ARCHAEOLOGICAL INVESTIGATIONS IN ELLSWORTHAND RICE COUNTIES, KANSAS

CARLYLE S. SMITH

Reprinted fromAMBRICAN ANTIQUITY

Volume 14, No.4, April, 1949

ARCHAEOLOGICAL INVESTIGATIONS IN ELLSWORTHAND RICE:~tOUNTIES,KANSAS

CARLYLE s. to...::'.... I1' ..,.....~I'.

* The author was assisted in the field hisP. Smith, and by three Richardnand Okada, and Vescelius.

ARCHAEOLOGICAL investigations, wereI1 carried on by the Museum of NaturalHistory, University of Kansas, from June 19 toAugust 20, 1948, in the Kanopolis Reservoirarea' on the Smoky Hill River in EllsworthCounty and also along the Little Arkansas

,River in Rice County, Kansa's.. The River BasinSurveys of the Smithsonian Institution hadundertaken the preliminary reconnaissance andhad found more than twenty sites in the areaof the Reservoir. William O. Leuty of Ellsworthwas helpful in guiding the field parties of bothinstitutions to most of the sites; also he gath­ered surface collections which were turned overto us.

The Kanopolis Reservoir is ,f'!'+"'In+~:a.rl

highly dissected terrain which marksborder along the 98th meridian of Ih't'\I'rt,"""rto

central Kansas. The Hill River mean-ders eastward, fed by streamsand canyons. Outcrops sandstoneare common on the bluffs the vand trees are limited to the streams..

SMITH] ELLSWORTH AND RICE COUNTIES, KANSAS 293

sites in the reservoir are devoid of any evidenceof contact with Western civilization and mostof them are assignable to prehistoric culturalcomplexes previously defined as foUows: Wood­land, Hopewellian, Upper Republican, andGreat Bend. In Rice County one Great Bendsite was investigated and specimens were col­lected from the surfaces of others. Evidence ofoccupations by more than one cultural complexon some of the sites makes it impossible to pre­sent all of the data in historical sequence. Asfar as possible the approach is from late toearly.

PETROGLYPHS'

The Indian Hill site (14EW1) consists of anoutcrop of Dakota sandstone situated betweenthe mouths of Red Rock and Horse Thief can­yons on the north side of the Smoky Hill River.The vertical face of the cliff is covered to aheight of approximately 8 feet with petroglyph~~

which were carved in the soft rock. Threegroups of the drawings have been illustrated byMallery (1893, pp. 80-1, Fig. 44) and were re­corded prior to 1869. The only changes sincethen have been the additions of recent initials,names, and dates. The petroglyphs ~nclude

representations of owls, cervids, bison, bisontracks, beaver, conventionalized human figures,thunder birds, "turkey" tracks, serpentinemotifs, and ladderlike designs. In two places,not shown by Mallery, are equestrian figures.One of them shows a man, holding a long lance,astride a horse. At a point where the talusmeets the cliff a diamond-shaped knife, severalend scrapers, and a few chips of chert werefound. Probably such objects were used incarving the petroglyphs into the face of thecliff. Similar artifacts occur in Great Bend andUpper Republican sites in the Plains and offerlittle in the way of evidence which can be usedin determining the authorship of the drawings.

The presence of petroglyphs depicting eques­trians suggests a late date for many of thedrawings. The Cheyenne and, possibly, theArapaho may have contributed their share be­cause local traditions refer to the presence ofthese groups as late as the 1870's. Grinnell(1915, p:. 249) states that the Cheyenne wereattacking the stage line along the Smoky HillRiver just to the west of the area in 1867.

GREAT BEND SITES

The Great Bend or Paint Creek culture(Wedel, 1942) is represented in the Kanopolis

Reservoir by two stratified levels in the Thomp­son Creek site (14EW6). The site consists offour habitation ievels exposed in a road cutnear the confluence of Thompson Creek and theSmoky Hill River. Alternating sterile zones andlayers of dark soil containing hearths filled withwhite ashes extend along the' road for a dis­tance of approximately 75 feet. A profile 60 feetlong was trimmed along the bank and the layers~ere designated A, B, C, D from top to bottom

HUMUS

A

------.18-----__--.J C

-~-------JD

FIG. 71. Profile showing four cultural horizons at theThompson Creek site. Vertical scale exaggerated.

