

manufactured from chert. The base was thinned

slightly by the removal of short longitudinal flakes

from one surface. Lateral blade edges basically are

straight and are reworked on the distal end. It is

probable that this specimen and the Pandora type in

general may represent a knife or preform (Turner

and Hester 1993:170). It is assigned to the Middle to

Late Archaic of the Lower Pecos, south and central

Texas.

Carrizo (N = 2; Figure 10.12b–c) The Carrizo

point is an unstemmed triangular point with a single

notch in the center of the basal edge. The one com-

plete specimen (Cat. No. 325) recovered from the

surface of site 41BX544 has concave to slightly

Figure 10.9. Late Pleistocene/Early Holocene Angostura

projectile point.


Figure 10.10. Early Holocene Early Triangular projectile point.

Figure 10.11. Early Holocene (Early Archaic Notched-Shouldered) projectile points: (a) Uvalde; (b) LaJita; (c) Bulverde;

and Martindale.


Figure 10.12. Middle Holocene (Transitional-Middle Archaic Unstemmed and Triangular) projectile points: (a) Pandora;

(b–c) Carrizo; (d–e) Tortugas; (f) Kinney; (g) Anthon; (h) Matamoros; and (i) Refugio.


recurvate alternately beveled blade edges. The sec-

ond specimen is a proximal fragment of an impact-

damaged specimen (Cat. No. 83) recovered from

the surface of site 41BX528. Both specimens are

chert. The Carrizo type has a very localized distri-

bution in southwest Texas (Black and McGraw

1985:127). A single specimen was recovered from

the Panther Springs site from a mixed Archaic con-

text. This point style is believed to be associated with

the Middle to Late Archaic (Turner and Hester

1993:84), but the exact dates are unknown.

Tortugas (N = 4; Figure 10.12d–e) All speci-

mens have triangular blades with straight to con-

cave lateral edges. Basal edges on complete speci-

mens vary from straight to slightly concave and were

thinned by the bifacial removal of longitudinal flakes.

This created a basal edge that is distinctly wedge-

shaped in cross section. There are three complete

specimens, one distal fragment, and one proximal-

medial fragment with an impact fracture. Two com-

plete specimens have no edge beveling, while the

one complete and all the fragmentary examples have

alternate edge-beveling of the lateral blade edges.

The angle of edge beveling on these points varies

from 36º to 49º. All were manufactured from chert.

One heavily patinated point was recovered from site

41BX793 (Cat. No. 96); three were recovered from


2), and 41BX872 (Cat. No. 1).

The same typological problems that are associ-

ated with the definition of other triangular dart point

forms in south Texas are also associated with the

Tortugas type. Black and McGraw (1985:129) feel

that the confusion surrounding triangular point forms

results from geographical and temporal overlap, as

well as attributes of blade beveling and basal thin-

ning, which are highly variable.

Turner and Hester (1993:188) place the Tortugas

point in the Middle Archaic period and possibly ear-

lier. Weir (1976:51–52) attributes this point type to

the San Geronimo and Clear Fork phases in central

Texas. At the Panther Springs site no distinctions

were made between Thinned-Base Early Triangu-

lar and Tortugas (Black and McGraw 1985:129–133),

and all specimens were attributed to Local Period 5,

between 4000 and 2500 B.C.

Recent research at the Loma Sandia site has

expanded our knowledge of this point style (Highley

et al. 1995:433–441). The majority of Tortugas points

from Loma Sandia were recovered from a late Middle

Archaic context dated to 850–550 B.C. The sample

of recovered Tortugas points from Loma Sandia in-

dicates that both beveled and unbeveled forms were

found and that they served as both projectile points

and knives (Wright 1995). Wright (1995:578) con-

cludes that, based on the scarcity of consistent pat-

terns of wear associated with cutting or sawing,

Tortugas points were probably not used as knives;

the majority were used as projectile points. This ar-

gument fails to address the purpose of alternate bev-

eling that extends in most cases along the complete

length of the lateral edges and suggests that Tortugas

points typically were removed from the haft for

resharpening. The single specimen from site 41BX534

(Cat. No. 1), which exhibits an impact fracture, also

exhibits this type of complete edge beveling. A dual

function for Tortugas points is suggested by the ob-

servations from the Applewhite Reservoir, Loma

Sandia, and Panther Springs sites.

Kinney (N = 1; Figure 10.12f) One point was

recovered from the surface of site 41BX526 (Cat.

No. 22) and is manufactured of chert. The lateral

blade edges are strongly convex to recurved and the

tip is finely retouched to a needle-like point. One

basal corner is broken and the opposite corner has

traces of probable asphaltum. This specimen is con-

sidered a Kinney point because of size, lack of alter-

nate beveling on the blade edges, and the amount of

basal concavity. Turner and Hester attribute Kinney

points to the Middle Archaic (1993:137). A single

Kinney point also was found at the Loma Sandia

site (Highley et al. 1995:422–423), but was out of

context.

Anthon (N = 1; Figure 10.12g) The Anthon type

is a medium-sized triangular point with straight to

convex edges and a straight basal edge. This speci-

men (Cat. No. 11) was recovered from site 41BX669

and was manufactured from heat-treated chert. The

lateral edges and tip are reworked considerably, and

there is slight edge smoothing or grinding along one

edge close to the base. Some confusion in the iden-

tification of Anthon, Kinney, Tortugas, and other simi-

lar triangular point styles from the Middle to Late


Archaic makes identification, especially of single

specimens, very difficult. Davis (1991:30) assigns

the Anthon to a time period ranging from 2000 to

1500 B.C.

Matamoros (N = 1; Figure 10.12h) This single

specimen was recovered from a surface context at

site 41BX534 (Cat. No. 1). It is an unstemmed tri-

angular biface with convex lateral edges and a

straight basal edge. There is slight alternate bevel-

ing of the lateral blade edges and the chert was heat

treated. The point is considered to be a Matamoros

because the dimensions are smaller than either

Tortugas or Early Triangular, although a consider-

able degree of overlap exists.

