Archaeological Investigations at the Pine Snake Site, an Allen Phase Settlement on Flat Creek in Northwestern Cherokee County, Texas

Repository Citation Perttula, Timothy K.; Nelson, Bo; Walters, Mark; and Feathers, James (2013) "Archaeological Investigations at the Pine Snake Site, an Allen Phase Settlement on Flat Creek in Northwestern Cherokee County, Texas," Index of Texas Archaeology: Open Access Gray Literature from the Lone Star State: Vol. 2013 , Article 43. https://doi.org/10.21112/.ita.2013.1.43 ISSN: 2475-9333 Available at: https://scholarworks.sfasu.edu/ita/vol2013/iss1/43


INTRODUCTION
The Pine Snake site is a re<:ently dis<:overed late 17'h to early IS'h century Caddo Indian archaeological site located on private land in the northwestern part of Cherokee County, Texas, in the valley of a westwardflowing tributary to the Neches River.This is an area of the Pineywoods of East Texas that contains extensive numbers of Caddo archaeological sites along all major and minor streams (see Anderson et al. 1974;Kleinschmidt 1982;Perttula and Nelson 2007, 2009a, 2009b;Walters and Perttula 2012).Post-A.D. 1400 Frankston phase and post-A.D. I 650 Historic Caddo Allen phase sites, especially cemeteries dating to either phase.are particularly abundant in th is part of East Texas (Cole 1975;Kleinschmidt 1982; Marceaux 2007Marceaux , 2011;;Perttula 2007a. 2007b, 200R. 2010, 20 II).This article summarizes the findings from archaeologi<:al investigations we completed at the Pine Snake site in late 2008.They have produced important information on the domesti<: an.:haeological record at a well preserved Allen phase habitation site.

SETTfNG
This historic Caddo site has an spatially extensive archaeological deposit.It extends for more than 600 111 along a narrow ridge top and ridge slope (370-400 feet amsl), as well as a swale or saddle.reminiscent of the setting of the nearby Kah-hah-ko-wha site (41CE354) on Flat Creek (Perttula and Nelson 2007), and a probable spring-fed intermittent drainage area, now dry (Figure l).The current channel of Flat Creek is more than 800 111 to the north; a seasonal tributary to Flat Creek also lies ca. 1 SO m to the east and southeast.
The vegetation on the ridge top and ridge slope landforms is a pine plantation that is approximately 10 years old.Consequently, surface visibility is less than 10% due to a thick understory of brush and vines.There is evidence of ridge rows of soil on the site, evidence of the clearing preparations in advance of planting the pine trees in rows.

WORK COMPLETED AT THE SITE
Thirty-nine shovel tests were first excavated on the ridge top and ridge slope to define the spatial limits of the site (see Figure 1 ).A total of 29 of the shovel tests excavated at the Pine Snake site were found to contain archaeological materials that occurred between 0-60 <:m bs in Nacogdoches fine sandy loam sc.diments (see Mowery and Oakes 1959).These we.re principally ceramic vessel sherds (64%) that were concentrated in two areas at the northern (ST 576 and ST 581) and southern (552)(553)(554)(555)(557)(558)(559)(560)(561) ends of the site (see Figure 1).These con<:entrations surely relate to archaeological evidence of two or more His-tori<: Caddo farmstead compounds preserved at the Pine Snake site.Other shovel tests were noted to contain lithic artifacts and animal bone.with most of the latter found to be in the southern ceramic concentration.The estimated spatial extent of the site is 20.100 m 2 , or approximately 5.0 acres.
During the course of shovel testing, one shovel test (ST 558) in the southern ceramic con<:entration (see Figure 1) encountered a lens of darker soil, a mussel shell fragment, and a high number of ceramic sherds between 0-40 em bs.Another shovel test (Unit 778), rectangular-shaped and about 70 x I 00 em in size to perrnit a better examination of the profile.was ex<:avated adjacent to it (Figure 2).Encountered in Unit 778 was a zone of ash between 32-40 em bs, followed by a hard.compact oxidized soil between 40-46 em bs .,..,..-... --...... i ,

