YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� CHAPTER�5� excavated�24YE14�in�Gardiner, �Montana,�another�Late� Archaic�site�in�the�study�area.� These�four�sites— LATE�ARCHAIC�LITHIC� Yellowstone�Bank�Cache,�RJP� 1,�2 4YE14,�and�Little�Trail� TECHNOLOGY�AND�LAND�USE� Creek—provide�detailed�information�regarding�Late� Archaic�use�of�the�Gardiner�Valley,�Montana,�between� WITHIN�THE�GARDINER� BASIN,� 3,000�and�1,500�uncalibrated�years�ago.�W e�focus�upon� UPPER�YELLOWSTONE�RIVER,� the�lithic�technological�organization�of� Late�Archaic� MONTANA/WYOMING � hunter�gatherers�in�the� Gardiner�Basin,� results�of�which� compliment�papers�in�this�volume�by�Adams� and�Park.�� � Douglas�H.�MacDonald�and�Lester�E.�Maas� � � Archaeological�excavations�within�the�Gardiner�Basin� of�Montana�have�yielded�abundant�evidence�of�Late� Archaic�Native�American�occupation. �This�paper�reports� on�several�sites�in�the� Gardiner� Basin,�Montana�(Figure� 1),�that�provide�insight�into�hunter�g atherer� use�of� the� Yellowstone�River�near�its�headwaters�in�Yellowstone� National�Park�between�3,000 �and�1,500�uncalibrated� radiocarbon �years�ago.�Lithic �analysis�of�Late�Archaic� artifacts�indicates�active�use�of�local�Crescent�Hill�chert� and�Obsidian �Cliff�obsidian,�both�with�sources�20�2 5� miles�from�the� site.�Large�bifaces�were�transported�from� the�material�sources�to�the� Gardiner�Basin�en�route� northward�to�the�Paradise�Valley� and�beyond.�Lithic�data� indicate�that�the�Gardiner� Basin�was,�thus,�a�staging�area� for�Late�Archaic �hunter�g atherers�entering�and�exiting� Figure�1.�Project�Location �with �Sites�Mentioned� in�Text.� the�nearby �Yellowstone�Plateau�and�its�rich�and�diverse� � ecosystem.��� Background� and�Setting� Introduction� The�study�area� for�this�paper� is�the�Gardiner�Basin�of� The�Late�Archaic�was�a�peak� period�of�Native� the�Yellowstone�River,�an�approximately �15�mile�stretch� American�use�of�the� Greater�Yellowstone�Ecosystem,� of�the� river�bound�by�its�confluence� with�the�Gardiner� with�numerous�sites�along�the�Yellowstone�River�and� River�on�the�south�(in�Gardiner,�Montana)�and�Yankee� Yellowstone�Lake�yielding �Late�Archaic�Pelican�Lake� and� Jim�Canyon�on �the�north�(Figure�1).� To�the� north,�t he� Besant�projectile�points�that�date�to�between�3,000�and� Yellowstone�River�opens�up�into�the�vast�and�wide� 1,500�uncalibrated�radiocarbon� years�ago�(Sanders�2000,� Paradise�Valley,�while�to�the�south,�the� river�exits�the� 2001;�MacDonald �et�al.�2010).�This�paper� summarizes� rugged�Yellowstone�Plateau�and�the�river’s�headwaters� recent� work�at� four�Late�Archaic�sites�in�the�Gardiner� at�Yellowstone� Lake,�some� 30�miles�south.�The�area�of� Basin,�including �results�of�University�of�Montana’s� (UM)� the�study�is�mostly�within �the�confines�of�Yellowstone� work�at�the��Yellowstone�Bank�Cache� Site�(24YE355),�first� National�Park�(YNP),�but�also�includes�sections�of�the� reported�upon�in�MacDonald�et�al.�(2010).�In�addition,� Gallatin�National�Forest�(GNF).�T hree�of�the�key�sites�in� UM�excavated�L ate�Archaic�occupations�at�the�RJP�1�Site� discussion�are� within �YNP�(Yellowstone�Bank�Cache,� (24YE190)�along�the�Yellowstone �River�and�the�Little�Trail� 24YE14,�and�RJP�1),�while�o ne� (24PA1081)�is�within� Creek�Site�(24YE1081)�in�an�upland�stream�tributary�of� nearby�uplands�of�the� GNF.� � the�Yellowstone �River.�Finally,�Sanders�et�al.�(2000)� - 56-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� The�three� valley�floor�sites—Yellowstone�Bank�Cache,� surrounding�areas.��The�current�project� is�within�portions� RJP�1,�and�24YE14—are�located �on�gently�sloping�alluvial� of�the� winter�range� of�b ison,�elk,�mountain�sheep,� deer,� terraces�comprised�of�Pleistocene�gravels�with�a�thick� and�antelope�which�extends� down �valley�along�the�river� (ca.�2�m)�mantle�of�Holocene�a ge�alluvium�(Good�and� to�Livingston,�Montana,�where�the�Yellowstone�River� Pierce�1996;�MacDonald�et�a l.� 2010;�Pierce� 1973;� exits�the�mountains�(YNP� 1993).�This�winter�range�for� Rodman�et�al.�1996).�Glacial�moraines,�colluvial� large�ungulates�provided�good�hunting �for�prehistoric� landslides,�and�other�features� of�Late� Pleistocene�age� hunter�gatherers,�as�well�as�sheltered�areas�for�winter� occur�in�the�valley�floor�and�in�uplands� near� Little�Trail� camps�due�to�the�limited� snowfall�in�comparison�to�the� Creek.�Little�Trail�Creek�is�within �an�upland �stream�valley� nearby�mountains.