Refugio (N = 1; Figure 10.12i) The Refugio point

is typically lanceolate in appearance with a rounded

basal edge. The lateral blade edges of this specimen

are slightly convex and the tip is very pointed due to

retouch. Secondary retouch or resharpening produced

a very thick cross section in relation to the width of

the point, and one edge is distinctly beveled. The

basal area is well-thinned for hafting, and resharp-

ening does not proceed past the haft area. The chert

is heavily patinated, and the specimen was recov-

ered from site 41BX793 (Cat. No. 64). The Refugio

point is usually assigned to a Middle to Late Archaic

time span (Davis 1991:158). Turner and Hester

(1993:178) consider the possibility that some of these

points are merely preforms. The degree of

resharpening on this specimen precludes that possi-

bility in this instance.

Middle Holocene (Transitional-Middle Ar-

chaic Notched and Shouldered Forms). Ensor

(N = 3; Figure 10.13a) All specimens were manu-

factured from chert and none exhibit definitive evi-

dence of heat treatment. Both of the specimens from

site 41BX669 (Cat. No. 12 and 14) are proximal-

medial fragments with missing tips. The point from

site 41BX537 (Cat. No. 137) is complete and heavily

patinated. Blade edges vary from straight to con-

vex; stems are corner-notched and expanding. Two

of the blades were beveled alternately along the edge

(Cat. Nos. 137 and 12).

Woerner and Highley (1983:9–11 ) differentiate

three varieties of Ensor points based on the shape of

the basal edge of the stem: Variety I has a concave

base, Variety II has a convex base, and Variety III

has a straight base. The specimen from site 41BX537

(Cat. No. 137) and one from site 41BX669 (Cat.

No. 12) can be classified as Variety I. The second

Ensor from site 41BX669 (Cat. No. 14) is a Variety

III type.

Turner and Hester (1993:114) assign Ensor points

to the Transitional Archaic (200 B.C. to A.D. 600).

The Ensor and Frio types are diagnostic of the Twin

Sisters phase of the Transitional Archaic in central

Texas from A.D. 200–700 (Weir 1976:136). Ensor

points also were recovered from deposits described

as Local Period 9 (Twin Sisters and Driftwood

phases) at Panther Springs (Black and McGraw

1985:284).

Marcos (N = 4; Figure 10.13b–c) All specimens

are fragmentary and manufactured of chert, and only

one was heat treated. Two are relatively complete,

but are missing the barbs. One is heavily burned and

potlidded and is missing a large portion of the base,

and one has a missing barb and a transverse snap at

the tip. All exhibit varying degrees of resharpening

along the blade lateral edges. Blade edges vary from

straight to slightly recurved, and basal edges are

straight. The two relatively complete Marcos points

exhibit distal impact fractures consistent with crush-

ing. The heavily burned specimen also has a crushed

distal tip due to projectile impact. Three examples

were recovered from site 41BX528 (Cat. Nos. 81,

82, and 87) and one from site 41BX858 (Cat. No.

2).

Marcos points are assigned to the Transitional

to Late Archaic (600 B.C. to A.D. 200) by Turner

and Hester (1993:147). In central Texas this type is

attributed to the San Marcos phase (Weir 1976:51).

At the Panther Springs site, Marcos points were

found in deposits equated with Local Period 8 (650

B.C. to A.D. 250; Black and McGraw 1985:112),

which are stratified above Castroville points.

Marshall (N = 4; Figure 10.13d) A total of three

fragmentary examples and one relatively complete

point were analyzed. All were manufactured of chert

(one heat treated and one burned) with the excep-

tion of one made from a fine-grained purple quartz-


Figure 10.13. Middle Holocene (Middle Archaic Notched and Shouldered) projectile points: (a) Ensor; (b–c) Marcos;

(d) Marshall; (e) Montell; (f) Pedernales; (g) Travis; (h) Frio; and (i) Darl.


ite. Stem edges are slightly to moderately expand-

ing, and basal edges vary from straight to convex.

All specimens were strongly corner notched, although

none have intact barbs. Transverse blade fractures

are present on two fragments, and the relatively com-

plete specimen has a reworked blade and a basal

fracture from projectile impact along the lateral edge

of the stem.

The type description provided by Suhm and Jelks

(1962:211) incorporates a wide range of variation.

Specimens from Panther Springs exhibit certain

morphological similarities to Castroville points (Black

and McGraw 1985:111). Weir (1976:135) proposes

that Marshall points are a morphological derivation

of Pedernales, resulting from the stylistic influence

of Plains Archaic peoples. Three specimens were

recovered from site 41BX526 (Cat. Nos. 24, 25, and

26) and one from site 41BX539 (Cat. No. 11).

Marshall is a typical Middle Archaic dart point

(Turner and Hester 1993:149) with an estimated age

of 1000 B.C. or somewhat earlier. Suhm and Jelks

(1962:211) estimated an earlier time span from 4000

or 3000 B.C. to A.D. 1000. Weir (1976:54–55) as-

signs the Marshall type to both the Round Rock and

San Marcos phases of central Texas. At Panther

Springs, this style is associated with Local Periods 7

and 8 (Black and McGraw 1985:111).

Montell (N = 3; Figure 10.13e) These points

have a fairly broad triangular blade with straight lat-

eral edges. All stems are corner notched, expand-

ing, and have a vertical V-shaped basal notch (site

41BX526, Cat. No. 23; site 41BX528, Cat. No. 84;

and site 41BX872, Cat. No. 4). The distal tips of all

examples are broken in transverse snaps, barbs are

missing, and the stem of one specimen is broken. All

were manufactured from chert and one was heat

treated.

Montell is a Late to Transitional Archaic dart

point type (Turner and Hester 1993:157) dating

roughly to between ca. 1000 B.C. and A.D. 200.

Weir (1976:135) assigns the Montell type to the San

Marcos phase of central Texas. At Panther Springs,

the Montell type consistently occurs in San Marcos

phase deposits along with Castroville specimens.

Pedernales (N = 5; Figure 10.13f) Only one

complete specimen was recovered; there were four

proximal-medial fragments. All points were manu-

factured from chert with two exhibiting heat treat-

ment. The blade edges of all points were reworked

and are convex. Shoulder shapes vary from tapered

to well-barbed depending on the degree of blade

resharpening. All stems are contracting, but lateral

stem edges vary from straight to convex. The basal

concavity varies from shallow to U-shaped. Two

specimens were recovered from site 41BX528 (Cat.

Nos. 88 and 96) and one each from sites 41BX540

(Cat. No. 5), 41BX546 (Cat. No. 2), and 41BX872

(Cat. No. 5).