(Fig. 71). All four layers dip toward the streamand are more deeply buried where they reach it.Samples of the contents of each layer were ob­tained by cutting back the face of the profile.Later, a bulldozer was used to remove some ofthe overburden. The heavy rains which occur­red during the field season caused the reservoirto rise and cover the site before the investiga­tions could be completed.

Level A, overlain by humus and massive sed­iment, consisted of a dark line containing oc­casional flecks of charcoal and traces of whiteashes. At the northern end of the profile thehorizon was 1 foot 6 inches below the surfaceand dipped to a depth of 3 feet 9 inches at thesouthern end. The layer yielded two sherds ofGreat Bend pottery identifiable as GeneseoPlain (Wedel, 1948c) , one fragment of anelongate mano, one small flake retouched onall edges, and numerous animal bones.

Level B, overlain by a deposit of compactsilt, was made up of a layer of stained soil,charcoal, and white ashes averaging 2 inches inthickness. Several basin-shaped hearths com­posed of white ashes underlain by burned earthwere noted. The layer dipped from a contactwith Level A to a depth of 4 feet 6 inches atthe south end of the profile where a thin layer ofsilt interv~ned, dividing the layer into twoparts. Level B was the richest in artifacts. The

294 AilIJJJ?ICAiV AiVTIQUITY [4, 1949

Great Bend culture is represented by 100 sherdsof Geneseo Plain (Fig. 72, A, n), 17 sherds ofGeneseo Simple Stamped, one triangular pro­jectile poi.nt with a straight base, three crudeside scrapers, two sandstone abraders, one frag­ment of polished bone, and numerous animal

bones.Level C, overlain by a deposit of massive

sediment, was marked by a thin zone of stainedsoil containing a few basin-shaped hearths filledwith white ashes and underlain by burned earth.The layer joined Level B at the northern end ofthe profile and dipped to a depth of 5 feet 9inches at the southern end. 1"'wo small sherdswith little or no outer surface remaining werefound. Comparative study suggests that thesherds belong to a ceramic tradition other thanGreat Bend. The temper and texture are coarse,suggesting Upper !{.epublican or some Wood­land manifestation. In this connection it isnoteworthy that a large body sherd of Wood­land pottery was found in the talus \vhen thebank was being cleaned to reveal the profile.Broken animal bOl1es and fragn1ents of charcoalcomplete the inventory of Level (~.

Level D, overlain by a deposit of compactsilt, consisted of a thin layer of stained soil,charcoal, and a few basin-shaped hearths filledwith white ashes underlai.n by burned earth.LevelD did not contact Level C at the northernend of the profile where I.Jevel I) appeared at adepth of 3 feet 3 inches and dipped south\vardto a depth of 6 feet 9 inches. 'The only objectsfound were broken animal bones. 'rhe layerwas underlain by a deposit of massive sedi-

ment.Additional specimens were collected from the

surface of a plowed field a few hundred feet tothe northwest of the main exposure. They in­clude 12 shercls of Geneseo Plain, one unclassi­fied cord-roughened sherd, one unclassified in­cised sherd, and one broken plano-convex endscraper. In establishing the profile at the mainexposure the following artifacts were found: 15sherds of Geneseo Plain, eight sherds of C;eneseoSimple Stamped, one cord roughened Woodlandsherd, two side scrapers, one flake knife, onebone punch or flaker, and one bone awl.

To summarize the data from the ThompsonCreek site: Four superimposed occupation levelswere found. No artifacts were found in theearliest horizon, Level D. The meager yieldfrom Level C suggests an occupation by apottery-using culture other than Great Bend.Level B represents the heaviest occupation and