�� bounded�by�basalt�outcrops�upslope� and�glacial�moraines� Prior�Research�� onsite.�Soils�at�the�site�are� largely�a�mix�of�Little�Trail� Overall,�Native�American�use� of�the� northwestern� Creek�alluvium� and�upslope�colluvium.�Large�glacial� Plains�and�Rocky�Mountains�increased�during �the�Late� moraine�features�are�located�adjacent�to�each�of�the�four� Archaic,� reflecting�abundant�and�diverse�resource� sites�in�the�study �area,�suggesting�some�Late�Archaic� availability�due� to�the�comparatively�moist�and�cool� preference�for�site�establishment� near�these�features,� climate�(Frison�1991;�Husted� and�Edgar�2002;�Lahren� perhaps� as�protective�landforms.��� 1976;�Sanders�2001).�Referred�to�as�the� Lamar�Valley� Various�researchers�have�established�a� baseline�of� Subphase�by�Reeves�(2006)�for�YNP�sites,�the�Pelican� paleoenvironmental�data�for�northern�portion�of�YNP� Lake�and�Besant�phases�of�the�Late�Archaic�period� (Beiswenger� 1991;�Bright� 1966;� Meyer�et� al.�1995;�Rhode� (3,000�1,500�B.P.)�was�a�time� of�active�use�of�the� 1999;�Sanders�2001;�W hitlock�1993).�These� studies� Yellowstone�ecosystem�by�Native�Americans� (Johnson� provide�a�fairly�reasonable�approximation �of�the� 2001;�Sanders�2000,�2001).� Several�archaeological�sites�in� prehistoric�environment�in�the�Yellowstone�Valley.��The� the�study�area�and�greater� Yellowstone�have�yielded�Late� Late�Archaic�period—ca.�3000�to�1500�B.P.—was�a�period� Archaic�Besant� and�Pelican�L ake�projectile�points,� of�active�Native�American�use�of�the�Yellowstone�Valley,� including�the�four�sites�discussed�in�the�current�study.� � coinciding�with��a�fairly�mesic �climatic�period�of�cooler� During�the�UM�survey�of�a�3,000�acre�portion�of�the� and�moister�conditions,�favorable�to�grasslands�in�the� Gardiner�Basin�in�2007�2008�(Maas�and�MacDonald� northern�Plains�and�vicinity.�Recent�analysis�of�botanical� 2009),�the�Late� Plains�Archaic� accounts� for�more�than� remains�and�pollen�from�Gardiner�Valley�sites�by�Gish� 50%�(n=48)�of� the�projectile�points�recovered�from�both� (this�volume)�suggests�the� presence�of�largely�modern � excavated�and�surface�contexts� (Figure�2).�Overall,�these� suites�of�vegetation�in�the�valley�and�nearby�uplands,� point�data�show�a�substantial� and�strong �occupation�of� largely�dominated�by�sagebrush�and�grasses�in�the� the�Gardiner�Basin�during� the� Late�Plains�Archaic,� as�well� valleys�and�pines�and�s agebrush�in�the�uplands.�� as�during�the�preceding�Middle�Plains�Archaic�and� In�confirmation�of�these�mesic�conditions�(compared� subsequent�Late�Prehistoric�periods, �trends�supported�by� to�preceding�warm�and�dry �periods),�Sanders� (2000:58� chapters�by�Park�and�Adams� in�this�volume.�These�data� 60)�and�Albanese�(1999)�documented�two�paleosols�at� are�corroborated�by�other�portions�of�the�park�as�well,� Site�24YE14.� The�two��paleosols �dated�to�ca.�1600�and� including�Yellowstone�Lake� and�the�nearby�Hayden�Valley� 2300�B.P.,�similar�to�the� dates� achieved�on�the�four� (MacDonald �and�Livers�2011;�Sanders�2001).�� prehistoric�features�from�a� paleosol�at�another�site�in� this�study,� the� Yellowstone�Bank�Cache�site� approximately�three�miles�north.�The�buried�soils�at� these�sites�attest�to�i ncreased�moisture�and�vegetation� on�stable�landforms�during �the�Late�Archaic�period�in�the� Gardiner�Basin.�� Habitation�of�the�upper�portions�of�the�Yellowstone� Plateau�by�prehistoric�peoples �was�likely�seasonal�due�to� the�heavy�winter�snows�that�b lanket�most�of�YNP�and� - 57-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� Archaic�projectile�points�at� both�the�Six�Point�site�and�the� Yellowstone�Bank�Cache�site�attests�to�their� repeated � and�conjoined�use�during�prehistory.� The�remainder� of�this�paper�provides�an�overview�of� the�lithic�artifacts�recovered�at�the� four� key�sites�in�the� study�area,� including�Yellowstone�Bank� Cache,�RJP�1,� 24YE14,�and�Little�Trail�Creek.�The�focus�is�on �the� lithic� technology �of�site�occupants�during �the�Late�Archaic� occupations; �other�studies�provide�additional�site�details� which�supplement�those�provided�below�(MacDonald�et� al.�2010;�M aas�a nd�M acDonald �2009).� � � � Figure�2.�2007�2 008�Projectile�Points� by� Period�(n=93� � total),�UM�Gardiner�Basin�Survey.