This type is quite common throughout central

Texas and the Lower Pecos area (Black 1989c:49).

Black and McGraw (1985:113) note that the

Pedernales type was subdivided into several variet-

ies by past researchers (Sorrow 1969; Sorrow et al.

1967). They also stress that the chronological and

geographic significance of this variation is not fully

tested. Areas of future research should also con-

sider that breakage, as well as reworking or recy-

cling, influences intra-type variation among

Pedernales points.

Pedernales is a distinctively Middle Archaic dart

point type. Weir (1976:51) includes this type as a

diagnostic feature for the Round Rock phase in cen-

tral Texas. At the Panther Springs site this type is

assigned to Local Period 7 (2000–600 B.C.). Turner

and Hester give Pedernales a time range of 2000–

1200 B.C. (1993:171).

Travis (N = 1; Figure 10.13g) A single specimen

was recovered from site 41BX526 (Cat. No. 28)

and was manufactured of heat-treated chert. The

blade is narrow and triangular with convex lateral

edges. The stem is slightly contracting with straight

stem edges and a convex basal edge. The stem edges

do not exhibit the steep alternate beveling that is

characteristic of Early Archaic Nolan points. The

blade was resharpened by pressure flaking, which

has created a sharp point. Travis points are found

primarily in central Texas, but are not as common in

adjacent areas, and may have a temporal distribu-

tion ranging from 2650 to 2050 B.C. (Turner and

Hester 1993:189).


Frio (N = 2; Figure 10.13h) These specimens

are medium-sized dart points with triangular blades

and expanding stems. The basal edges of both ex-

amples are moderately concave. Shoulders are well

developed, but are not barbed. The material used in

both examples is chert and neither was heat treated.

Each was reworked distally. One Frio point was re-

covered from site 41BX528 (Cat. No. 86) and one

from site 41BX872 (Cat. No. 21). Frio points char-

acterize the Transitional Archaic period (ca. 200 B.C.

to A.D. 600, perhaps later in some areas; Turner

and Hester 1993:122).

Darl (N = 2; Figure 10.13i) Darl points are clas-

sified based on technology and comparisons with

other examples. Two specimens were found at site

41BX526. Both are burned heavily and exhibit potlids

and crazing. One is a proximal-medial fragment (Cat.

No. 29) and the other is a medial fragment with only

a portion of the blade and stem (Cat. No. 27). The

stem fragment has alternately beveled lateral edges,

very slight shoulders, and a concave basal edge. The

blade edges of the medial fragment are also alter-

nately beveled. According to Turner and Hester

(1993:101) Darl points are from the Transitional Ar-

chaic, ca. A.D. 200.

Unidentified Dart Points. (N = 7) A single

point was recovered from each of the following sites:

41BX346 (Cat. No. 1), 41BX347 (Cat. No. 1),

41BX537 (Cat. No. 94), 41BX540 (Cat. No. 6),

41BX554 (Cat. No. 21), 41BX669 (Cat. No. 10),

and 41BX834 (Cat. No. 1). All dart points in this

category consist of heavily retouched whole points

that could not be assigned to a specific known type,

so their variability is described as a group. Three

specimens are lanceolate or narrow triangular in

shape. Each has a distinct basal edge; one is straight,

one is convex, and one is concave. The point with

the concave basal edge also has very slight shoul-

ders with concave lateral stem edges. Another lan-

ceolate specimen has one lateral blade edge that was

reworked by light direct percussion while the other

edge is well finished by parallel pressure flaking. One

point has been reworked or recycled into a beaked

or grooving tool. The last two specimens are corner/

side-notched forms and may represent heavily

resharpened Ensor points. All specimens are chert;

one is crazed and potlidded.

Miscellaneous Dart Point Fragments. (N =

12) This group is composed of proximal, medial, dis-

tal, and other fragments of dart points. Breakage

and reworking have precluded specific type identifi-

cation and the fragments represent breakage during

use or manufacture. All specimens are chert.

Proximal/Proximal-Medial Fragments (N = 5)

Two proximal fragments were recovered from site

41BX534. Cat. No. 8 has a parallel stem with a con-

cave base. Evidence of moderate to heavy grinding

can be seen on the haft element and it has alter-

nately beveled edges. Cat. No. 11 has a parallel stem

with a straight base. The fragment recovered from

site 41BX539 (Cat. No. 12) was burned; it has a

straight stem with evidence of alternate beveling. A

basal fragment from site 41BX554 (Cat. No. 17)

displays a slightly expanding stem with a straight base.

Site 41BX670 (Cat. No. 1) produced a proximal frag-

ment with an expanding stem and concave base.

Proximal fragments all have transverse snap frac-

tures suggesting breakage during use and later re-

moval from the dart foreshaft during tool repair.

Barb Fragments (N = 1) The single barb frag-

ment recovered from site 41BX863 (Cat. No. 26) is

chert and may be from a heat-treated dart point.

The width, shape, and thickness of this fragment

suggest that it is from a dart point type, such as

Castroville or Marcos, on which the barbs are rather

long and narrow.

Distal Fragments (N = 7) All distal fragments

are chert and one is heavily burned. Heat treatment

is evident on only one specimen. Two specimens

exhibit alternate beveling of the lateral blade edges,

one has a perverse fracture, and three have trans-

verse snap fractures. The specimen with the per-

verse fracture shows evidence of soft-hammer flak-

ing, but not pressure flaking, suggesting that it is a

dart point preform fragment resulting from a manu-

facturing failure. Specimens were recovered from


18), 41BX664 (Cat. No. 306), 41BX793 (Cat. No.

88), 41BX853 (Cat. No. 1), and 41BX863 (Cat. Nos.

28 and 29).

Late Holocene (Late Prehistoric-

Protohistoric/Historic). Alba (N = 1; Figure


10.14a) The point has a triangular blade with very

fine pressure flaking. The stem is slightly contract-

ing and the basal edge is straight. The point and one

barb are broken. This specimen was recovered from

site 41BX544 (Cat. No. 20). The distribution of Alba

points is primarily throughout central and east Texas,

the Coastal Plain, and Louisiana (Turner and Hester

1993:200), but extends occasionally into north-cen-

tral Texas (Suhm and Jelks 1962:263). This type also

shows some resemblance to the Bonham type, which

seems to extend into west Texas as far as the Pecos

River (Suhm and Jelks 1962:264). The Alba type is

dated to between A.D. 800 and 1200 (Turner and

Hester 1993:200).