is assigned to the C;reat Bend culture..Level Ais assigned to the salnc culture on the basis oftwo shercls. 'fhe 1)05ition of the site at the junc­tion of t\VO streanlS suggests that the sterilezones of sedilnent may have been deposited intin1es of ilood. vVedel (1947a,F'ig. 51) showsa range in tin1e from approxirnately 1475 to1675 for the C;reat 13end Aspect. ]"'he sedimentsoverlying Levels j\ andB suggest that theSmoky FIill I~iver and 'Tl10nlpson (~reek \vereflooding their valleys and depositing sedimentsat this time. Since the cultural identity of thepeople \vho occupied the site at Levels C: andI) cannot be determiner!, any atten1r)t to cor­relate the various periods of flooding \vith cli­matic cycles in the Pla.ins \vould be prenlature.()n the basis of the data recovered, the distri­bution of the (;reat Bend l\spect is extended\vestvvard froln Pai nt Creek a.nd north\vardfroD1the Ijttle I\rkansasRiver to the Sn1.oky1-1ill .R.iver, but the yield of artifacts is toorneager to suggest any regional or telnI)oralvariation f[(Jln the culture as defined elsc\vhere.'fhe site appears to nlark periodic can1ps andit is doubtful \vhet her conlplcte excavation\vould have furnished data con1parable to thosefound in the large village sites. 'l'he filling ofthe "Kanopolis l{cservoir precluded further\vork.

rrhe Majc)r 5i te (14 R ('2) yielded a consider­able amount of data on the (;reat. Bene} culture.]'1he site consists of a series of slnall 1110unds ofrefuse scattered along a l()\v ridge bet\veen theIjttle i\rkansas ]{iver and IJorse C~reck in theto\vnship of (;al t, {{ice ('Iou 11 ty. ()ne refusemound \\':1S tested by digging a trench 4 feet\vide and 28 feet long through it. j\nothernlound \\'as excavated in g-inch levels \vithin5-£00t squares. In hoth Inou nels t.he refusc \vasfrom 16 to 36 inches thick and \vas cornposedof stained soil n1ixcd \\'ith fire-cracked stones,broken anin1al bones, charcoal, potsherels, andimplelnents of flint, of sandstone, and of bone.SOlne portions of the second rnound had. beendisturbed by the intrusion of large bell-shapedcache pits filled \vith sin1ilar detritus.

1'1he artifacts closely resenlhle those found atother (~reat Bend sites such as the discoverysite on Paint (~reek (Udden, l<)()(); \Vedel, 1935c) and those found on C:()\V C~reek ancl theLittle ArkansasH.ivcr by \Vedel (1942). 1"'hepottery falls readily into t.he types establishedat the Fifth I)lains l\rchaeological C:onference("'Tedel, 1948c). C;eneseo Plain (Fig. 72, C, D, E)

predominates over all of the other types.

SMITH] ELLSWORTH AND RICE COUNTIES, KANSAS 295

Geneseo Simple Stamped (Fig. 72, G) is secondin prevalence. Geneseo Red Filmed and CowleyPlain are also present. I have added anothertype, Little River Cord Roughened (Fig. 72, F),which includes the cord-roughened sherdsmentioned by Wedel (1942, p. 5). The potteryfrom the site is characterized by fine grit orsand tempering and occurs in the form of jarswhich are usually less than one foot in height.A few sherds are shell tempered. The rims maybe straight or flaring. The shoulders are roundedand the bottoms often flattened. Handles arepresent in three principal forms: strap, per­forated lug, and loop (Fig. 72, c, D, E). Most ofthe vessels lack decoration on the rim but somebear incised lines or notches on the lip. Plainvessels outnumber those with roughenedsurfaces. Of secondary importance are thosebearing the impressionsl of a grooved and/orthong-wrapped paddle. Some vessels are coatedwith a fugitive red paint. A smaller pr~portionhave cord-roughened surfaces suggestive of theceramic tradition of the earlier Upper Republi­can culture.

A study of the stratigraphic distribution ofthe types in the undisturbed portions of Mound2 indicates that Geneseo Plain increased inpopularity from the bottom to the top of thedeposit. It predominates at all levels, however.Cowley Plain, the shell tempered pottery, israre but tends to occur near the top. All of theother types decrease toward the top of themound. When the statistics are available fromother Great Bend sites it may be possible toarrange them in a seriation which will suggesttheir relative chronological positions.