� � � � As�discussed�in�more�detail� below,�the�Yellowstone�Bank� HHoorriizzoonnttaall SSccaallee Cache� site,�Little�Trail�Creek,�and� EElleevvaattiioonn 11 iinncchh==335500 fftt.. 24YE14�fit�many�of�the�Late� FFtt.. aammssll NN Archaic�site�type�criteria,� 55335500 OOlldd YYeelllloowwssttoonnee RRooaadd RRaaiillrrooaadd CCuutt including�use�of�basin�shaped� HHoolloocceennee aaggee tteerrrraaccee 55330000 YYeelllloowwssttoonnee BBaannkk CCaacchhee SSiittee rock�filled�hearths.�While�there�is� PPlleeiissttoocceennee no�evidence�of� Late�Archaic�use� 55225500 MMoorraaiinnee eeddggee YYeelllloowwssttoonnee RRiivveerr of�stone� circles� sites�in�the� SSiixx PPooiinntt SSiittee 55220000 ((MMoorraaiinnee ww//ggllaacciiaall Gardiner�Basin,�several�stone� oouuttwwaasshh ffeeaattuurreess)) circle�sites�were�identified�in�the� 55115500 UM�study�area� (Livers,�this� � volume)�and�evidence�of�game �hunting �features�are� Figure�3.�Schematic�Geomorphological�Cross�Section�of� present�at�the�nearby�Six�Point�Site� (24YE176),� Yellowstone�River�at�the�Yellowstone�Bank� Cache�Site.� immediately�west�of�the�Yellowstone�Bank�Cache� site� discussed�herein�(Figure�3).� UM�collected�27�projectile� � point�and�biface�fragments�at� Six�Point,�which�is�located� Yellowstone�Bank�Cache�Site� atop�a� glacial�moraine��at� an�elevation�of�approximately� UM�survey� results�of�the�Yellowstone�Bank�Cache� Site� 5240�ft.�amsl,�overlooking�the�Yellowstone�Bank� Cache� indicate�long �term�use�of�the� site�during �the� middle�to� site�below.�The�moraine�contains�large�glacial�outwash� late�Holocene,�with�most�active�use�during�the�Late� chutes�lined�by�glacial�boulders�associated�with�Late� Archaic�and�Late�Prehistoric�periods.�Avocational� Pleistocene�glacial�melting�in�the� valley�(Pierce�1973).��� archaeologist�Tom�Jerde�identified�the�Yellowstone�Bank� Native�Americans�apparently�drove�game�from�west� Cache� Site�during�the�summer�of�1986�when�he� observed� to�east�up�the�Late�Pleistocene�glacial�outwash� several�concentrations�of�f ire�cracked�rock�(FCR)�and� channels—using �them�essentially�as�natural�drive�lines— lithic�artifacts�on�the�ground�surface�and�eroding�from� to�the�top�of�the�moraine, �where�the�fauna�were� the�banks�of�the�Yellowstone�River.�� dispatched�by�waiting�hunters.�During�the�Late�Archaic,� Details�of�the�cache� find�location�are�provided� and�multiple�other�time�periods,� game�apparently�were� elsewhere�(MacDonald�et� al.�2010);�our�focus�herein�is� hunted�at�the�S ix�Point�Site� on�the� moraine�t op�and� on�the�lithic�technology� of�the�two�bifaces�in�the� cache� processed�below�at�the�Yellowstone�Bank�Cache� Site� that�were� available�to�study,� both�of�which�are�similar�in� (MacDonald �et� al.�2010).�The�presence� of�several�Late� -58-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� morphology�and�technology �of�production�(Table�1;� Figure�4).� Both�bifaces� were� produced�by�removal�of� large�biface�thinning�flakes�to� produce� high�width�to� thickness�ratios�of�5.44�a nd�5.04,�respectively.�Both� bifaces�weigh�between�48�49�g� and�measure��90�95�mm� long,�46�48�mm�wide,�and�8�1 0�mm�thick.�The�bifaces� were�also�both�produced�from�high�quality�cherts�and� were�heavily�retouched�for�use�as�cutting/butchering� tools.�Hafted�b iface�retouch�indexes�(HRI; �Andrefsky� 2006)�are�moderately�high�(>0.25)�for�both�bifaces,� suggesting�long�term�curation �of�the�bifaces�in�toolkits.�� FS�5 43�was�produced�from� material�that� matches� hand�samples�o f�Crescent�Hill�chert� (see�A dams,�this� volume).�The�material�is�a�white�chert/chalcedony�with � large�areas�of�pink/reddening�that�is�largely�opaque� except�for�slight�translucency �on�its�edges.�FS�544�was� produced�from� moss�agate,�available�in�south�central� Montana�and�north�c entral�W yoming�(although�more� proximate�sources�are�possible).��Use�wear�is�present� along�the�lateral�margins�of�both�bifaces,�suggesting�their� use�as�knives�or�other�cutting� tools.�In�particular,�FS�543� was�clearly�hafted�as�a� knife.� Both�bifaces� in�the�cache� are�reminiscent�of�knife�forms�shown�in�Frison� Figure�4.�Two�Bifaces �from�Yellowstone� Bank�Cache (1991:129)�dating�to� the� Late�Archaic�from� other�sites�in� � the�northwestern�Plains.�� � � � Table�1.�Measurement�Data�for�Yellowstone�Bank� Cache�Bifaces. � Based�on�the�high�concentration�of�features� and� � artifacts�in�the� northern�portion�of�24YE355,�as�well�as� � Tom�Jerde’s�cache�location,�Area�A�was� deemed�a�high� � priority�area� for� test�unit�excavations.