Fresno (N = 1; Figure 10.14b) This is a small,

thin triangular biface with a snap fracture at the

medial portion of the blade. It was manufactured by

pressure flaking from a thin flake of heat-treated

chert. The basal edge is straight, the basal corners

are slightly rounded, and the lateral blade edges are

straight. Along one lateral edge is the distal end of a

lateral macrofracture resulting from impact. One

basal corner is slightly damaged, which also may be

the result of impact. This point (Cat. No. 27) was

recovered from the surface of site 41BX863. Fresno

points are most prevalent in central, east, and south

Texas (Turner and Hester 1993:213) and are broadly

associated with the Late Prehistoric period. Fresno

points are very similar morphologically to Cameron

points, but the Cameron type is limited geographi-

cally to the Lower Gulf Coast and the Lower Rio

Grande Valley. Both point styles overlap in time and

may represent the same point style with raw mate-

rial differences determining, or limiting, size. Smaller

pieces of raw material, as well as their scarcity along

the coast and Lower Rio Grande Valley could ac-

count, in part, for the size difference between Fresno

and Cameron points.

Scallorn (N = 6; Figure 10.14c) Two of the

Scallorn points were recovered from site 41BX528

(Cat. Nos. 73 and 77), two from site 41BX554 (Cat.

Nos. 14 and 23), and one each from sites 41BX669

(Cat. No. 6) and 41BX854 (Cat. No. 2). These speci-

mens have long, slender blades with straight lateral

edges. The blade edges of three specimens are finely

serrated. All stems are corner notched and widely

expanding, and three have broken stems. Basal edges

are straight, convex, or slightly concave, and all ex-

amples have pronounced barbs. Four of the points

are chert, one is chalcedony, and one is petrified

wood. The chalcedony specimen (Cat. No. 77) was

manufactured from a material that originates in the

Three Rivers area of Live Oak County, Texas. One

point was heavily burned (Cat. No. 73). Turner and

Hester (1993:230) state that Scallorn points prob-

ably occurred during the time frame of A.D. 700–

1200. Scallorn points were associated with the Aus-

tin phase (A.D. 1000–1300) in Texas (Suhm and

Jelks 1962) and were commonly observed in depos-

its or contexts that predate the occurrence of Perdiz

points.

Perdiz (N = 16; Figure 10.14d–e) All of the

specimens included in this group were pressure

flaked from small chert flakes. The majority retain

some portion of the ventral surface of the original

flake blank, although eight specimens were almost

entirely flaked bifacially. Eight (50 percent) are com-

plete or relatively complete (sites 41BX347, Cat. No.

6; 41BX528, Cat. Nos. 78 and 94; 41BX534, Cat.

No. 95; 41BX537, Cat. No. 182; 41BX544, Cat. No.

15; 41BX546, Cat. No. 1; and 41BX669, Cat. No.

7). Of these only one was substantially resharpened

(site 41BX669, Cat. No. 7) and three others were

minimally resharpened. A majority of the chert flakes

selected for manufacture into Perdiz points do not

exhibit any identifiable evidence of deliberate heat

treatment, unlike Perdiz points from the Hinojosa site

in Jim Wells County, Texas. Black (1986:57) reports

that more than 90 percent of all complete specimens

from Hinojosa were heat treated. Unlike the blades,

the stems of all Perdiz points recovered during the

Applewhite Reservoir project area survey and test-

ing are bifacially flaked. This indicates that minimal

effort during manufacture was put into blade shap-

ing. The shoulders of all specimens are well barbed

and stems are acutely contracting and either pointed

or rounded. Perdiz fragments were recovered from

sites 41BX528 (Cat. Nos. 76, 93, 95, and 98) and

41BX544 (Cat. Nos. 17, 18, 21, and 28).

This point style is very widely distributed through-

out Texas and portions of Louisiana (Turner and

Hester 1993:227). Perdiz points are associated with

Local Period 11 at the Panther Springs site (Black

and McGraw 1985:99). Scallorn and Edwards points


Figure 10.14. Late Holocene (Late Prehistoric/Historic) projectile points: (a) Alba; (b) Fresno; (c) Scallorn; (d–e)

Perdiz; (f) Edwards; (g) Guerrero; and (h–i) arrow point performs.

predate the occurrence of Perdiz in the Choke Can-

yon area (Black 1989c:52; Hall et al. 1986). Corbin

(1974) notes that Scallorn, Fresno, and Padre types

date earlier than Perdiz points along the Coastal Bend

area of Texas. At Panther Springs, Scallorn and

Edwards types occur in deposits below Perdiz (Black

and McGraw 1985:286). Perdiz points are dated to

A.D. 1200–1500 (Turner and Hester 1993:227).

The breakage pattern of Perdiz points from the

Applewhite Reservoir project area compares well

to other similar studies of Perdiz points (Black

1986:159–163; Brown et al. 1982:42–43). There are

two basic fracture patterns observed on arrow points

readily identified as Perdiz: (1) snap fractures at the

juncture of the blade and stem and across the blade

(n = 5) and (2) snap fractures at the juncture of the


blade and stem with the blade intact (n = 3). In each

fracture group, barbs may also be missing. Data from

the Hinojosa site and 41LK67 indicate that the most

common areas of fracture are across the blade and

the upper stem area (Black 1986:161; Brown et al.

1982:42). Brown further suggests that Perdiz points

that impact against a hard substrate (rock or bone)

are more likely to shatter than snap, whereas points

that embed in softer material (flesh and muscle) are

more susceptible to bending stresses. The high de-

gree of fracture patterning on broken Perdiz points

suggests that all were broken in nearly identical

modes of use, i.e., via impact. The patterns of break-

age and the thinness of Perdiz points strongly sug-

gests that these implements were function specific.

Perdiz points were also subjected to resharpening

and repair if breakage was not severe enough to

end their use-life. Black (1986:Figure 5c, i, m–n) il-

lustrates several examples that represent resharp-

ened examples. Based on an assessment of the stem/

blade length, there are four from the Applewhite

Reservoir project area that represent proximally or

distally repaired specimens.