No artifacts of European or Pueblo originwere found at the Major and Thompson Creeksites but Wedel (1942) reports such items aschain mail and painted pottery from other sitesin the region. Marjorie F. 1"'ichy (letter, Sep­tember 10,1948) of the Museum of New Mexicohas identified a number of sherds found in theLittle River area. A surface collection from theHayes site (14I<'C3), one mile west of the Majorsite, contains a Pueblo shercl associated withGreat Bend pottery. It is of Rio Grande glazedware, Late Glaze E or F, dating from theseventeenth century. At Spriggs Rocks (14RC1)approximately 3 miles south of the Major site,a collection of Pueblo sherds was gathered fromthe surface of a sandy slope devoid of otherceramic remains. The sherds seem to haveformed part of one oUa identifiable as LateGlaze C to Early D, characteristic of the

Galisteo' Basin and dating from the pre­Spanish portion of Pueblo IV. These" datescompare favorably with those suggested byWedel (1947a) for the Great Bend culture,ca. 1475-1675. The absence of objects ofEuropean and late Pueblo origin at the Majorand Thompson Creek sites suggests that bothof them date from the period prior to theCoronado expedition of 1541.

In connection with Southwestern potsherds itis worthy of note that the painted sherds foundby Williston and Martin (Martin, 1909) in1898 at the Scott County Pueblo have beenidentified by Tichy and indicate a late seven­teenth century date for the ruin. \Vedel (1947a)reports Great Bend sherds along with a sherdof Tewa Polychrome and an incised pipe stemwhich are assigned to the latter part of theseventeenth and the early part of the eighteenthcenturies. The Williston and Martin sampleshows the presence of Tewa and PojoaquePolychrome, "Kapo" black ware, late red wares,and Rio Grande culinary wares. Taken as agroup, these varieties indicate a late seventeenthto early eighteenth century date for the site.Along with the Southwestern pottery and thefew sherds of Great Bend pottery were numer­ous specimens of Dismal River pottery (Hilland Metcalf, 1942). Some of the sherds pre­viously classified along with sand-temperedsherds as Dismal River are identified by Tichyas "late Rio Grande micaceous culinary ware."I suggest, however, that the mica-temperedware of New Mexico may represent an instanceof diffusion from the Plains into the Southwest.The need for a detailed study of Rio Grandeutility ware is apparent. The data indicate,then, a partial overlap in time in Kansas forGreat Bend, Dismal River, and refugee Pueblogroups from the Upper Rio Grande.

UPPER REPUBLICAN SITES

The Upper Republican culture is representedin surface collections from a number of smallsites throughout the Kanopolis Reservoir.Probably the remains of earth lodges arepresent at sites along Clear Creek on the northside of the Smoky Hill River but the limitedamount of labor and time available caused ourattentions to be directed toward the collectingof data pertinent to the definition of cultureswhich are less well known in the Great Plains.Upper Republican sherds were found at theElm Creek site (14EW12) described below.The only rim sherd pertaining to the Upper

296 AMERICAN ANTIQUITY 14, 1949

c D

F

INCHES

H

FIG. 72. Pottery from Ellsworth and Rice counties, Kansas. (See facing page for identifications.)

SMITH] ELLSWORTH AND RICE COUNTIES, KANSAS 297

Republican culture found at the Elm Creek siteis of the braced variety and is cord roughened(Fig. 72, H). The sherds are medium in texture,flaky in structure, and appear to be temperedwith particles of shale. Two triangular projectilepoints with straight bases suggest assignmentto the same culture. The greater prevalence ofcord-roughened sherds (Little River CordRoughened) in the deeper part of a refuse moundat the Major site suggests a carry-over of anUpper Republican ceramic trait into the GreatBend complex. Otherwise the Great BendAspect is without known antecedents in Kansasas far as our present knowledge extends.

A HOPEWELLIAN SITE

The Hopewellian culture is represented atthe Ward site (14EW17), which is situated ona branch of Ash Creek, a tributary of the SmokyHill River, some ten miles south of Ellsworth.Our attention was drawn to the site throughthe finding. of rocker stamped pottery by Wil­liam Leuty and by the field party of the RiverBasin Surveys. A limited area in a plowedfield was marked by a concentration of burnedfragments of Dakota sandstone and brokenbison bones. When the plowed soil was st.rippedfrom the area a circular feature approximately20 feet in diameter composed of jet-black soilwas revealed.