�In�particular,�YNP� � was�interested�in�the�excavation�of�three�of�the� � prehistoric�features—Features�3,�4,�and�6—that�were� in� � danger�of�eroding�from�the�edge�of�the�Yellowstone�River� � terrace�escarpment�(Table�2).�� � � � � � � � - 59-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� Table�2.�Summary�Data�from�Features,�Yellowstone�Bank�Cache�Site.� � � � � During�the�excavation�of�those�three�features,�an� Feature�36�Lithics.�I n�the� northern�p ortion�of�Area�A� additional�prehistoric�fire�feature—Feature�36—was� of�24YE355,�UM �focused�on�the�excavation�of�features� identified�in�Test�Unit�(TU)�3.�A s�such,�four� prehistoric� eroding�from�the�Yellowstone�River�terrace.�Along�with� features�were�excavated,�each�of�which�was� identified� the�nearby �Feature�3�(discussed�below),�Feature�36�was� approximately�20�2 5�cm�below�ground�surface�(bgs) � the�furthest�north�of�these�f eatures.�Feature� 36�was� within �the� buried�paleosol.�As� discussed�below,�each�of� comprised�of�a� concentration�of�f ire�cracked� rock�and� the�features�yielded�radiocarbon �dates�and�Pelican�Lake� charcoal�in�close�proximity�to� the�location�of�Jerde’s� artifacts�which�place� occupation�of�the�buried�surface� original�cache.� A�wood�charcoal�sample�returned�a� between�500�B.C.�and�400�A.D.�(Table�2).�� conventional�radiocarbon�age�of�2280±40�B.P.�(Beta� In�association�with�the� four�features,�UM�recovered� 250835)�with�a� 2�sigma�calibration �of�Cal�B.C.�400�to �350� 2,824�prehistoric�artifacts�from�the�four�features�at�the� and�C al�B.C.�3 00�to�2 10.�This�date�p laces�u se� of�Feature� Yellowstone�Bank�Cache�site,�including�1,381�faunal�and� 36�squarely�within�the�Late� Archaic�period.� � 1,443�lithic�artifacts�(Table�2;�Table�3).�In�particular,� � excavations�yielded��a�flintknapping �feature� adjacent�to� Features�36�and�3,�which�matches�the�general� provenience�of� Jerde’s�original�cache�location.�� Nearly�75�percent�of�all�prehistoric�artifacts,�including� faunal�and�lithic�materials,�are� from�Feature� 36.� As� discussed�in�MacDonald �et�al.� (2010),�excavations� revealed�a�dense�concentration�of�faunal�and�lithic �debris� within �this�densely�packed�rock�lined�fire�feature.�The� following�discussion�is�focused �on�the� lithic�organization � of�Late�Archaic�occupations.�Table�3�summarizes�lithic� artifact�assemblages�from�the�features� by�material�type.� � � Figure�5.�Pelican�Lake�Points,� Table�3.�Lithic� Artifacts,�Yellowstone� Bank�Cache�Site. � Yellowstone� Bank�Cache�Site,� Feature�36.� - 60-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� � (n=11),�jasper�(possibly�Crescent� Hill�chert;�n=49),� and� Excavation�of�Feature�36�and�its�associated�perimeter� untyped�chert�( n=35).�� yielded�more� than�2,000�lithic�and�faunal�artifacts,� As�reflected�in�Figure�6,�there� was�a�significant� including�966�lithic �artifacts �and�1,181�bone�fragments� difference�in�the�use�of�Crescent �Hill�chert�and�obsidian� (Table�2).�Given�the�association �of�faunal�remains� by�people�sitting�adjacent� to�Feature� 36�(x2=14.437;�d f=1;� (discussed�in�MacDonald�et�al.�2010),�it�is�likely�that�the� p=0.000).�One� or�more�individuals�produced�primarily� feature�was�utilized�as�a�roasting�pit�for�a�variety�of� middle�late�stage�bifaces�from�Crescent� Hill�chert,�while� medium�and�large�game.�� finishing�bifaces� and�projectile �points�from�obsidian.� In�direct�association�with� the� Late�Archaic� Eleven�obsidian �flakes�from� the�feature�were�submitted� radiocarbon�date,�UM�recovered�two�Pelican�Lake� for�XRF�analysis�(Hughes�2008b),�with�all�coming�from� projectile�points�within�feature�matrix�(Figure�5).� The� Obsidian�Cliff�(18�miles�south).�Three�additional�flakes� points�both�were�produced�from�high�quality�varieties�of� were�submitted�for�analysis�from�the�adjacent�Late� Crescent�Hill�chert.�FS�128�was�produced�from�a�bluish� Archaic�Feature�3�(discussed�below),�each�of�which�also� black�chert�with �red�inclusions, �while�FS�140�was� yielded�an�Obsidian�Cliff�source�designation.�� produced�from� a�dark�red�chert/jasper� with�white� � inclusions.�Both �of�these�materials�match�hand�samples� 120 collected�by�the�senior�author�from�the� Crescent� Hill� 105 Crescent Hill source�in�YNP.�As�discussed�by�Adams� (this�volume),�the� Obsidian 100 Crescent�Hill�chert�source�area�is�located��15�miles�east� southeast�of�the�Yellowstone�Bank�Cache� site.