Cox and Smith (1991) suggest that Perdiz points

recovered from the McGloin House site (41SP135)

in San Patricio County, Texas, occasionally were

retrieved for use as hafted cutting tools. Their argu-

ments are interesting, but failure to consider attritional

damage through (1) repeated use as projectile points,

(2) soil wear, and (3) the wear effect produced along

lateral edges during use as projectile points, makes

their results tenuous (Dockall et al. 1993). It is felt

that the size and technology of Perdiz points sug-

gests that they were rarely selected for use in butch-

ering tasks. A major premise of the research con-

cerning the function of Perdiz points is that if they

were employed as projectile points, then there must

be evidence of impact fractures (Cox 1993:307–308;

Cox and Smith 1991:293). Cox (1993:308) made the

novel observation that edge attrition was observed

on more whole Perdiz points than on broken points,

which suggests that other functions must be consid-

ered in addition to use as projectile points. A simpler

and more elegant scenario is that increased wear on

whole Perdiz points is associated with a longer use

life than that of broken points. The rebuttal to Dockall

et al. (1993) by Cox (1993) also demonstrates that

the authors had not considered the effects of

resharpening and rejuvenation on projectile points,

even on styles such as Perdiz. Cox (1993:308) raised

the question of “why are there so many dull, poorly

shaped arrow points?” He is probably correct in

thinking that these could no longer be used as arrow

points, in fact, he was discussing Perdiz points that

had met the end of their use-life without suffering a

major impact that had rendered them useless ear-

lier. The process of rejuvenation must be considered

carefully; not all projectile points will reach the end

of their use-life via projectile impact. Some remained

unbroken and were resharpened or retipped to the

point of discard.

The segments represented within the current

sample include medial and distal portions that may

have been returned to the sites within portions of

meat or within whole game animals (Black 1986:163).

Probable Perdiz Fragments (N = 9) This small

series of fragments is described here because of

their technological and morphological similarities to

Perdiz points. There is also a distinct similarity in

fracture patterns between these fragments and the

incomplete Perdiz points described above. The

sample consists of four tip fragments, one lateral

edge/barb segment, two medial segments missing

the tip and stem, one complete blade minus the stem,

and one heavily reworked blade missing the proxi-

mal end of the stem. All tip fragments, and one com-

plete blade minus the stem, are unifacially pressure

flaked. The medial segments and one complete blade

segment exhibit some bifacial flaking. All specimens

were broken in transverse fractures probably result-

ing from stress during impact (as discussed above).

There were five fragments recovered from site

41BX544 (Cat. Nos. 16, 19, 23, 24, and 25), two

from site 41BX669 (Cat. Nos. 15 and 17), and one

each from sites 41BX528 (Cat. No. 97) and 41BX554

(Cat. No. 15).

Edwards (N = 3; Figure 10.14f) These points

were manufactured from heat-treated chert flakes

through a combination of bifacial thinning and pres-

sure flaking techniques. The blades are triangular

with straight to slightly concave edges. Stems are

deeply corner notched, expanding, and have widely

flaring barbs. Bases are concave and flare as widely

as the barbs. The distal portion of the blade is bro-


ken on two specimens; one by a transverse fracture

and one by an impact fracture. The third specimen

is missing one barb. Three specimens were recov-

ered from sites 41BX525 (Cat. No. 2), 41BX538

(Cat. No. 61), and 41BX554 (Cat. No. 20).

Edwards is one of the largest arrow point types

and is also felt to be the earliest arrow point form in

south and central Texas (Turner and Hester

1993:212). Black and McGraw (1985:101) note a

morphological similarity between Edwards and

Scallorn types and assign Edwards points to Local

Period 10 at Panther Springs. Turner and Hester

(1993:212) date the Edwards point type to A.D. 960–

1040.

Guerrero (N = 1; Figure 10.14g) The specimen

from site 41BX528 (Cat. No. 74) is a small triangu-

lar stemless point manufactured of chert and is pres-

sure flaked. The lateral edges of the blade are slightly

convex. The basal edge is concave and was created

by the removal of small longitudinal pressure flakes.

The chert exhibits a gloss from thermal alteration

which indicates that the point was manufactured

from a heat-treated flake. The basal concavity and

the lateral edges of the proximal area were ground.

Grinding on the lateral edges extends from the base

upward about 6.6 mm and creates very slight shoul-

ders.

Turner and Hester (1993:216) report the distri-

bution of Guerrero points throughout south and south-

east Texas and into Coahuila, Mexico. McGraw and

Hindes (1987:169) indicate that a burned basal frag-

ment identified as a Guerrero point was encountered

at site 41BX528. The addition of the current speci-

men now brings the total of this point style from that

site to two. Guerrero points are known to have been

used into the eighteenth century.

Arrow Point Preforms (N = 9; Figure 10.14h–i)

These artifacts represent the early stages of arrow

point manufacture. All of these preforms were manu-

factured from chert flakes. Only one has evidence

of bifacial thinning prior to pressure flaking; the oth-

ers show only pressure flaking. There are several

indications that these preforms were abandoned or

broken during manufacture. Three preforms have

transverse snap fractures across the blade and one

has a lateral snap fracture along the blade edge.

Another is complete, but was abandoned due to fail-

ure in thinning. Four resemble the Cliffton point style

and are here considered to be probable Perdiz point

preforms because they exhibit the early stages of

Perdiz stem manufacture. The remaining two speci-

mens are unstemmed triangular preforms with rela-

tively straight basal edges. Three preforms were

recovered from site 41BX528 (Cat. Nos. 7, 80, and

25), two from site 41BX534 (Cat. Nos. 9 and 14),

and two from site 41BX669 (Cat. Nos. 8 and 9);

one preform was recovered from site 41BX544 (Cat.

No. 29) and one from site 41BX554 (Cat. No. 19).

Miscellaneous Arrow Point Fragments (N =

6) These proximal-medial and distal fragments are

portions of broken arrow points that could not be

assigned to a specific known type. Fragments were

recovered from sites 41BX539 (Cat. No. 13),

41BX544 (Cat. Nos. 22 and 26), 41BX554 (Cat. No.

16), 41BX669 (Cat. No. 16), and 41BX872 (Cat.

No. 22).

Edge-Modified Tools (N = 114)

This broad classification is used to describe both uti-

lized flakes and purposefully edge-retouched tools.