The deposit of dark soil contained numerousfragments of sandstone which were rarelylarger than the size of a man's fist. It waslenti cular in cross section, being 8 inches thickat the center and tapering in all directions. Thesoil below it was stained as though it had beendisturbed and the same coloration extended atleast 5 feet to the southeast of the margin of theoverlying black deposit.. Within the blackarea four pits occurred. Two were circular;one was rectanguloid, and the other irregularin outline. All of them had rounded, basin­shaped bottoms and contained highly con­centrated masses of burned sandstone, brokenanimal bones, traces of white ashes, charcoal,and black soil. They ranged in depth from 17 to24 inches. One centrally situated pit had alarge amount of white ashes at the bottom and

was intrusive into another pit to the southwest.N a traces of post molds or other features could.be found. The presence of a pit use-d as a fire­place near the center of a circular area showingintensive human occupation suggests the formerexistence of a dwelling but if such ever existedit must have been of a perishable nature.

The excavations yielded two sherds of brown­ish black pottery heavily tempered with grit.One is a rim sherd which slopes inward abovea slightly constricted neck-area (Fig. 72, I).The lip is flat and there are three exterior bossesformed by deep punctates entering from theinterior surface. The surface between the bossesand the lip is covered with broad, deep incisedlines arranged in a crosshatched pattern.Below the bosses on the area of the neck thereis a rectanguloid plat filled with verticallyoriented lines of dentate stamping, bordered ontwo sides by two broad incised lines which areU-shaped in profile. The other sherd is from thebody of a vessel and bears an area of dentatestamping bordered by a broad incised line.There can be little doubt as to the Hopewellianaffiliations of the two sherds. The rim sherdcannot be duplicated in all its features in theKansas City Hopewell culture but the char­acteristics are distributed among the severalvarieties of sherds found by Wedel (1943).

Other artifacts from the excavations includea side notched projectile point and one whichis lanceolate with a straight base; an elongateknife with a convex base, nearly parallel sides,and a rounded tip; two flake knives; threeplano-convex end scrapers; two chipped disks orblanks; one quartzite chopper; several sandstoneabraders; two polished splinter awls. The sur­face of the site yielded ten grit-tempered sherds.Eight are plain; one has large deep exteriorpunctates; another exhibits a zone of smoothrocker stamping. Projectile points include onewith an expanded base and another which iscorner notched. Two plano-convex end scrapers,a fragmentary chipped celt, and a sandstoneabrader complete the list.

The Ward site offers the first definite evidenceof a Hopewellian habitation site west of the98th meridian. Wedel (1940a) has reported the

------------------------------------_._--_..._-

EXPLANATION OF FIG. 72. A, B, Geneseo Plain, Thompson Creek site; c, D, E, Geneseo Plain, handles, Major site; F,

Little River Cord Roughened, Major site; G, Geneseo Simple Stamped, Major site; H, cord-roughened braced rim, Elm Creeksite; I, incised, stamped, and bossed rim, Ward site; J, cord-roughened, belled rim, Elm Creek site; K, cord-roughened bodysherd, Red Rock Canyon site. A-G, Great Bend Aspect; H, Upper Republican Aspect; I, Hopewellian Phase; }, K, Wood­land Pattern.

298 A~IERICAN ANTIQUITY [4, 1949

finding of one Hopewellian sherd in associationwith Woodland pottery on Salt Creek in LaneCounty, west of the lOOth meridian, whereUpper Republican houses were superimposedupon a site of Woodland affiliation. Too fewspecimens were found at the Ward site topermit detailed comparison with material fromother Hopewellian sites found in the vicinityof Kansas City'. More information is needed onthe nature of the Hopewellian invasion of thePlains, for if Griffin (1946, p. 67) is right, it mayhave contributed materially to the develop­ment of Upper Republican pottery, giving itthe incised and collar~d rim.

WOODLAND SITES

The Elm Creek site (14EW12) is situated onthe east bank of Elm Creek near its junctionwith the Smoky. Hill River. The principal oc­cupation is referable to a culture affiliated withthe Woodland Pattern. A secondary occupationby the Upper Republican culture has beendiscussed above. Extensive tests made in thesite failed to reveal any evidence of prolongedoccupation by either culture, however. Mostof the specimens were found on the surface orwithin the plowed soil. A thin deposit of refusecontaining Woodland pottery was found at thesouthwestern corner of the site, whereas UpperRepublican pottery was common toward thenortheast on the surface and in the plowed soil.Two varieties of Woodland pottery are distin­guished on the basis of tempering material.One variety contains angular particles of calciteand the other contains rounded sand grains orcrushed particles of quartz. The paste of bothvarieties is coarse and granular. All of the sherdsare cord roughened but some bear verticallyoriented impressions while others bear diagonalimprints which often cross each other. Althougha few sherds from neck-areas suggest flaringrims, the only rim sherds found are straight inprofile with flattened lips. One sherd (Fig. 72, J)is slightly belled, suggesting the cambered rimof the Hopewellian culture. Nearly all of theprojectile points are equilateral in proportions,having barbed blades and stemmed or cornernotched bases. Other artifacts are less diag­nostic and could belong either to a Woodlandor to an Upper Republican occupation; theyare plano-convex endscrapers, a chipped celt,and numerous side scrapers of shapes seeminglydetermined by the outline of the originalflake.

The Red Rock Canyon site (14EW13) issituated on a bluff between the canyon of the

same name and Elm Creek on the north side ofthe Smoky Hill River. The specimens seem toindicate an affiliation with the Woodlandpattern. Test excavations were made in threeareas where burrowing animals had unearthedflint chips and fragments of fresh water shells.Tests made with a soil auger at other placesin the site failed to reveal any other con­centrations of cultural material. At the south­ern edge of the bluff two disturbed cairns builtof sandstone slabs were excavated. They hadrectanguloid cists at the center. Probably theyonce contained burials which had been removedwhen loose stones had been gathered from thearea for the purpose of building a stone-walledcorral farther to the east. One flint chip and afragment of a plano-convex end scraper werefound in removing the soil from around thecairns.

The remainder of the site yielded potteryclosely resembling that found at the Elm Creeksite. The sherds are cord roughened with theimpressions more often arranged vertically(Fig. 72, K) than in a criss-cross pattern. Most ofthe sherds are grit tempered but a few containfragments of calcite and still others fragmentsof shale. The shale-tempered sherds may beUpper Republican rather than Woodland inculture. One sherd has a hole drilled through it.The only rim sherd found is straight in profileand has a rounded lip. Some of the sherds haveirregular interior surfaces covered with brushmarks and impressions of finger nails. No sherdsdefinitely identifiable as Upper I<.epublicanwere found.

The projectile points are typically small,averaging less than three-quarters of an inchin length, with sharp barbs and expanded stems.One has a serrated blade. They suggest theprojectile points found in Hopewellian moundsalong the' Republican River in Kansas (Schultzand Spaulding, 1948, PI. XXIX) in associationwith Hopewellian and Woodland pottery.Other artifacts include a T -shaped drill witha concave base, plano-convex end scrapers,amorphous side scrapers, a flat sandstonegrinding slab, a perforated shell hoe, and a smallshell pendant with two perforations.

The Woodland culture appears at more sitesin the Kanopolis area than do any of the othercultures. Even so, less is known of the Woodlandculture because of the meager yield of data fromanyone site. The projectile points and a fewof the ceramic traits suggest. a relationship toHopewellian, but it is not clear whether thisindicates contemporaneity of the two cultures

SMITH] ELLSWORTH AND RICE COUNTIES, KANSAS 299

or whether the Woodland culture lasted longenough to receive diffused traits prior to theactual entry of Hopewellian groups into thePlains. The sites suggest the presence of smallhunting and gathering gr"oups who did not lingerlong at one place.

BURIAL CAIRNS

The bluffs along the Smoky flill River werefavorite places for the erection of cairns over thebodies of the dead. Unfortunately most of thecairns have been destroyed by persons whogathered stones for walls as well as by collectorsof artifacts. Two cairns were investigated at'the Red Rock Canyon site and our attentionwas directed to a group of seven cairns on theHudson site (14EW24), which is situated on thetop of a grassy butte south of the Smoky HillRiver, overlooking a branch of Thompson Creek.The cairns are barely discernible in the grass,a condition which undoubtedly led to theirsurvival. Four cairns were excavated. In oneof them a flexed skeleton, crushed by theweight of the overlying rocks, was found. Thecairn containing the burial was a roundedmound of Dakota sandstone measuring approxi­mately 12 feet in diameter and 15 inches inheight at the center. The base rested upon sterilesubsoil but humus lapped up on the sides of thecairn to a thickness of 9 inches. The rocks form­ing the top of the mound were exposed to viewon the surface over an area measuring 8 feet indiameter. The perimeter of the cairn wasmarked by flat slabs which may have beenplaced there to retain the small, irregular frag­ments of sandstone of which the cairn wascomposed. At the center of the cairn was arectanguloid cist measuring approximately 66inches from east to west and 30 inches fromnorth to south. In a few places vertically placedslabs marked the edges of the cist but in mostplaces the mass of sandstone fragments formingthe cairn ended irregularly, giving way to thefill in the cist which was composed of dark soilmixed with rock fragments. The overlyingstones had sunk into the cist and had com­pressed and broken the bones of a human skele­ton of indeterminate sex. The skeleton restedupon its right side with the head to the east,facing north. The legs were partly flexed.The right hand rested upon the pelvic regionwhile the left hand was at, or near, the chin.Most of the skull had disintegrated and theother bones were fragmentary. There were noartifacts associated with the burial.

The three other cairns at the site were built

of larger stones and did not contain recognizablecists. Under one of them was an area of stainedsoil with a filled animal burrow leading into it.If any burial had been present below the cairn,burrowing mammals probably have removedall traces of it. Perhaps they were commemora­tive monuments or protective coverings forcaches of perishable material. No artifacts orother cultural materials were found at the site.The presence of these culturally unidentifiablecairns in the area suggests the westward exten­sion of mound building, but the lack of artifactsprecludes identification of the cairns with anyknown cultural complex.

SUMMARY AND CONCLUSIONS

A study of the archaeological remains fromEllsworth and Rice counties, Kansas, reveals aprogression of cultural complexes which oc­cupied the area on different time levels. General­ly speaking, the cultures contrast sharply witheach other and very little evidence of transitionfrom one to the other is present. Thus far thereare no data on preceramic cultures in the areastudied. Although the burial cairns which wereexcavated are devoid of artifacts they probablywill be linked with a pottery-using culture whenmore work has been done.

Sites attributable to the Woodland Patternare common in the area but produce few speci­mens and therefore do not lend themselves tofurther classification at this time. The distribu­tion of the Hopewellian Phase is now extendedwest of the 98th meridian but the sites arerare, suggesting a brief period of occupancy.Sites of the Upper Republican Aspect are fairlycommon but no earth-lodge remains were found.The Great Bend Aspect is w·ell represented inRice County. The discovery of a Great Bendcamp site on Thompson Creek extends theknown distribution of the complex farther northand west than previously known. The strati­graphy at the Major site implies that it will bepossible to arrange many of the Great Bend sitesin chronological order, thus giving us timeperspective within the culture. The identifica­tion of the sherds from the Rio Grande pueblosadds additional support to the dates arrived atby Wedel (1947a). The data gathered from theIndian Hill petroglyph site will be of greatercomparative value when other petroglyphsites in the Plains are reported on.

Using Wedel's dates for the ceramic cultures,we find that the area covered by Ells\vorth andRice counties was occupied from approximately1000 A.D. on. The earliest cultures are those

300 AMERICAN ANTIQUITY [4,1949

affiliated with· the WoodlandPattern.~Towardthe ·close of the twelfth century, Hopewelliangroups had ventured into the area. rrhe UpperRepublican Aspect may have been present asearly as the beginning of the thirteenth century.The cord-roughened sherds in the Great Bendculture suggest contact with, or a genesis in, theUpper Republican culture. No data establish­ing contemporaneity are present. The GreatBend Aspect must~ have existed from the late

fifteenth until the latter part of the seventeenthcentury. Data on the occupation of the areaby bison-hunting nomads are limited archaeo- Ilogically to petroglyphs showing men ridinghorses. The last aboriginal occupation was thatof the Cheyenne and Arapaho, which extendedinto the last half of the nineteenth century..

Muscurn of Na.tural lIistory(Jnivcrsity ()fKan!'iasLawrence, Kan~n.s