����� 80 In�addition�to�the�two�Pelican�Lake�projectile�points,� 62 60 lithic�artifacts�from�Feature� 36�i nclude�a�high�density� 60 concentration�of�flaking�debris,�reflecting �an�intensive� 38 episode�of� lithic�tool�production.�Tool�production�was� 40 oriented�toward�the�manufacture�of�middle�and�late� 20 stage�bifaces�and�projectile�points.�Based�on�an�analysis� of�lithic�raw� materials,�as�well�as�their�close�provenience,� 0 Feature� 36’s�intensive�lithic�reduction�area�appears�to�be� Biface Reduction Flakes Pressure-Shaping Flakes � associated �with �the�previously� identified �cache.�The� Figure�6.�Comparison �of�Crescent�Hill�Chert �and� presence� of�these�lithics�support �the�hypothesis�that� Obsidian�Use,�Feature�36.� individuals�produced�large�bifaces�and�Pelican�Lake� � projectile�points�while�seated� adjacent� to�Feature�36.�� As�discussed�further�by�Adams�(this�volume),�these� Among�the�lithic�assemblage� recovered�in�association� differences�in�lithic�raw�m aterial�use�between�the� with�Feature�36�are�945 �flakes,�including�599� obsidian �and�Crescent�Hill�chert�reflect�variable�land �use� indeterminate� flake�fragments.�Of�the�typed�flake� cycles�of�distinct�individuals�or,�alternatively,�distinct� assemblage�(n=346),� biface�reduction�(n=151,�43.6%)�and� trips�by�the�same�individual�to�the�respective�lithic � pressure/shaping�flakes�(n=138,�39.9%)�are�the�dominant� sources.�In�either�scenario,�Crescent�Hill�chert�entered� varieties,�indicative�of�the� middle�late�stages�of�biface� the�site�in�an�earlier�stage�of�biface�reduction�than�the� and�projectile�point�manufacture.�Of�these�middle�and� obsidian,�reflecting� variability�in�m obility�and�land�use� late�stage� flakes�(n=289,�84%� of�typed�flakes),�the� during �the�Late�Archaic�s ite�occupations.�� majority�(n=167,�58%)�are�the �byproducts�of�Crescent�Hill� Feature�3�Lithics.�UM�excavated�Feature�3�in�the� chert�biface�manufacture�(Figure�6).�Other� lithic� adjacent�1x1�m � test�unit�to�the�east�of�Feature�36� materials�recovered�from�Feature�36�include�obsidian� discussed�above.�Feature�3�was� eroding�into�the�river� (discussed�below)� and�an�assortment�of�other�materials� and�the�ultimate�goal�was�to�salvage�its�contents�prior�to� in�low�quantities,�including�d acite�(n=1),�orthoquartzite� its�complete�disappearance.�While�Feature� 36�contained� - 61-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� more�than�2,000�artifacts,�Feature�3�yielded�only�181,� proportion�of�Crescent�Hill�chert�than�obsidian �(x2=7.777;� including�96�lithic�and�85�faunal�artifacts.�� df=1;�p=0.005).�� While�its�overall�artifact�densities�were�less�than�the� Feature�4/37�Lithics.�A pproximately�8.84�m�(28�ft.)� adjacent�Feature�36,�Feature� 3�contained�a� very�high� south�of�Features�3�and�36�discussed�above,�upstream� density�of�FCR�and�charcoal,� yielding�a� conventional� along�the�Yellowstone�River,�the�MYAP�team� excavated� radiocarbon�date��of�1670±50�B .P.�(2�sigma�calibration �of�� another� burn�feature—Feature�4/37—within�TUs�4�and� Cal�AD�250�4 50,�A D�450�4 60�a nd�A D�480�5 30)�( Beta� 6.�FCR �was�packed�with�charcoal�throughout�Feature� 238180).�This�date�provides�a�d irect� age� for� the�obsidian � 4/37�fill.�As�with �Feature �3�discussed�above,� Feature�4/37� Pelican�Lake� projectile�point�base�found�in�the�feature� dates�to� the� terminal�portion�of� the� Late�Archaic�period,� matrix�(Figure�7).�In�contrast�to�t he� two�Crescent�Hill� as�determined�by�a� radiocarbon�date�of�1600±70�B.P.� points�in�the� nearby�Feature�3 6,�the� Pelican�Lake�point� with�a�2�sigma�calibration �of�Cal�AD�260�2 90�and�AD�320� from�Feature�3�was�produced�from�Obsidian�Cliff� 610�(Beta�238179).�� obsidian�and�was�serrated�along�its�only�intact�lateral� The�overall�artifact�counts�from�Feature� 4/37�are�low� edge.�Its�intact� lateral�tang �is�also�more�strongly� compared�to�Features�36�to�the�north,�but�fairly�similar� pronounced�than�the�two�in�Feature�36,�although �it�is� to�Feature�3.�Statistically,�no�significant�difference�was� within�the� range�of� variation�of�Pelican�Lake� points�found� observed�in�lithic�raw�material�use�between�Features�3� elsewhere�in�the�northwestern�Plains�and�vicinity�(Foor� and�4/37,�reflecting� similar�stone�tool�manufacturing� 1982,�1998;�Frison �1991:�104).�� activities�in�the�two�feature�areas.�The�similar� radiocarbon�dates�(ca.�1,600�y ears�ago)�and�lithic� technology �suggests�contemporaneous�occupation�and� use�of�the�two�features�at�the�e nd�of�the�Late�Archaic� period.