Some of the edge-retouched tools probably repre-

sent a continuum of resharpening episodes on ini-

tially unmodified flake tools as a result of use. Still

others represent deliberately flaked pieces for spe-

cific purposes, whether the retouch was to facilitate

hafting, manual prehension, or tool performance in

use.

Tools were employed in a variety of tasks such

as cutting, scraping, shredding, planing, and graving.

These tool motions are suggestive of working a va-

riety of materials, including various types of plant

resources. Tasks would have included the manufac-

ture and maintenance of a variety of wooden imple-

ments and woven pieces in addition to food-process-

ing activities.

All of the edge-modified tools were made from

chert. A thick tool is considered to be any specimen

thicker than 1.0 cm.


Thick/Thin Beaked Tools. (N = 2) Beaked

tools can be manufactured from either cobbles/

pebbles or flakes. A key attribute is the presence of

one or more lateral projections created, typically, by

minimal percussion retouch. These beaks or projec-

tions usually represent a significant portion of one or

more lateral edges of the tool. Use-wear can be

present on the tip of the projections or in the con-

cavities between them. These tools were used in a

variety of different tasks such as scraping, cutting,

and engraving tasks.

Sites 41BX528 (Cat. No. 59) and 41BX863 (Cat.

No. 19) produced thick beaked unifaces. They were

both plano-convex in cross section and ovoid to ir-

regular in planview. One was manufactured from a

decorticated flake, the other from a cortical flake.

There is percussion flaking along the edge and light

to moderate use-wear on both tools.

Thick/Thin Notched Denticulate or Denticu-

late Tools. (N = 7; Figure 10.15) This category is

composed of two different, but morphologically simi-

lar implements. Denticulates have small, multiple

notches along one or more edges, while notched

denticulates have a single larger notch along one

edge. The denticulate edge is characterized by small

regularly spaced concavities that create a serrated

appearance of the edge. Denticulates are used in

various cutting or sawing tasks requiring a toothed

edge. Use-wear on notched denticulate flakes is typi-

cally located in the notches and is characterized by

moderate rounding and unifacial step fractures as-

sociated with scraping hard materials.

Of the five notched denticulates in the sample,

only one is considered thin (site 41BX528, Cat. No.

61). This decorticated flake has a single notch along

one margin that has been shaped by percussion and

pressure flaking, as well as use-wear. The other four

notched denticulates (site 41BX526, Cat. Nos. 11

and 18; site 41BX534, Cat. No. 36; and site 41BX544,

Cat. No. 4) are ovoid to irregular in shape with more

invasive flaking, as well as evidence for more inva-

sive use that is not restricted to the outermost mar-

gins. Three of these are cortical flakes. One flake

tool (Cat. No. 18) was possibly used as a spoke-

shave.

Two denticulate tools were collected from sites

41BX528 (Cat. No. 51) and 41BX863 (Cat. No. 22).

Both were manufactured from thin flakes; one is a

decorticated flake and the other is a blade. Pressure

and percussion flaking are evident along one mar-

gin.

Thick/Thin Endscrapers. (N = 35; Figure

10.16, 10.17a–c) Endscrapers are typically manu-

factured from flakes and have one or more retouched

edges. Retouched edges are usually quite uniform in

contour and may be convex, concave, or straight.

The distal or proximal end of the flake or tool blank

is often the location of this retouch. Size is variable

and may be related to heavy- and light-duty scrap-

ing tasks or hard and soft materials. Hafting or other

modes of prehension wear are sometimes observed

opposite the scraping end.

A total of 35 specimens was classified as end

scrapers. These are presented in Table 10.1.

Planview shapes range from triangular to ovoid and

elongate. Most of these tools (n = 21) can be char-

acterized as light-duty tools used on soft materials.

Eight of these are less than 1 cm in thickness. The

other 15 scrapers probably were used on harder

materials such as wood or bone.

Thick/Thin Side Scrapers. (N = 3; Figure

10.18a) The morphological and technological char-

acteristics of these types are similar to end-scrapers

except that retouch is along one or more lateral edges.

Three thick side scrapers were recovered from


2), and 41BX669 (Cat. No. 1). The blanks for these

side scrapers were cortical flakes. These ovoid to

elongate tools displayed percussion flaking and

usewear. Two of these can be characterized as

unhafted scrapers used on hides or some similar soft

material. The third scraper probably was used on a

harder materials such as antler or wood.

Thick/Thin Multi-Edged Tools. (N = 5; Fig-

ure 10.18b) Morphology, technology, and function of

artifacts assigned to this category can be quite vari-

able. This category includes artifacts that often dis-

play combinations of various other implement types.


Examples include flakes with cutting/scraping or

scraping/adzing wear, tools with multiple edges with

scraping wear, etc.. Artifacts were assigned this

designation because they could not be assigned an-

other category. One edge is often modified.

A total of five multi-edged scrapers was col-

lected from sites 41BX528 (Cat. Nos. 30, 32 and

63), 41BX833 (Cat. No. 1), and 41BX863 (Cat. No.

24). These cortical flakes are irregular in shape with

percussion and pressure flaking, as well as usewear,

along the unifacial margin. Tools were used for both

light and heavy activities on variable materials.Figure 10.15. Notched denticulate or denticulate tool

from site 41BX526.

Figure 10.16. Example of an edge-modified flake from the project that was probably used as an end scraper.


Thick/Thin Nibbled/Utilized Tools. (N = 51;

Figure 10.19) Nibbled or utilized tools are often un-

modified or only minimally modified flakes or other

pieces of raw material. They are considered a form

of expedient tool. A few implements may show some

very minor edge damage, but it is not substantial

enough to alter the outline of the initial piece or flake.

Typically, utilized flakes were expediently produced

through a variety of flake production techniques that

may have included hard-hammer, soft-hammer, or

Figure 10.17. (a–c) Example of selected edge-modified flakes from the project that were probably used as endscrapers.


indirect percussion. The goal was to produce a ser-

viceable flake that could be used with minimal trim-

ming and shaping for immediate tasks. Function is

quite variable, but commonly includes cutting and

scraping. Identification of nibbled/utilized tools is

usually via microscopic usewear analysis, although

a first rough sort often can be performed on the ba-

sis of flake size and shape.

A total of 51 tools was categorized as nibbled/

utilized (Table 10.2). Of these, 26 were thicker than

Table 10.1. Thick and thin endscrapers.