I
.
The overall density of artifacts noted in the positive shovel Lesls (excluding the irregular-sized Unit 778) is considerable, at 2.69 per positive shovel test, or ca.21.5 artifacts per m 2 • In tem1s of the ceramic sherds, the highest densiLies noted in the shovel tests included 14 sherds in ST 558.nine sherds in ST 555, and eight sherds in ST 550, all in the southern concentration at the site (see Figure 1).Also noted in this same area were several Perdiz arrow point fragments (see below).
Subsequent to the completion of the shovel testing, additional controlled excavations were done .in the southern part of the Pine Snake to better define the archaeological character of the hearth fealure, as well as to examine the archaeological deposits in more detail around the hearth .The additional work included the excavation of five I x I m units (Units 782-786) around ST 558 and Unit 778 (see Figure 2) as well as a ca.70 cm x l m unit (Unit 781 ).Various line-screen samples were taken from Units 781-783 and the post hole features exposed in the work (see below and Appendix 1).Additionally, a 40 x 40 em line-screen unit was excavated just south of ST 558 (see Figure 2).This work exposed a large hearth feature (Feature 1) that most likely represents a central hearth inside a Caddo domicile.along with seven post holes, as well as an abundance of Historic Caddo Allen phase artifacts discarded in and around the hearth.
TI1e archaeological deposits in this part of the Pine Snake site range between 30-50 em in thickness overlying a red clay subsoil (Figure 4).The uppermost deposit (Zone 1) is a dark yellowish-brown ( IOYR 3/4) sandy loam A-horizon that ranges from 12-40 em in thickness.Outside the area of the Feature I hearth, the second soil zone (Zone 2) is a strong brown (7.5YR 4/6) sandy loam E-horizon about 15 em in thickness.Below this is a red (2.5YR 4/6) clay B-horizon (Zone 3) (Figure 4).In the area of the Feature I hearth,lying below Zone l is a ca.10-20 cm thick very dark grayish-brown (lOYR 3/2) sandy loam (Zone 4) occupational deposit that rests directly above a lens of ash and an oxidized deposit, but this charcoal-stained sediment also extends in a 50-80 em wide band outside of, but adjacent to, the oxidized basin (see Figure 2).Below Zone 4 is a thin gray ash lens (Zone Sa), between 3-5 em in thickness, that accumulated through hearth use atop the oxidized basin.The oxidized basin sediments (Zone 5b) are a hard baked red (2.5YR 4/8) sandy loam that are a maximum of 16 em in thickness (see Figure 4).The oxidized basin appears to have been excavated into the Zone 3 B-horizon approximately 12 em in the deepest part of the hearth basin.