� In�total�346�lithic �and �faunal�artifacts�were�recovered� in�Feature�4/37,�including�280�lithics�and�66�faunal� remains.�The�lithic �assemblage�from�Feature� 4/37�is� comprised�largely�of�flaking�debris�from�the� manufacture� of�obsidian�(79.6%)�b ifaces.�Compared�to �Feature�36�to� the�north,�Crescent� Hill�chert� (15%)�is�infrequent�in� � Feature� 4/37,�but�c ompares� favorably�with�Feature� 3� Figure�7.�Obsidian�Pelican�Lake�Point,�Feature�3,� which�also�had� higher�ratios�of�obsidian�than�Crescent� Yellowstone�Bank�Cache�Site.� Hill�chert.�As�such,�it�a ppears�as�if�several�individuals� Of�the�96�lithic�artifacts�collected�from�Feature�3,� used�obsidian �predominantly�at�this�site,�while�one�or� flaking�debris�comprise�a�majority�(n=91),�with�only�two� more�preferred�or�simply�carried�with�him�or�her�more� bifaces,�two�unifacial�tools�(both �sidescrapers),�and�a� Crescent�Hill�chert�to� Feature�36.� The�lithic�assemblage� single�core� comprising�the�remainder.�Both �sidescrapers� from�Feature�4/37�yielded�predominantly�flaking�debris� were�produced�from�obsidian �and�show� extensive�use� (n=273,�97.5%),�with� the�only�stone�tools�being�five� along�their�lateral�margins�for� cutting/scraping.�The� bifaces�and�two �unifaces.�Typed �flakes�are�dominated�by� single�core�is�a�large�fragment�of�red�ochre,�or�hematite,� middle�late�stage�biface� reduction�(n=28/75,�3 7.3%)�and� perhaps� used�in�a�ritual�or�other�purpose�by�Native� pressure/shaping�flakes�(n=28/75,�37.3%).�Clearly�the� Americans�at�the�camp�site.�The�flake� assemblage�(n=91)� emphasis�of�flintknappers�at�this�location �was�the� from�Feature�3�is�comprised�largely�of�obsidian�flake� reduction�of� middle�and�late� stage�bifaces�and�projectile� fragments�and�indeterminate� flakes� (n=45),�with� points,�predominantly�from�obsidian.�� Crescent�Hill�chert�flakes�(n=19)�not�nearly�as�abundant� Among�the�bifaces�are�three�obsidian�projectile� as�in�Feature�36.�These�lithic�data�are� significantly� points,�including�a� midsection/tip�fragment� (FS�214)�and� different�than�the�nearby�Feature� 36�which�had�a�higher� two�untyped�notched�base� and�midsection�fragments�(FS� 213).�Each�of�the�projectile �points�was�recovered�from� - 62-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� feature�fill.�Each�of�t he�points�is�broken,�with �only �one�of� Gardiner�Basin�prior�to�venturing�northward�into�the� the�points�possibly �diagnostic.� FS�214�is�a�small�obsidian� Paradise�Valley�and�the�Great �Plains�and�Rocky� point�base�that�resembles�the�Pelican�Lake�form�in�size� Mountains�beyond.�� and�shape;�however�the�fragment�is�of�the�very�proximal� RJP�1� Site� Lithics� base�and�lacks�most�of�the�notching�and�all�of�the� blade.�� The�Yellowstone�Bank�Cache� site�provides�an� Feature�6�Lithics.�As�with�Features� 3,�36,�and�4/37� outstanding�window�in�the� organization�of�lithic� discussed�above,�Feature�6�is�another� FCR �laden�fire�pit� technology �for� Late�Archaic�hunter�g atherers�in�the� eroding��from�the�Yellowstone�River�terrace�in�the� Gardiner�Basin.�Three�other�sites�in�the� study �area—RJP� northern�portion�of�Area�A�of� 24YE355.� Feature�6�is� 1,�24YE14,�and�Little�Trail�Creek—also�provide�Late� approximately�4.4�m� (14�ft.)�south,�upstream�along�the� Archaic�data�by�which�to�better�understand�a�wider� river�from�Feature�4/37.�Charcoal�collected�from�the� range�of�lithic�r eduction�activities�in�the�study�area.� feature�received�an�AMS�conventional�radiocarbon�date� In�2008,�the� UM�team�identified�another�Late�Archaic� (Beta�2 50834)�of�2530±40� B.P.�with�a�2�sigma�calibration � lithic�reduction�site—RJP�1�(24YE190)—approximately� of�Cal�B.C.� 800�530.�This�date� is�in�close�correspondence� one�mile�southwest�of�Yellowstone�Bank�Cache�Site�(see� to�Feature�36�(ca.�2300�B .P.),�an�indication�of�long�term� Figure�1).� The�RJP1�site�is�a� moderate�to�high�density� use�during�the� entire�terminal� portion�of�the�Late�Archaic� lithic�scatter�located�on�the� northern�edge�of�a�long� period,�between�approximately�500� B.C.�and�400�AD.�� terrace�that�roughly�parallels�the�Old�Yellowstone�Trail� UM�recovered�130�artifacts�from�Feature�6�and� Road�that�runs�along�the� Yellowstone�River�between� associated�excavation�levels,�including�99�lithics�and�31� Gardiner�and�Corwin �Springs.�� The�western�edge�of�the� faunal�artifacts.�The�lithic�artifact�assemblage�from� terrace�has�a�series�of�boulders�and�associated�debris� Feature� 6�is�comprised�entirely�of�flaking�debris.