Site Catalog Type Thick

(mm)

41BX526 12 Thick end scraper 17.5







41BX528 27 Thin end scraper 8.6


41BX528 29 Thick end scraper 12.


















41BX670 1 Thin end scraper -

41BX670 2 Thick end scraper -

41BX670 3 Thin end scraper -






Figure 10.18. Selected edge-modified flakes: (a) side

scraper; and (b) multi-edge tool, site 41BX568.

Figure 10.19. Example of nibbled/utilized edge-modified


1 cm. Thick flakes are typically angular or blocky

and have significant dorsal cortex with relatively few

dorsal flake scars, which suggests that they were

removed from nonspecialized cores. All flakes have

regular, non-denticulated, unifacially, or bifacially

modified edges. Usewear is light and consists pri-

marily of various combinations of rounding, smooth-

ing, and blunting.

Burin Spall Core. (N = 1) It is often difficult to

determine if these artifacts are cores for the pro-

duction of burin spalls or are tools used in a variety

of scraping tasks. These small artifacts usually were

manufactured from flakes or other types of ex-

hausted tools, occasionally bifaces and projectile

points. The identifying characteristic used to assign

tools as burins is the presence of one or more burin

spall facets. Usually a suitable flake was proximally

or distally unifacially retouched to create a striking

platform for spall removal. The first or primary spall

was removed by direct percussion along one lateral

edge of the flake, creating a flat facet. The chisel-

like edges and durable spall facet ridges were at

times utilized in working medium to hard materials

such as wood, bone, or antler. Occasionally, a burin

exhibits several such spall removals and resembles

a small microblade core.

A single burin spall core was collected from site

41BX916 (Cat. No. 3). A decorticated core, it is

plano-convex in cross section. This specimen has a

step/hinge fracture termination where the burin spall

removal was unsuccessful and twisted onto the dor-

sal surface rather than following the lateral edge.

This specimen may have been used as a scraper.

The ventral surface also displays modification along

the lateral edge, opposite from the burin scar, that

may be related to hafting.

Other Specialized Tools. (N = 7) This cat-

egory is left for assigning unusual specimens or one-

of-a-kind forms that cannot be included in the broad

categories defined above. Assignment here was done

on a case-by-case basis only after other possibilities

were eliminated.

Cobble Chopper/Cobble Tool. (N = 3; Fig-

ure 10.20a) Cobble tools are manufactured from

medium pebbles or medium cobbles. They frequently

Table 10.2. Thick and thin nibbled/utilized flake tools.

Site Cat. No. Type Thickness (mm)

41BX525 1 Thin 8.0

41BX526 13 Thick 17.9

41BX526 14 Thick 13.1

41BX526 15 Thick 19.3

41BX526 16 Thick 22.2

41BX526 17 Thick 10.0

41BX528 11 Thin 9.6

41BX528 21 Thick 24.3

41BX528 34 Thin 8.9

41BX528 35 Thick 11.1

41BX528 46 Thick 10.0

41BX528 49 Thick 21.3

41BX528 50 Thin 5.2

41BX528 56 Thin 8.6

41BX528 57 Thin 9.5

41BX528 60 Thick 14.6

41BX528 62 Thin 3.4

41BX528 64 Thin 5.4

41BX528 65 Thin 9.7

41BX528 66 Thin 6.8

41BX528 67 Thin 8.9

41BX528 68 Thick 15.0

41BX534 37 Thick 12.9

41BX534 39 Thin 5.5

41BX534 142 Thick 11.9

41BX537 37 Thick 12.9

41BX538 5 Thick 11.3

41BX538 6 Thin 5.4

41BX539 8 Thick 20.4

41BX540 1 Thin 5.2

41BX544 1 Thin 2.7

41BX544 6 Thin 7.1

41BX554 10 Thin 8.8

41BX568 1 Thick 28.5

41BX670 10 Thick 17.8

41BX793 124 Thick 10.7

41BX863 21 Thick 18.0

41BX863 23 Thick 11.1

41BX872 240 Thin 6.1

41BX872 14 Thin 8.2

41BX872 17 Thick 14.6

41BX872 18 Thin 4.9

41BX872 13 Thin 4.2

41BX916 7 Thick 10.0

41BX916 4 Thick 10.2

41BX916 2 Thick 11.9

41BX916 6 Thick 17.7

41BX916 5 Thin 8.5

41BX916 1 Thin 4.2

41BX916 8 Thin 2.6

41BX916 9 Thin 3.2


Drills/Perforators. (N = 2; Figure 10.20b)

Morphology and use-wear were used to assign indi-

vidual artifacts to this category. These flake drills/

perforators are more asymmetrical and unpatterned

than their bifacial equivalents. Collected from sites

41BX664 (Cat. No. 427) and 41BX670 (Cat. No.

256), one is made from a decorticated flake and the

other from a cortical flake, respectively. Both drills

exhibit percussion and use-wear.

Flake Possibly Used for Notching Practice.

(N = 1; Figure 10.21) It is not certain if the notched

flake from site 41BX528 (Cat. No. 52) represents

an actual tool. This broken decorticated flake is

plano-convex in cross section and has a unifacial

notch created by small percussion flakes on the dis-

tal end. Usewear is evident along one lateral mar-

gin, but not in the area of the notch.

Backed Microtool. (N = 1) Collected from

site 41BX347 (Cat. No. 4), this decorticated flake

appears to be the proximal end of a small broken

blade. It is pressure flaked along one lateral margin;

use-wear is evident along the other. It may have

been hafted.

Undetermined. (N = 3) Three unifacial arti-

facts (site 41BX528, Cat. No. 44; site 41BX544, Cat.

No. 8; and site 41BX568, Cat. No. 12) could not be

classified with any of the categories described above.

Neither type nor function could be determined. Cat.

No. 44 is a unifacially resharpened decorticated flake.

Cat. No. 8 is the proximal fragment of a decorti-

cated blade with usewear along one lateral margin.

Cat. No. 12 is a cortical flake that may have use-

wear modification on one margin; percussion flak-

ing was used to remove tiny flakes from the dorsal

surface.

Groundstone (N = 13)

The groundstone assemblage reflects a variety of

implements including hammerstones, grinding slabs,

perforated stone, notched stone, ochre, and

slickstones. Although not abundant, there is consid-

erable technological variability represented. Each

basic category is discussed below.

have one or more edges retouched by coarse hard-

hammer percussion, either unifacially or bifacially.