FEATURES
The features documented in the limited archaeological investigations at the Pine Snake site include the Feature 1 hearth and seven post holt:~ (PH 1-7) (see Figure 2).One of the po~t holes (PH 6) intrudes into the Zone Sa ash deposits in the hearth (see Figure 4), and clearly postdates its accumulation.The top of PH 7 was exposed underneath Zone 4 deposits, but is coeval with the Feature I ash and oxidized basin deposits (see Figure 4), and thus it is not a central structure post; our excavations did not identify a central post associated with Feature l.
Feature 1 consists of several assm:.iateusediments.First is an intensely oxidized red sandy loam (Zone Sb) hearth basin that is approximately 1 W em (north-south) x 180 em (east-west) (see Figure 2).1n the center of the hearth, the basin extends from 31-48 em bs (see Figure 4).The thickness and hard-baked character of the basin indicates that the hearth was repeatedly used and re-used for a number of years by the Caddo peoples living at the Pine Snake site.
In the approximate center of the Zone Sb basin is a concentrated 60 x 65 em area of gray ash (Zone Sa).The top of the ash is exposed at 28 em bs (see Figure 4 ).Above this is the Zone 4 deposits (13-28 em bs) of very dark grayish-brown sediments (with charcoal flecks and staining) that are likely to have heen raked out of the hearth to clean and empty the basin.This band of charcoal-stained sediments is present to the immediate north.west, and south of the oxidized basin (see Figure 2).
The Feature I hearth was probably located inside a structure, although the excavations were not sufficiently intensive or of a large-enough scale to identify a structure wall post pattern.Such a hearth.with a concentrated and well-preserved ash deposit in the central part of the basin, would have been "used almost exclusively for warming, as suggested by white ash smoldering fires.which are less suitable for cooking than the higher-intensity outdoor fires with their constant flow of oxygen to fuel the flames" (Shafer 2003:37).
Post holes PH 1-5 are arranged in a semi-circle along the northern and western parts of the Fe.a~ureI hearth.Given their placement, and their generally small size (Figure 5), they may mark the construction of a scree.lilalong one side of the heat1h while it was in use .. The exposed tops of the post holes are at 33-35 em bs, and they extend to 38 ern bs (PH 5), 41 em bs (PH 4), 44 em bs (PH 2 and PH 3), or 47 em bs (PH 1).Only PH 1, PH 4, and PH 5 were set in the Zone 3 clay B-horizon.The post hole fill is either a dark grayish-brown sandy loam with charcoal tlecks (PH 1, PH 3), a dark yellowish-brown (IOYR 4/4) sandy loam with charcoal flecks (PH 2), or a dark brown (IOYR 3/2) sandy loam with no charcoal flecking.With the exception of PH 1, which is 19 em in diameter and a typical size for a wall or smaller interior post support of a Caddo house.post hole diameters range from 10-13 em, much too small for them to have likely served as wall posts.
PH 6 intersects the Zone Sa ash in Feature I , and postdates the accumu~ation of the ash (see Figures 2 and 4).This post, with straight sides and a flat bottom.is approximately 19 em in diameter, and extends from 2R-37 em bs.Its fill is a very dark grayish-hrown sandy loam (Zone 4 ).
PH 1 PH 7 is a !luger post feature (26 em in diameter), probably a support post, that had been placed along the western side of the oxidized basin.probably cutting through it, or less likely set next to it when it was in use (see Figures 2 and 4).The fact that the post was first recognized only below the Zone 4 hearth debris suggests it had been placed at approximately the same time as the hearth was in use; hearth sediments were incorporated into the fill.This post has sloping sides and a rounded bottom, and extends from 35-66 em bs.The fill of this post is primarily a strong brown sandy loam (Zone 2) that is mottled with very dark grayishbrown sandy loam (Zone 4) and chunks of oxidized soil.