�No� that�form�a�northwestern�“hook”,�within�which�is�nestled� bifaces,�projectile�points,�or�other�stone�tools�were� a�dense� lithic�scatter�identified�by�UM�as�the� RJP�1�Site� recovered�in�Feature�6.�The� flake�assemblage�is� (24YE190).��Based�on�the� distribution�of�cobbles�upon�the� comprised�of�obsidian�(60.6%)�and�Crescent� Hill�chert� terrace,� the� landform�resembles�a�glacial�outwash� (31.3%),�with�small�amounts�of�r ed�jasper�(4%)�and� channel�with�cobbles�washed�off� the�upper� terrace�down� untyped�chert�(2%).� As�with �Feature�3�to�the� north,�small� onto�the�lower� terrace�in�the�Late�Pleistocene,�similar�to� nodules�of�hematite�(red�ocher)�were�also�recovered�in� the�Six�Point�Site�(24YE0176)�identified�by�U M �in�2007� Feature� 6.�The�hematite�was�likely�used�for�ritual�or� and�discussed�briefly�above�(MacDonald� 2008).� other�decorative�purposes.�Most�of�the� recovered�flakes� Lithic �data�from�RJP�1�provide�support�for� the�model� are�indeterminate�fragments�(72.8%),�with�biface� of�Late�Archaic�technological�organization�provided� reduction�and�shaping�flakes�a ppearing�in�higher�counts� herein.�UM�collected�a�total� of�199�lithic�artifacts�from� (n=15)�than�early�stage� reduction�flakes�with �cortex� the�RJP1�Site.�Most�of�the� artifacts�observed�and� (n=4).�� collected�were� obsidian�(N=193,�97.47%),�with�chert� These�data�confirm�similar�lithic �reduction�strategies� comprising�only�a�small�amount�of�the� total�materials� for�individuals�at�Feature�6�as� those�at �the�other�three� collected�(n=4,�2.02�%).� �A�single �dacite�biface�rounds�out� features�excavated�at�the�site.�The�focus�of�Late�Archaic� the�lithic�assemblage.�Projectile�points�comprise�the� site�occupants�was�clearly�on�t he�middle�to�late�stages�of� most�significant �portion�of�the�non� flake�artifacts�(n=6,� biface�reduction,�using�materials�collected�at� the� 40%),�with�late� stage�bifaces� having�a�strong �presence� Obsidian�Cliff�and�Crescent� Hill�chert� sources.�Based�on� (n=4,�26.67%).�Of�the�debitage�collected,�all�was� lithics�from�these�four�features�at�the�Yellowstone�Bank� produced�from� obsidian, �with�the�high�number�of� Cache� site,�the� Gardiner�Basin�was�a�staging�area� from� shaping�and�biface�reduction�f lakes�(n=42,�22.95%)� which�Late�Archaic�hunter�gatherers� exploited�the� indicative�of� extensive�late�stage�biface� and�point� abundant�and�diverse�ecosystem� of�the�Yellowstone� production,�similar�to�the�pattern�described�above� at�the� Plateau�to� the� south.�After�completing�hunting�and� Yellowstone�Bank�Cache�site.� � gathering�activities�in�the�plateau�(including� lithic�raw� material�collection),�hunter�g atherers�re�grouped�in�the� - 63-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� Late�Archaic�projectile�points�were�the�only�type� obsidian�projectile�point�that� shows�heavy�retouch�on � recovered�at� the�site,�suggesting�its�Late�Archaic�age� both�lateral�edges�of�the�obverse�face.��There�is�also�light� (Figure�8).�No�features�were�identified�at�the�site.�Several� retouching� on�the�reverse�face.��A s�this�artifact�is�an� of�the� recovered�tools�at� RJP� 1�are�worthy� of�note.� FS�2�is� intact,�functional�projectile�point� it�is�not�known�why�it� a�broken�Obsidian �Cliff�obsidian �Late�Archaic �corner� was�discarded,� though�loss�due�to�human�error�is� notch�projectile�point�that�shows�no�evidence�of�use.�� possible.�� The�two�intact� edges�show�no� evidence� of�retouching,� As�shown�in�F igure�8,�FS�6�is�an�Obsidian�Cliff�obsidian� and�it�is�likely�that�the�p rojectile�point�broke�from�end� biface�fragment�tip�of� a�late� stage�or�ultra�thin�biface.�� shock�during �final�stages� of�manufacture.� �FS�4�is�a� The�artifact �resembles�the�Late� Archaic�bifaces�from�the� Figure�8.�Lithic� Artifacts,� RJP�1 �Site. broken�opaque�black�obsidian�Late �Archaic�Pelican�Lake� Yellowstone�Bank�Cache�discussed�above.��The�artifacts� projectile�point.��The�artifact� shows�light�to�moderate� exhibits�little�retouching�but�there� is�a�suggestion �of�end� retouching� on�a ll�lateral�edges�and�regular� maintenance� shock�break� resulting�in�transverse� fracture.� �FS�7�(Figure� prior�to�breakage�during�use�as�a�projectile.��XRF�analysis� 8)�is�a� white�chert�Late�Archaic �Pelican�Lake� point�broken� determined�that�the� obsidian�was�from�Big�Southern� by�end�shock�during�use.