Usewear and technological studies of similar imple-

ments may have consisted of both hafted and

unhafted versions (Dickens 1994:126). Although

usewear and edge morphology is not distinctive

enough to allow for reliable functional inferences,

these large core tools may have been used for cut-

ting, ripping, tearing, or chopping tasks involving plant

or animal materials (Dickens 1994; Hayden 1977,

1979).

Cobble choppers were recovered from sites

41BX554 (Cat. No. 18), 41BX793 (Cat. No. 74),

and 41BX853 (Cat. No. 3). All three are ovoid in

shape and have been bifacially modified through

percussion flaking. Use-wear is evident on two of

the choppers (Cat. Nos. 18 and 74). Cat. No. 74 has

battering on the proximal end, which is the chopper;

there is also battering on the distal cortical end where

it was used as a hammerstone.

Figure 10.20. Selected edge-modified flakes: (a) cobble

chopper; and (b) drill/perforator.


Hammerstones. (N = 3; Figure 10.22a) None

of the three chert hammerstones were heat treated;

two are complete. All three implements represent

lightly used specimens. Two probable hammerstones

were collected; one fragment was recovered from

site 41BX528 (Cat. No. 69) and a complete one was

recovered from site 41BX568 (Cat. No. 11). The

other complete specimen came from site 41BX527

(Cat. No. 14). Cat. No. 14 was a tested cobble that

was later used as a light-duty hammerstone. There

is light battering along one edge and the bottom of

the cobble.

Grinding Slab Fragments. (N = 2; Figure

10.22b–c) Both fragments were recovered from site

41BX793 and were manufactured of Wilcox Group

sandstone (Cat. Nos. 71 and 143). The abrasive

character of use-wear is different for each speci-

men. Cat. No. 71 has a slight polish and grain

faceting, while Cat. No. 143 has an abundant sur-

face polish and multidirectional striations indicative

of a rotary motion. Both fragments were initially

employed as grinding slabs; when broken the frag-

ments were recycled into thermal features.

From site 41BX831, a rather complete grinding

slab manufactured from the same raw material was

recovered from a backhoe trench in the lower por-

tion of the Elm Creek paleosol and dated roughly to

8200 B.P. (Thoms et al. 1996:27, Dockall and Pevny

2007). This specimen is significant because it docu-

ments a rather long history of use of the Wilcox sand-

stone as a source of raw material for expedient grind-

ing implements in the Applewhite Reservoir project

area. A similar range of grinding implements of both

sandstone and limestone were identified during CAR-

UTSA survey and testing in the Applewhite Reser-

voir project area (McGraw and Hindes 1987:335–

336).

Notched Tabular Sandstone. (N = 1) A single

artifact was recovered from site 41BX833 (Cat. No.

3). The material appears to be fine-grained Wilcox

Group sandstone. The edges of this rectangular frag-

ment are smooth, perhaps abraded. A bifacial notch

was flaked by direct percussion along each long edge,

and the concavity of the notch was dulled by batter-

ing. There is no visible abrasion or other wear on

either surface. Although the function of this artifact

is not known, morphological analogy would suggest

some type of weight, perhaps for netting. The envi-

ronmental setting of this site is along the upland edge

of the Applewhite Terrace on the east bank of the

Medina River and suggests the possibility that this

notched artifact was used to recover riverine or up-

land resources (as in seine nets for fish or casting

nets for fowl). The author is not familiar with any

other occurrences of similar artifacts in the study

area.

Ochre (N = 1) This specimen is a small hema-

tite nodule, fractured on one end. One surface has

small linear scratches that are randomly oriented and

overlapping. These striations may indicate the use

of this mineral as a pigment source. It was recov-

ered from the surface near Test Pit 5 at site 41BX669

(Cat. No. 4).

Limestone or Sandstone Pebble. (N = 1)

This is a small water-worn, weathered pebble with

no apparent alteration or wear. The pebble is oval in

shape and is slightly burned. It was recovered from

site 41BX539 (Cat. No. 1).

Battered/Smoothed Stone or Slickstone. (N

= 1; Figure 10.22d) A single artifact was recovered

from the surface of site 41BX669 (Cat. No. 3). This

artifact is a spherical, very smooth chert cobble. The

entire surface of the stone exhibits overlapping cone

Figure 10.21. Example of an edge-modified flake used

for notching practice.


Figure 10.22. Selected groundstone tools from the project area: (a) hammerstone; (b–c) grinding slabs; and (d)

battered/smooth stone or slickstone.

fractures and ring cracks that may have resulted

from use as a hammerstone. It stands apart from

that category because of subsequent smoothing that

may have resulted from grinding. Three areas of

smoothing may be the result of wear. The exact func-

tion of this implement is unknown.

Perforated Stones. (N = 2) This category in-

cludes two small pebbles with a hole just above the

center. One is sandstone and one is limestone. The

hole in each specimen appears to be natural because

there is no microscopic evidence of drilling. Although

there is no wear from suspension, this does not ne-


gate their use in ornamentation. Both are heavily

weathered and the sandstone specimen appears

burned. The limestone specimen was recovered from

site 41BX554 (Cat. No. 13) and the sandstone speci-

men from site 41BX670 (Cat. No. 68).

Shell Artifacts. (N = 2) One fragment of

Quadrula sp. exhibits a notch along the shell edge

near the umbo and was recovered from site 41BX872

(Cat. No. 3). The notch along the edge of this speci-

men resembles similar notched freshwater mussel

shells recovered from prehistoric sites in south and

central Texas (Dockall and Dockall 1996) and may

represent an expedient form of cutting implement. A

second specimen was recovered from site 41BX853

(Cat. No. 1) and represents a fragment of a marine

shell pendant with a drilled suspension hole.

Conclusions

Analysis of lithic material collected from the

Applewhite Reservoir project area demonstrates

evidence of the entire prehistoric sequence from the

Late Paleoindian to Historic periods. Clearly, the

sample is dominated by Archaic materials followed

by Late Prehistoric collections. The lithic assem-

blages from both Archaic and Late Prehistoric sites

include a comprehensive range of tool manufacture,

as well as evidence for the deliberate thermal alter-

ation of raw material prior to manufacture.

Lithic Technology and Technological Organization: Results of Lithic Analysis

Documents