THERMOLUMINESCENCE DATING, by James Feathers
Two ceramic sherds and a chunk of the intensely oxidized Feature 1 hearth basin were submitted to the University of Washington Luminescem:e Dating Laboratory for thermoluminescence dating (Feathers 2010).The two sherds-both parallel brushed cooking jar body sherds-are from Unit 785 (24 em bs) and Unir 786 (28 em bs).The sample of oxidized soil from Feature I is from 39-41 em bs in Unit 785.Dose rate measurements were made on each sample as well as an associated sediment sample.Due to time limitations, measurements on three of the sediment samples did not get completed, but variation in radioactivity among them was not great, so an average from the same site was used for nge calculation fur those particular samples.The difference in radioactivity between sherds and associated snlimenls was also not great in most cases, suggesting the ceramics were made from similar material as the sediments.Dose rates were determined using alpha counting and flame photometry.The bera dose rate calculated from the~e measurements was compared with the beta dose rate measured directly by beta counting.These were in agreement for nine of the samples, but differed for the other seven.In the latter case. the beta dose rate from beta counting was used for age calculation.The cause of the discrepancy in beta dose rates is not certain, but it could relate to disequilibrium in the uranium decay chain.Moisture content was esrimnted as 80 ± 20 % of saturated value for the sherds, and I 0 ± 5 percent for the sediments.Table 1 gives all relevant data, including the total dose rate for each sample.Equivalent dose was determined by TL, IRSL and OSL.The TL measurements, in general.were characterized by poor plateaus (Table 2).In only two samples did the plateau (region of constant equivalent dose) extend beyond 320°C, symptomatic of rclativcly low firing conditions.ln half the samples there was some sensitivity change with second glows.All but three growth curves were fit with linear functions.Anomalous fading was ubiquitous, and apparent in all but three samples (Table 2).OSL was measured on 5 or 6 aliquots per sample.The OSL signal was generally strong.at least 10 times the intensity of the lRSL signal (in the SAR protocol).This is typical for ceramics and because A glow refers to luminescence as a function of temperature; a second glow comes after heating to 450°C.**g-value is the fading rate expressed as %per decade, where a decade is a power of 10. lRSL mainly stems from feldspars, which art: involved in anomalous fading, this probably means the OSL signal does not fade appreciably.Weak signal allowed equivalent dose for fRSL to be measured om more than two aliquots for only o ne of the samples.Equivalent dose values are given in Table 3.The equivalent dose values differ among TL, lRSL and OSL, which is not surprising given that the b-values (which reflect the lower efficiency of alpha irradiation in produc ing luminescence) also differ.However, the OSL equivalent dose is unusually low, when compared with the TL and lRSL values, for most samples , the s ignificance of which will be taken up next.The b-values are also given in Table 3.These are fairly typical values, although the TRSL values are based on limited data due to a weak signal.The IRSL/OSL b-values were not measured for UWI 964, due to an oversight; an average value was used for OSL age calculation.tht: additive dose intercept was used for equivalent dose.
Table 4 gives the ages calculated separately for TL, lRSL, and OSL for each sample.For samples with evide nce of fading of the TL signal, the correction procedure of Huntley and Lamothe (2002) was applied.In some cases, this correction did not produce a significantly older age, although part of the reason for this may be low precision in the fading data.No fading tests were done for either IRSL or OSL (because of exorbitant amount of machine time required) , so no correction can be applied to them.
The IRSL signal mainly comes from feldspars, which often fade, so the IRSL ages musF be considered a minimum.The weak IRSL, and therefore feldspar, signal suggests that the OSL s ignal probably comes mainly from quartz and does not fade, as mentioned earlier.4 shows that the OSL age i~ typically younger than, or equivalent in value to, the TL and IRSL ages.lt is dearly younger for two samples from the Pine Snake site, and for UW1965, it is probably younger because the TL age is underestimated due to insignificant fading correction because of poor precision.On one sample (UW1966), the OSL age is older, even when fading of the TL signal is considered.
The younger OSL ages are unusual.The OSL is usually older, if anything, because of fading for Tl and IRSL.A possible reason is the low original firing temperature of the pottery.It is well known in the ~•asc of light exposure that the traps associated with TL and IRSL do not bleach as rapidly as they do for OSL.It is possible the original firing of the pottery was not sufficient to deplete the TL and IRSL signals to the s<.~me extent as the OSL signal.In other words, the TL and IRSL still contain a residual signal from the raw material.That would make the OSL ages the most likely ages for the manufacture of the pottery.but OSL is also known to contain slower bleaching components, so it is possible tha[ a residual is present ev~:n with the OSL.
Assuming the OSL is the best estimate in most cases, Table 5 gives the final ages for the P111e Snake samples.Some additional comments: the OSL age for UWl966 seemed unreasonably old, so the TL w<~' taken as the best estimate.

OSL OSL TL
The ages for samples from Pine Snake are all considerably older [han expectations, as they range from A.D. 1380-1550 at one standard deviation and A.D. 1330-1630 at two standard deviations, as compared to the post-A.D. 1650 temporal estimate derived from the analysis of the recovered ceramics from the site (see below).Insufficient heating may he causing the age overestimation.Many of the ceramics were reported as "soft'' hy the students preparing them for measurement.I do not consider this a fully satisfactory explanation (the last batch of Caddo ceramics we processed [from the Lang Pasture site in Anderson Coull[y, Texas, see Feathers 20 II] also had poor plateaus indicative of low heating, but the OSLand TL agreed for most of them and the ages were not known to be overestimates), but I can think of no other reason why the OSL ages should be younger.A possible test could be coarse-grain analysis on UW 1964. the sample of burned L:arth from the hearth feature.Coarse grains of quartz were isolated from this sample.It should be possible to measure only the fast component of these quartz grains -the ones most likely reset to zero -to see if an age more in line with expectations can be obtained.

RECOVERED ARTIFACTS
More than 580 artifacts arc in the assemblage of recovered material culture remains from the Pine Snake site.The majority of the artifacts are ceramic sherds (n=462) from plain, utility ware, and fine ware vessels.There are also engraved ceramic elbow pipe sherds (n=3) and burned clay pieces (n=64) in the collection.Lithic artifacts include seven arrow points or arrow point preforms, a flake tool, a manu fragment.47 pieces of lithic debris, and one core.
Plotting the distribution of artifacts recovered in situ al the site, with one exception (a single sherd at 22 em bs), all of these artifacts (n=45) are found between 24-36 em bs.The highest densities of artifacts by depth occur between 27-28 em bs ( 18%), 30 em bs (11 %), 32 em bs (11 %), and 35-36 em bs (27%).These artifacts arc at depths that indicate they are from the large hearth or in deposits immediately above and/or coeval with the hearth (i.e., they are in deposits raked out of the hearth and deposited adjacent to it).Mussel shell pieces are particularly abundant below 31 em bs, in the lower part of the basin hearth.

Ceramic Sherds
A total of 462 ceramic sherds were noted in the various positive shovel tests and excavated units at the site.This comprised 157 plain rim, body, and base shcrds and 305 decorated rim and body sherds, among them several Patton Engraved sherds, including a rim from ST 778 ( 10-20 em bs).The plain to decorated sherd ratio is 0.52, consistent with an Historic Caddo occupation in the upper Neches River basin (Perttula and Nelson 2007:Tahle 4; Walters and Perttula 20l2:Table 31) as well as the latest Frankston phase (ca.A.D. 1600-1650) Caddo ceramic assemblages (see below).
Most of the decorated sherds have brushing on them, either as the sole form of decoration (n=253), or in combination with punctated (n=3) and incised (n=4) lines; these are all from utility ware jars (i.e., cooking jars).The brushed sherds that were noted include vertical brushing on tbe rim (n=S); horizontal brushvng on rims (n=4); horizontal brushing on the rim and vertical brushing on the body (n=1); parallel brushed (n=208); overlapping brushed (n=l2); vertical brushed on the body (n=18); and opposed brushing (n=S).Overall, 85% of the decorated sherds from the site have brushed marks on e.ither the rim and/or the body of utility ware jars; this is one of the highest proportion of brushed sherds documented to date in any Caddo site vn the upper Neches River basin other than a series of historic Caddo sites on nearby Stone Chimney Creek (see also Perttula and Nelson 2007:Table 4;Walters and Perttula 20 12:Table 31).
Other decorated utility ware sherds include brushed-punctated sherds.These include vessels with a tool punctated rim and horizontal brushing on the body (n=2), ns well as one body sherd with parallel brushing , and a row of tool punctates pushed through the bmshing, and parallel brushed-overlapping and/or diagonal incised lines over the brushing (n=2).There are also two Spradley Brushed-Incised body she.rds, a new utility ware ceramic type documented in Historic Caddo contexts in the Neches-Angelina river basin (Marceaux 2011).These sherds have parallel bmshing overlain with closely-spaced perpendicular incised lines.
Other utility ware sherds are present at the Pine.Snake site, although in low proportions (as is also the case at the Kah-hnh-ko-wha site 141CE354], see Perttula and Nelson [2007:Table 5]).There.are five shcrds (including two rims) from LaRue Neck Banded jars, and two lxxly sherds with closely spaced pinched rows from Killough Pinched jars.
A Maydelle Incised rim with triangular incised zones tilled with punctations was found in ST 554 (20-40 em bs), and there arc two body sherds with tool punctated rows, and five lxxly sherds with parallel or opposed incised lines.Another body sherd has a vertical appliqued fillet and opposed incised lines, and one body sherd has a straight appliqued fillet.A third appliqued sherd in the decorated shcrd assemblage has a straight appliqued fillet with a row of tool punctations placed adjacent to the fillet.
There are a number of rim (n=3) and body sherds (n=6) with horizontal or parallel grooved ridges in the Pine Snake utility wares.Very similar decorated sherds have been recovered at several other historic Caddo sites in the Angelina and Neches River basins in East Texas-including at the Deshazo site (41 NA27) (Fields 1995:199-200 and Figure 75d-e)-and this kind of grooved utility ware has recently been dubbed Lindsey Grooved because of its abundance at the Lindsey site (41CE293) (Marceaux 2011).Although categorized as " ridged" pottery at the Deshazo site, Fields ( 1995: 199-200) dcscribe,s the decorati\< e technique as "the manipulation of the vessel surface to form shallow horizontal grooves altemati,ng with mounded ridges.The grooves are trough shaped and the ridges are not appliqued."To avoid confusion with the Belcher Ridged type, and its very different form of ridged decoration, we have chosen to emphasize the.horizontal grooves as the defining criteria for this new Caddo utility ware type.
In the fine wares at the Pine Snake site, there are five Patton Engraved rim and body sherds.These are all apparently from Patton Engraved, var.Allen vessels (see Perttula 2008:Figure 2) with horizontal rows of engraved lines with triangular tick marks.
There are 13 other fine ware sherds recovered in the investigations.These include a bottle sherd with opposed engraved lines; two sherds with multiple curvilinear engraved lines; two possible carinated bowl sherds with part of a scroll element; two possible Poynor Engraved, var.Hood sherds with hatched panel dividers (see Pcrttula 2008:Figure I); a body sherd with opposed engraved lines; and a rim to an inverted rim carinated bowl (likely a Patton or Poynor Engraved vessel, st:t: Suhm and Jelks 1962) with one horizontal engraved lint: showing on it.Two other rims have a single horizontal engraved lim: under the lip, and one of these is also lip notched; this is a rare form of lip lrt:atment in the upper Neches River basin Caddo ceramic tradition (Perttula 2011).Another sherd from a carinated bowl has a single horizontal engraved line above the carination, and horizontal brushing on the vessel body.
The last line ware engraved sherd from the Pine Snake site is a Humc Engraved.var.Allen sherd (see Perttula 2008:Figure 2).This sherd is from a bowl or carinated bowl with at least one horizontal engraved line that has a series of hatched pendant triangles.
As previously mt:ntioned, tht: Pine Snake site has one of the highest proportions of brushed sht:ni'> in any documented Caddo ceramic assemblage in this part of the upper Neches River basin (Table 6), and thus it falls into the Group 1 assemblages, tht: youngest Caddo sites in our present sample along with the ceramic sherd assemblage from the nearby Kah-Hah-Ko-Wha site (41CE354) and several sites on Stone Chimney Creek (41CE421 and 41CE429).The Group 1 sites are all identified as having Historic Caddo Allen pha&c components.The Pine Snake ceramics are tempered with grog or crushed sherds/fi red clay (Table 7).Grog is a temper in each of the sherds analyzed in detail, regardless of whether the sherd is from a plain ware, utility ware, or fine ware vessel.Less than 6% of the shen.lsfrom the site (with the highest proportions found in the plain wares) are from vessels with bone temper inclusions (see Table 7); the limited use of bone as a temper is characteristic of the upper Neches River basin Caddo ceramic tradition (Perttula 2011;see Table 6).Crushed pieces of hematite or ferruginous sandstone are also regularly used (29.9%) as a temper in the Pine Snake ceramics, especially so in the tine wares.Two sherds (2.3°,{,) have charred organic remains in the paste, suggesting they are from vessels where the clay paste was not well-cleaned before tiring and/or that they are from low-tired vessels where the organi\.: materials in the paste were not completely combusted during firing.
More than 33% of the sherds from the Pine Snake site are from vessels that have been made witti a naturally sandy clay.Each of the different ceramics wares at the site relied on roughly comparable proportions of sandy clays (i.e., sandy paste, 31.6-42.9o/o),with their greatest use seen in the engraved line wares (see Table 7).
The majority of the sherds from the Pine Snake site are from vessels fired in a low oxygen or reducing environment (Table 8).The sherds fired in this manner comprise between 57.9% (utility wares) and 71.5% (fine wares).Plain wares were generally fired in the same manner as the fine wares, as 69.5% of the plain sherds are from vessels tired in a redu\.:ingenvironment.In particular, the highest proportion of sherds are from vessels fired in a low oxygen environment.but then the vessels were pulled from the fire and allowed to cool in the open air.Depending upon the ware, between 42.9-56.YJ'c of the sherds from the site are from vessels fired in this manner; plain wares \vere most commonly fired this way./\hout 23% of the sherds are from incompletely oxidized vessels (see Table g); that is. the ve~ sels were not fired a sufficient duration ro completely oxidize the pa::;tc.Almost 5% of the sherds analyzed in detail for firing conditions are from utility ware vessels that have been either sooted, smudged.or reheated.leaving a thin dark reduced band along one or both vessel sherd surfaces.

Instrumental Neutron Activation Analysis of Selected Sherds
Six ceramic sherds and one ceramic elbow pipe sherd from the Pine Snake site (Table 9) were submitted to the Archaeometry Laboratory at the Research Reactor Center at the University of Missouri-Columbia for instrumental neutron activation analysis (INAA) (Ferguson 2009a).The purpose of these analyses is to detennine, in so far as is possible, the production locale of the analyzed sherds, and identify any ceramic sherds that arcfrom vessels made from non-local clay sources (i.e., are from a trade vessel).ln the analysis, Ferguson (2009a) novaculite (ST 558,20-40 em bs); a possible Bassett point fragment of gray chert (ST 778,20-30 em b); arrow point tip and blade fragments (n=2, ST 556,40-60 em, quartzite, ST 778,30-40 em bs, gray chert)-and a distally-retouched flake tool of gray chert (ST 565,0-20 em bs).

Ground Stone Tools
A single ferruginous sandstone mano fragment (with grinding on both surfaces) was found in Unit 785 between 20-30 em bs.This two-sided mano is at least 40 nun in length, 52 mm in width, and I R mm in thickness.

Lithic Debris
The lithic debris from the site includes flakes from several different raw materials: gray chert (n=3l), brownish-gray chert (n= I), dark gray chert (n=3), honey-colored chert (n= I), a translucent brown chert (n= l ), dark brown chert (n=l), black chert (n=2), a glassy black siliceous material (obsidian, n=2, ST 778, 10-20 em bs and Unit 785, 10-20 em bs), and quartzite (n=5).These flakes are the product of the rcsharpening or final pressure-flaking of completed or ncar-completed tools that must have been brought to the site.except for the quartzite pieces; the latter raw material occurs as pebbles in local stream gravels, and they were reduced through knapping to obtain flakes for tool use or arrow point manufacture.There is a small tested quartzite pebble from Unit 784 ( l0-20 em bs); it is 46 mm in length, 29 mm in width, and 2R mm in thickness.
Chert comprises about 85% of the lithic debris at the site, all from non-local raw material sources.These in dude Central Texas sources 150 miles or more to the west (i.e., the gray, brownish-gray.dark gray, honey-colored, translucent brown, and dark brown cherts; several flakes have a limestone-covered cortex) or Ouachita Mountains sour<.:es(i.e., the black chert) more than 100 miles to the north.
The obsidian (Texas Obsidian Project No. 234a-c) is from a long-distance non-local source, namely from the Obsidian Ridge (Cerro de Toledo rhyolite) source in the Jemez Mountains of northem New Mexico (Thomas R. Hester, April 27, 2009 personal communication).According to Ferguson (2009b), the obsidia11 pieces were analyzed by the University of Missouri Research Reactor Center with a hand-held portable Bruker XRF: the instrument has a rhodium-based X-ray tube which we operate at 40 kV and a thermoelectrically cooled silicon dete<.:tor.We calibrated the instrument using a suite of 45 well-known obsidian sources with data from previous XRF and NAA measurements.Your samples were counted for three minutes to measure the minor and trace clements present.The elements measured include Mn, Fe, Zn, Ga, Pb, Th, Rb, Sr, Y, Zr, and Nb ... The resulting data were compared to our database of obsidian sources using both hierarchical cluster analysis and bivariate plots.Due to the small size of the artifacts, the concentrations of lower atomic weight elements are inflated with the smallest samples, thus the source assignment was based primarily on elements Rb, Sr, Y, Zr, and Nb.

ANIMAL AND PLANT REMAINS
Animal bone pieces (n=8) were noted in seveml shovel tests, including ST 576 (0-20 em bs) in the northern area and ST 550 (0-20 em bs), 555 (0-15 em bs).and 558 (20-40 em bs: among them deer teeth) in the southern ceramic concentration (see Figure 1).ST 558 (20-40 em bs) also had a mussel shell fragment.Burned animal bone, wood charcoal, nutshell, and mussel shell fragments were also noted between 10-40 em bs in ST 778.They are also relatively abundant in the I x I m excavation units placed in and around the Feature 1 hearth.These remains have not been studied to date.

Figure
Figure l .Map of the Pine Snake site.

Figure 2 .
Figure 2. Plan of excavations in the southern p<trt of the Pine Snake site.

Figure 3 .
Figure 3. Looking down at the top of the oxidized wne in <1 he<1rth fe<1ture exposed in ST 778 at the Pine Snake site.

figure 5 .
figure 5. Profiles of post holes 1-5 at the Pine Snake site.

Table 1 .
TL and OSL data for samples rrom the Pine Snake site.
*Dose rate calculated for TL.It will he slightly lower for OSI.hecause of lower alpha cfticiency.

Table 2 .
TL measurements.Refers to slope ratio between tht: first and second glow growth curves. *
* Due to a 2"<1 glow growth curve with a large negative Lm ercept (not expected.from current understanding),

Table 4 .
A~e cal' cul'ations for Tl,, IRSL, and OSL for Pine Snake site samples.
*Corrected fo r fa ding using Huntley and Lamothe (2002) method.Other TL ages reflect either no measured fading or a correction that was not significantly different ti:-om the uncorrected age.** Sample fades, but correction produced intinite value Table

Table 5 .
OSLand TL final age estimates for Pine Snake site samples.
* OSLffL refers to a weighted average of the two ages **negative value reflects aBC age

Table 6 .
Comparative sherd assemblage data from Lake Palestine Caddo sites, nearby Caddo sites, and the Pine Snake site.

Table 6 .
Comparative shcrd assemblage data from Lake Palestine Caddo sites, nearby Caddo sites, and the Pine Snake site, cont.

Table 7 .
Temper and paste attributes of the Pine Snake ceramics.