��� Butte,�Idaho,�approximately�220�miles�southwest�of�the� FS�5��is�of� special�note,�as�it�is�the�lone� Knife�River�flint� site.�FS�5�is�a�Late�Archaic�Pelican�Lake�Obsidian �Cliff� (KRF)�artifact�recovered�by�UM �in�its�work�in �the� Gardiner� - 64-� � YELLOWSTONE�ARCHAEOLOGY,� VOLUME�1� Basin�(Figure� 8).�Its�source�material—KRF—is�evidenced� Site�24YE14� � by�its�macroscropic�characteristics,�including�its�rootbeer� Sanders’�(2000)�work�at�Site� 24YE14�provides� brown�chalcedonic�character�and�its�orange/green� complimentary�data�for� those�recovered� by�UM�at�Late� ultraviolet�response�which�match�hand�samples�of�Knife� Archaic�sites�in�the�Gardiner� Basin,�Montana.�Site�24YE14� River�flint�collected�by�the�lead �author�in�the�source� area.�� is�located�approximately�three�miles�south�of� The�artifact�is�a�heavily�retouched�utilized�flake�that�had� Yellowstone�Bank�Cache�and�a� mile�south�of�RJP�1.�As� the�tip�sharpened�and�broken�during�use�as� a�drill.��There� discussed�above,�two�paleosols�of� Late�Archaic�age�were� is�a�perforator� point�on�proximal�end�near�the�original� identified�at�the�site.�At� 24YE14,�Sanders�(2000)� flake�platform,�and�the� artifact�show� signs�of�being� recovered�Pelican�Lake�projectile�points�and�diverse� hafted.� �The� retouching�on�the�artifact� results�in�a�sharp� faunal�remains—including�deer�and�elk—in�two� angle�for�the� blade,�which�may�indicate� use�in�a�sawing � components�dated�to� between�1600�and�2300�B.P.,� function.��� precisely�the�two�periods�of� Late�Archaic�occupation� FS�9�is�a�very �large�obsidian �bifacial�core�(Figure�8).�� defined�at�the�Yellowstone�Bank�Cache�as�well. � The�artifact�experienced�heavy�use �as� a�cutting�tool,� Lithic �data�from�24YE14�include�317�lithics�from�test� likely�during �butchering�of�large�game.�This�artifact�was� units�and�58�from�shovel�test�pits�(Table�4),�comprised�of� sourced�to�Obsidian �Cliff,�Wyoming,�but�blood�residue� approximately�75�percent�(n=279)�obsidian�and�22� analysis�of�the�artifact�was� negative.�The�biface�measures� percent�(n=81)�chert.� All�lithic�data� from�all�Middle�and� 164�mm�long �and�weighs�more� than�1000�grams.��It�likely� Late�Archaic�occupations�from�24YE14�( Sanders�2000:�3 4,� functioned�as�a�core�for�flake� blanks�and�as�a�heavy�d uty� 51)�are�grouped�together� for�this�analysis.�� cutting�t ool.��� � In�summary,�the�RJP�1�Site�is�a�moderate�d ensity� Table�4 .�Lithic �Artifacts�f rom �24YE14. � obsidian�lithic�scatter�of� likely�Late� Archaic� origin,�based� � on�projectile�point�association.�Obsidian� dominates�the�lithic�assemblage,� reflecting� recent�travel�toward�the� Obsidian�Cliff�source�by�site�occupants.� The�lack� of�Crescent�Hill�chert� at�the�site� indicates�no�travel�southward�along�the� Yellowstone�by�Late� Archaic� hunter� gatherers�at�RJP�1;�this�stands�in� contrast�to�those �that�camped �at�the� Lithic �data�from�24YE14�compare�well�with�lithics� nearby�Yellowstone�Bank�Cache�site�discussed�above.� from�three�of�the�four�features�at�Yellowstone�Bank� The�site�likely�is�a�single�occupation�camp�for� Late� Cache,�with�the�high�percentage�of� chert�in�Feature� 36�at� Archaic�hunter� gatherers�returning �from�a�trip�that� the�latter� site�somewhat�anomalous�compared�to�the� brought�them�near� Obsidian�Cliff,�likely�up�the�Gardiner� various�Late�Archaic�occupations�in�the�valley�bottom� River.�� discussed�herein.�Sourced�obsidian�artifacts� from�24YE14� The�large�biface�fragments�from�RJP�1�confirm�data� (n=3)�were�exclusively�from�Obsidian� Cliff�(Hughes�2000),� from�the�Yellowstone�Bank�Cache� site�discussed�above,� similar�to�the�Yellowstone�Bank�Cache�site.�While� namely�that�large�bifaces�were �produced�at�the�quarry� Sanders�did�not �identify �Crescent�Hill�chert,�specifically,� sites�and�transported�as�usable�cores,�cutting �tools,�and� at�24YE14,�we�assume�that� a�large�percentage�of� those� preforms.�The�presence� of�two�exotic�artifacts�at�RJP�1 — chert�artifacts� are�from�Crescent�Hill,�the�most�proximate� a�blade�like�tool�sourced�to�the�Knife�River�flint�quarries� chert�source� to�the�site.� � of�western�North�Dakota� and�a�biface�sourced�to�Big� Southern�Butte�near�Idaho�Falls,�Idaho—suggest�long� Little�Trail�Creek�Site� distance� trade� across�the� Yellowstone�ecosystem�during � Finally,�UM’s�excavations�at� the�Little�Trail�Creek�site� the�Late� Archaic.�� (24PA1081)�provide�complimentary�data�regarding�Late� - 65-� �
Description: