Early Sites on the Peruvian Coast

Author(s): Frederic Engel

Source: Southwestern Journal of Anthropology, Vol. 13, No. 1 (Spring, 1957), pp. 54-68
Published by: University of New Mexico
Stable URL: http://www.jstor.org/stable/3629157
Accessed: 22/03/2010 23:35

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=unm.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

University of New Mexico is collaborating with JSTOR to digitize, preserve and extend access to Southwestern
Journal of Anthropology.

http://www.jstor.org
EARLYSITES ON THE PERUVIANCOAST
FREDERICENGEL

THIS PAPER is a first report on a surveyof the Peruviancoast from

Chimbote,420 kilometersnorth of Lima, to the mouth of the Rio Ica, 450
kilometerssouth of Lima, conductedover the last two years. This surveyhas re-
sulted in the discoveryof some twenty archaeologicalsites which were probably
occupiedin pre-ceramictimes. In May, 1955, the Peruviangovernmentwas kind
enoughto grant the authora permitto conductexcavationson the Peruviancoast,
with the purposeof contributing,if possible,to the knowledgeof early sites. Test
excavationshave accordinglybeen made in severalof these sites, and will be de-
scribedbrieflyin this paper.
Early sites on the Peruviancoast have been reportedby a numberof writers,
but few systematicexcavationshave been made to date. These include work by
Bird at varioussites on the north Chilean coast1 and at Huaca Prieta and Cerro
Prieto in the Chicamaand Viru valleys,2excavationsby Willey and Corbett in
the Asperosite, Supe,3and by Strong at San Nicolas de Marcona.4Certainof the
sites reportedin this paper have been mentionedin a previouspublication.5
Our field studies are still in progress,and the survey has covered less than
one third of the 3000 kilometersof the Peruvian coast. The present paper is
offeredin order to put on recorda numberof sites which are of interestbecause
of their early age, and to call the attentionof other studentsto some of the work
which remainsto be done on the coast of Peru. During the courseof the survey,
a large numberof shell mounds were examined,and many of these test-pitted.
The great majorityindicatea late occupationfrom top to bottom. Others, how-
ever, producedno sherds,and presumablydate to the pre-ceramicperiod. These
are shown on Figure 1 and brieflydescribedhere:
1. Santa. This site was pointedout to the authorby James Ford. It is located
1 Junius B. Bird, Excavationsin Northern Chile (Anthropological Papers, American Mu-
seum of Natural History, vol. 38, pt. 4, 1943).
2 Junius B. Bird, "PreceramicCultures in Chicamaand Viri" (in A Reappraisalof Peruvian
Archaeology, Wendell C. Bennett, ed., Memoirs, Society for American Archaeology, no. 4,
pp. 21-28, 1948), pp. 21-25.
3 Gordon R. Willey and John M. Corbett, Early Ancon and Early Supe Cultures (Columbia
Studies in Archeology and Ethnology, vol. 4, 1954), pp. 24-25.
4 John H. Rowe, Notes and News: South America (American Antiquity, vol. 19, pp. 106-
107, 1953).
5 Frederic Engel, Les amas de coquillages de la cote peruvienne (Ancon-Rio Ica) (Journal
de la Societe des Americanistes,n.s., vol. 44, pp. 39-47, 1955).
54

VOL. 13, 1957

 EARLY SITES ON PERUVIAN COAST 55

on the southernbank of the Santa River, one kilometer from the seashore.The
site containsa stone-facedterraceand stone walls. Hammerstonesand percussion-
flakedstone picks occuron the surface.

5- C

FIG. 1. Location of early sites on the Peruvian coast (Scale: 1 cm = 46.8 km).
56 SOUTHWESTERN JOURNAL OF ANTHROPOLOGY

2. Samanco.The site is composedof a series of shell mounds along the bay

wherethe Nepaiia Riverruns into the sea.
3. Huayanuma.Near Samanco;probablypre-ceramic.
4. Tortuga. Near Samanco;probablypre-ceramic.This site was pointed out
to the authorby Dr Donald Collier.
5 and 6. Sites with cemeteriesand the remainsof dwellings,in the bay where
the CulebrasRiver runs into the sea. Site 5, on the southernbank, is described
below.
7. Zorras. Shellmoundsclose to the dry riverbed and the bay of Las Zorras.
Nearby is a site with stone architecture,probablydating to the early ceramic
period.
8. Aspero.In the areaof Aspero,whereone pre-ceramicsite was excavatedby
Willey and Corbett,6lie variousother moundsin which we could find no pottery.
9. Monte Trigo. A shellmoundclose to the beachat the foot of Monte Trigo,
a hill south of PuertoSupe, about half way from the latter to Huacho.
10. Medio Mundo. A shellmoundlying on the beach, not far from site num-
ber 9.
11. Punta Quilca.A very interestingmoundsituatedon the south bank of the
SalinasBay, south of Huacho. The lowerlevels of the moundvery probablycon-
tain pre-potteryoccupation.The upperlevels yield undecoratedceramics,probably
pre-Chavin.Other moundscan also be seen in the Punta Lachaypeninsula.
12. Rio Seco. Where the Rio Seco canyon meets the sea, after crossing the
PanamericanHighway at Km. 85 north, there is an importantstratifiedshell-
mound.Nearby is an area of stone architecturewhich yields pottery,presumably
also early.
13. Playa Grande.This shellmoundlies on the southernslope of Cerro San
Pedro, west of the Early Lima settlementof Playa Grande.We found only one
sherd,after sifting morethan eighty cubic metersof refuse.
14. Ventanilla.Severalshellmoundswith no visiblesherds,close to Hacienda
Ventanilla,betweenPlaya Grandeand the ChillonRiver.
15. Chira-Villa.Describedbelow.
16. Zig-Zag. A site in Chorrillos,where test pits failed to indicate pottery.
6 WilleyandCorbett,loc.cit.
 EARLY SITES ON PERUVIAN COAST 57

17. Caleta de los Antiguos. Several shellmoundson the PucusanaPeninsula.

Stone architectureis visible.
18. Otuma. Describedbelow.
19. Punta de Asma. Two large shellmounds,badly wind-eroded.
20. Mouth of the Rio Ica. Describedbelow.

SITE 18. LAGUNA DE OTUMA

The ancient lagoon of Otuma lies at 14? south longitude, 76?16' west lati-
tude. It is 33 kilometerssouth of Pisco and 15 kilometerssouth of the CerroColo-
rado site on the ParacasPeninsula.The route to the site is a truck trail through
the desert,which joins the paved highwayabout a kilometersouth of the Paracas
Hotel.
The coastal range is interruptedat Otuma, and at some time in the past the
sea enteredthis break and penetratedabout four kilometersinland, in a general
northwest-southeastdirection, forming a bay some two kilometerswide. Subse-
quently, a gravel bar some four meters high accumulatedoffshore, isolating the
lagoon. This subsequentlyformed two separatesaltwaterponds, separatedby a
few hundredmetersof very slightly higherground.The outer of these still exists,
on the northwestern,seaward,side of the coastal range.It is now a drying marsh,
filling with blown-insand and organic matter, mainly decayed algae. The inner
pond, southeastof the coastalrange,is now nearlydry, but its highly saline water
still supportsmarinelife: algae, insects,fish, and a flock of flamingos.The bulk of
the old inner lagoon bed now also consists of mixed sand and decayed organic
matter,overlyingthe marineshalesthat formedthe bed of the old bay.
Tectonic movementshave changed the bottom level of the lagoon, probably
morethan once,so that the old bed of the lagoonnow slopes from 10 metersbelow
sea level to 15 meters above. The surroundinguplands, composedof shales and
sandstones of sedimentary- probably all of marine--origin, have been re-
modelledby tectonicmovementsand severewind erosion.They are entirelybarren
of flora, even of cactus or lichens.
The climate is arid; no trace of fog condensationcan be observed.Rain is
extremelyrare; night temperatureis never low; a strong southwestwind usually
blows in the afternoon, sand-blastingand cutting even the hardest stone. Foxes,
lizards,scorpions,and sea birdsstill manageto live here,but the area is completely
desert.Water has to brought over from Pisco, or from Ica, 50 kilometersto the
east. The nearest human habitationis a small communityof fishermenon the
northernend of the Bahia de la Independencia,16 kilometerssouth of Otuma,
which has no drinking water.
58 SOUTHWESTERN JOURNAL OF ANTHROPOLOGY

In prehistoric times,the areamusthave lookedquitedifferent.FromPisco

southto the northshoreof the ParacasPeninsula,thereareabundantremainsof
sedentary occupation, partlybasedon fishingandthecollecting
partlyagricultural,
of sea food. On the peninsulaproper,and farthersouthdownto the mouthof
the Ica valley,the authorhas founda seriesof majorsites,occupiedduringthe
early,middle,andlateprehistoric periods.
But the innerlagoonof Otumais the onlyareawhichcouldbe foundin this
zonewhichshowsevidenceof occupation by peoplelivingin a purelypre-ceramic
period.Theirpresenceis indicatedby somethirty-two shellmounds,in whichwe
havenot yet beenable to find potteryor loom-woven textiles,but whereashes,
carbon,bones,andimplements areabundant.
All signsof occupation areconcentrated on the north,east,andsouthshores
of whatwas the innerlagoon,perhapsbecauseit has a milderclimatethan its
outercounterpart, withbetterprotectionagainstwindandwaves.
All themoundsareestablished in thevicinityof whatwasthewaterlevelbefore
thecontactwiththeseawasbrokenandthelagoonbeganto fill.Exceptformounds
Nos. 2, 12,and 14,whichlie a fewmetershigherup the slopeof the surrounding
hills,the refuseis usuallynot morethanfivemetersabovethe ancientwaterlevel.
Someof the moundsof the northshoreareestablished on sandstonehillswhich
musthavelain immediately abovethe water,andof whichthe lowerslopeswere
inundated.
All of the moundsconsistof a mixtureof shells, sand, ashes,and organicmat-
ter. They varyin height from thirtycentimetersto two meters.As yet we have sys-
tematically analyzedthe contentof only a few mounds;but on the basisof test
pitsin nineof thesesites,andan examination it
of the surfaceof the remainder,
appearsthat thereare significantdifferencesin the contentof certainmounds,
whichmaybe usefulin establishing the relativechronology of the area.Mostof
themoundsarecomposed of wholeshellsof verylargePectenpurpuratus, mixed
withsandandash.Othermolluscanspeciesarerarein thesesites,and boneand
organicmatterarenot common.Brokenshelloccursin onlysmallquantities. Also
fairlyfrequentare moundscomposedof Pectenpurpuratus and clams (Mactra
sp.) in aboutequalquantities, witha considerably highercontentof brokenshell.
Site 12 is of thistype,butcontainsin additiona largequantityof woodandother
organicmatter,andhas a muchhigherartifactyieldthananyof the othersites.
It lies on higher ground, and this differencein compositionmay be due to less
moistconditions.
Therearealso a few moundsin whichothermolluscanspeciespredominate;
we havenot yet conductedtest excavations in these.MoundNo. 2 is uniquein
havinglittle shell. The depositconsistsalmost entirelyof ashes,carbon,and
 EARLY SITES ON PERUVIAN COAST 59

organicmatter,withanimalboneoccurringin moderatequantity.Number13 is
also unique,thoughothersof the sort probablyexist.It lies in the bed of the
ancientlagoon,nearthe shore,andis entirelycoveredby recentblownsand.The
deposit--which consistsof shells of Pectenpurpuratus,Chionesp., Tagelus
dombeyi,Tegulaatra,and Mactrasp., mixedwith decayedorganicmatterand
sand- is saturatedby the residualwaterof the lagoonbed,whichhasdestroyed
all but the mostdurablematerial.This site mayrepresentan occupationof the
lagoonbedafterthe retreatof thewaterfromthe old shoreline.
Wheremoundswithdifferentcontentsstandside by side,an analysisof the
components maywellserveto givea relativedateto the occupation of the respec-
tivesites.Pectenpurpuratus, for example,todaylivesin deep,coldwater,whereas
members of Mactragenusareshallow-living forms.Sincethewatersof thelagoon
musthavegrownprogressively shalloweras evaporationproceeded, it seemsprob-
ablethata sitewitha highcontentof Pectenshellswoulddateto an earlierperiod
thanone witha highcontentof Mactrashells- to a periodwhenthe watersof
the lagoonweredeeperin thatsection.Unfortunately,this argumentcan be ap-
pliedonly to sites lying neareachother;with the tectonicdeformationof the
lagoonbottom,thewateronoppositesidesof thelagoonmayhavebeenof different
depthsat any giventime.A secondline of reasoningderivesfromthe fact that
Pectenpurpuratus todayliveson bedsof crushedshell whichaccumulate along
opencoastline,6abut whichare not to be foundin the blownsand formingthe
presentmudbedof the Otumalagoon.It is quitepossiblethatthe moundscom-
posedalmostentirelyof Pectenshellsdateto a periodbeforethe formationof the
gravelbarand the isolationof the lagoonfromthe sea.Also of possiblesignifi-
canceis the fact that the enormous Pectenwhichcharacterize thesesitesarevery
rarealongthe Peruviancoasttoday,andthatsiteswhichshowhigherpercentages
of othermolluscan speciesalsocontainmoresmallPectenthanlarge.
The meatdiet of the inhabitants of Otumaconsistedof whale,dolphin,and
fish.Sealandsea lionbonesarenot common;birdboneis rareexceptwhereused
for artifacts;andlandmammalbonesarealmostnonexistent. Algaeoccurin large
quantitiesin certainsites,but the remains of otheredibleplantsarescarce.A few
gourdfragments from site 12 are the onlysuggestion gardeningfoundin any
of
of the ninesitesin whichtest excavations wereconducted.Cottonfish-netfrag-
ments occur commonlyat site 12, and a fragment of twined cotton cloth was
foundin the test trenchat thatsite.
A numberof the moundsshowshallowdepressionson the surface,nowfilled
withblownsand.Theseare usuallymoreor less square,abouttwo meterson a
of shell species,the authoris indebted
as well as the identification
6a For this information,
to Drs W. H. and M. Koepke of the Museo de Historia Natural, Lima.
60 SOUTHWESTERN JOURNAL OF ANTHROPOLOGY

side and a meter deep. It is probablethat they representthe remainsof semi-

subterraneandwellings.
As yet, we have not conductedexcavationson a scale that would permit the
recognitionof cultural differencesbetweenmounds.Except for No. 12, the arti-
fact yield is very low. Most of the artifactscome from site 12, and the few from
othersites do not differsignificantlyfrom the assemblageat that mound.
Stone artifacts are not particularlycommonin the Otuma mounds.There are
a few heavy percussion-flaked choppersand core scrapersof basalt and quartzite;
a sandstonegrindingstone from site 27 with a small, shallowbasin; and a couple
of fragmentswhichmay be from small abradingstones. Site 12 produceda single
obsidianprojectilepoint, of the broad, flat, triangulartype which occurs in the
Paracasculture. A numberof small ovoid pebbles are of interest, becausethree
occurredtogetherin the trenchat site 12, suggestingtheir use as bolas stones.
A puzzling artifact type consists of the distal or proximalend of the long
bone of a bird, cut off below the head. These are generally carefully cut, and
sometimessmoothedafter cutting; yet the tendonsare often still attachedto the
head of the bone. They occur in considerablequantity: in 7.7 cubic meters of
deposit excavatedat site 12, there were twenty-oneof these specimens- mostly
made from humeri-and a similar one made from the rib of a sea lion.
They do not seem to be by-productsfrom the manufactureof bone implements,
becausethere are only two other bone artifacts in the collection: a cut bird-bone
tube (whichmay be a by-productof the manufactureof the cut-off bone heads),
and a flat polishedimplementof bird bone, squaredoff at both ends. No explana-
tion as to the functionof these artifactsis apparent.
Our collectionfromsite 12 also containsa coupleof smallknivesand a serrated
scraperor saw of shell of Mytilus chorus;the tips of two pointedwoodenimple-
ments,probablyawls; and a short woodenplug, cut off at both ends. Cut lengths
of split cane are frequent.
Fish-net,whichoccursin somequantity,is mostlyof cotton.Cottonstringis not
uncommon,and site 12 yielded a single fragment of twined cotton cloth. The
only other artifacts in the collection are fragmentsof rope of the rush known
locallyas inea,and knottedbundlesof fiberof inea.
On the basis of the work conductedto date, we can assume that the Otuma
shell moundswereoccupiedby fishermenliving in the pre-ceramicperiod.We hope
that future studies will tell us if these fishermensettled in Otuma in very early
days, and continuedliving by the lagoon until the introductionof gardening,or
if they first arrivedin late pre-ceramictimes. As yet, only site 12 has yielded evi-
dence of gardening.Analysis of carbonsamplesshould give us an indicationof
the absolutedating of some of the mounds,and further excavationwill not only
 EARLY SITES ON PERUVIAN COAST 61

allow a fuller definitionof this early culture, but should tell us somethingabout
the relationshipbetweenthe variousmounds.

SITE 20. MOUTH OF THE RIO ICA

The shell mounds on the south of the former estuary of the Ica River have
been visited by various archaeologists,including Uhle and Kroeber. To our
knowledge, they have not yet been excavated.The author has already reported
his explorationof these mounds7 the surface of which shows ample remains
of very late occupation.These mounds consist primarilyof sand, and stand re-
spectively80 and 150 metershigh. While theremay be an early occupationrepre-
sented at their base, excavationwould be extremelydifficult.To examinea few
cubicmetersof deep depositwould requirean enormoustrench,with the construc-
tion of retainingwalls to prevent sand slides. In the light of this difficulty,the
authorsearchedthe area for smallermounds,which would be easierto approach,
and had the good fortune to locate a probablypre-ceramicsite on the north bank
of the ancientriverbed.
The Ica Riverno longer runsinto the sea, as its wateris absorbedby irrigation
in the upper valley. This extensiveirrigationis relativelyrecent, and some water
must have flowed down at the time of the occupationof the mounds. The area
today is completelydesert,with small dunes coveringthe landscapeand no vege-
tation to be seen.
The site is a shellmound,coveringabout one-fourthof an hectare.It has an
elongatedform, its axis oriented60? west of north. The mound forms two high
areas,separatedby a depressionnow coveredwith blown sand. The southeasterly
part shows signs of a later occupation,with cotton cloth, cane houses, and a few
sherds.The northwesterlysection, which containsthe remainsof the early occu-
pation, forms a mound some ten metershigh, and shows no signs of later reoccu-
pation. The surface of this northwesternmound is covered with broken and
wind-blastedstones, among which are to be found hundredsof small fist-sized
choppers,core scrapers,and other percussion-flakedimplements,made mostly of
quartziteand graniticmaterials.While the later- southeastern- portionof the
site is similarlycoveredwith brokenstones,a searchof its surface failed to reveala
single undeniablestone implement.Lying amongthe core tools on the earlysection
of the site are a numberof obsidianprojectilepoints and flakes,most of the latter
showingmarksof utilization.
The large quantityof core tools is reminiscentof the site of Taltal in north
Chile.8 However, our Site 20 cannot readily be assigned to the same culture as
7 Engel, Les amas de coquillagesde la cote phruvienne.
8 Bird, Excavationsin Northern Chile; see especiallypp. 286-289.
62 SOUTHWESTERN JOURNAL OF ANTHROPOLOGY

Taltal.The choppers
aregenerallymorefullyflaked,withno cortexremaining
on
the implement.A series of fist-sizecore scrapersfrom Site 20 retain cortex on
thelowerface,buttherearefewof thetruepebblechoppers so commonat Taltal.
Projectilepoint typologyis also different.Site 20 yieldssmall,thick, narrow,
leaf-shapedspecimenswhichprobablywereusedwiththe dartandspear-thrower,
butnoneof the broadertypesor stemmedspecimens whichoccurin northChile.
Unfortunately,we had only one day availablefor test excavations,and
have not yet had opportunityto returnto the site. A single pit two meterssquare
was begun,but couldnot be carriedto the bottomof the midden.The sample
is thereforesufficiently
smallthat as yet we do not knowwhetheror not other
elementsof the northChileanculturesoccurat Site 20.
The testpit showsa well-stratified
deposit.Onlya littleovera meterwasexca-
vated,comprising the uppertwo levelsandpartof the third.Thesethreelevels
representthe remainsof a culturewhichpracticedgardening,since seeds of
lucumaoccurin theuppertwolevels,andthethirdyieldeda gourdfragment.The
bulk of the diet seemsto haveconsistedof shell-fish,principallythe smallrec-
tangularclamMesodesmadonacium.Crustacean remainsand bonesof sea lion
occurin moderatequantities.Bonesof birdsand fish are rare,and no fish-net
wasfound.
The bulkof the artifactsfromthe testpit arepercussion-flakedchoppersand
corescrapers,whichoccurin highquantityin levels2 and 3, but of whichonly
twocomefromthe upperlevel.Commonin all threelevelsaresmallovoidwater-
wornpebbles,batteredat one or bothends,and utilizedflakesof quartziteand
obsidian.Level2 produceda fragmentof a roughdeep-basin metateof sandstone,
and Level 3 a large quartzitescraper.
Site 20 promisesto be interesting,
andthe authorhopesto returnto conduct
furtherexcavations.The depositis apparently a deepone,and the artifactyield
veryhigh.The siteis thusan excellentonefor thestudyof the pre-ceramic
period
in Peru.
SITE 15. CHIRA/VILLA
Site 15 lieson the HaciendaVilla,southof Limaandjust southof the large
Incaperiodsiteof Armatambo. The sitelieson thenorthern slopeof CerroChira,
inlandfromthe seawardslopeof the hill. It is protectedfromthe offshorewinds
by a lownorthernextensionof the hill, whichseparatesthe site fromthe beach.
To the west of the hill, alongthe sea side, thereonce existeda marshyplain,
wherewaterbirdswouldhaveflocked.Fishingis goodoff PuntaLa Chira,just
overthe hill fromthe site, and fishermenfromthe suburbsof Limaregularly
exploit this stretchof coast.
 EARLY SITES ON PERUVIAN COAST 63

The steephillslopeabovethe occupation siteshowsthe remainsof formerter-

races,undoubtedly levelledfor agricultural purposes.The site itself lies on the
lower,less steep,slope.Its surfaceis coveredwith shellsand ashes,with occa-
sionallargestonesindicatingthe locationof fieldstonewalls.It coversan area
aboutseventy-five metersin diameter,witha thinnerscatteringof shellandashes
extendingaway from the hill for anotherfifty metersto the northandeast.The
middenis denselycompacted andrichin blackash.It attainsa maximum depthof
abouttwometers,thoughit is generallysomewhatthinner.Excavations haveen-
counteredfield-stone wallsat variousdepths,and it is hopedto followtheseout
at the firstopportunity. Full-scaleexcavations are urgentlyneeded,as the site
is soonto belevelledas agriculturalland.
Three test trenchesand a numberof pits have revealedgood visiblestra-
tigraphy,and a culturalsuccessionof considerable interest.Althoughtwo Early
Limadecorated sherdsweremixedwiththesurfacedebris,the latestoccupation of
thesiteseemsto havebeenin pre-Chavinoid times,andthelowerlevelsrepresent a
peoplewhopracticed gardening, butlivedin pre-pottery times.
Five stratigraphiclevelsextendthroughoutthe site. Of these,the most im-
portantareLevel2, whichyieldsthe fullestsampleof thelater- earlyceramic-
culture,and Levels4 and 5, whichcontainthe pre-ceramic culture.Level 3,
althoughrichin blackashandorganicmaterial,andgenerallyquitedeep,hasan
extremelylow artifact content. Level 1, about ten centimetersdeep, is a mixture
of blownsand,midden,andartifactsof all ages.It evenproducessphericaliron
shotwhichprobablydatefromthe warwithChile.
Remainsof lucuma,pacae,gourds,andcotton,indicategardeningin boththe
pre-ceramic and the earlyceramicperiods,andpeanutsare foundin the midden
corresponding to the latter.A few corncobshavebeenfound,but thesearenot
associated
definitely withtheearlycultures,as theyoccurred onlyin Level1 andin
a deeperbutdisturbed section.
Gardening seemsto havebeena minoroccupation at Chira/Villa.Throughout
the depositenormousquantitiesof shellsand boneindicatethat subsistence was
on
basedprimarily hunting,fishing, and the gathering of sea food. Mesodesma
donaciumseemsto havebeenthe majorfood itemin all periodsof occupation.
Musselshells- particularly MytiluschorusandM. magellanicus - occurin some
quantity,as do theshellsof Concholepasconcholepas, Thaischocolata,Crepidula
onyx,andCrepidula excavata.Otherspeciesof shellfisharerare.
Bonesof birds,fish,andsea mammalsoccurin largequantities.Particularly,
thereare thousandsof bonesof cormorants, whichare abundantaroundPunta
La Chiratoday.Mammalbonesarealmostexclusively of sealion.
Gardeningland apparentlylay on the hillsideterracesabovethe site, and
64 SOUTHWESTERN JOURNAL OF ANTHROPOLOGY

probablyalsoin theflatbottomlandto theeastof thehill,whichtodayis irrigated

and cultivated.Sea food, includingshellfishand the algaepopularlyknownas
cirueladel mar,couldbe gatheredoff PuntaLa Chiraandalongthe coastto the
northand south.The sameareawouldserveas both huntingand fishingsite.
Strangely,there is little evidenceas to how cormorantsand sea lions were
caught.Overonehundredcubicmetersof excavated middenhaveproducedonly
oneprojectilepointanda fragmentof a second,anda singleartifactwhichmay
representa stone spear-throwerweight. At the same time, cotton net occurs in
fair quantity,thoughfishbonesare considerably less frequentthanbirdbones.
It is quiteprobablethatbirdswerealsonetted.Althoughno fishhookshavebeen
found,it is quitepossiblethat fish weretakenwith the line. Stout cottoncord
occursin variouslevels,and remainsof muimuy(Emeritaanaloga)probably
represent fish-bait.In addition,a sectionof Mytiluschorusshellwitha conically
drilledperforation, fromLevel4, probablyrepresents a blankfor the manufac-
tureof a fishhook.
Heavy stone implementsare of the quartzitewhich outcropslocally on Cerro
Chira,with a few specimensof basaltfromneighboring hills. These includea
fewpercussion-flaked choppers and hammerstones from both the pre-ceramic and
the earlyceramiclevels;a long percussion flakedpick fromLevel2; a number
of roundedcobblesusedas hammerstones andtwousedas pestles(thelatterboth
fromLevel2); a fragmentof a well-made basaltpestlefromthe surface;a few
large discoidaland sidescrapers; and a crudepercussion-flakeddrillfromLevel4.
Unmodifiedquartzcrystalsandtinywater-polished pebblesoccur withsomefre-
quency in the upperlevels, but arerare belowLevel3. Ovoidcobbleswithtraces
of redpigment- probably ocher- arealsocommon,anda numberof thecobble
hammerstones similarlyshowtracesof pigment.
The onlypressure-flaked stoneimplements fromVilla/Chiraaretheprojectile
pointandfragmentalreadymentioned. The completespecimenis similarin shape
to thosefromSite20, butthickerandverywellretouched, withfineparallelflake
scarsacrossbothfaces.It comesfromLevel4. Polishedstoneimplements aremore
frequent.They includetwostonespindlewhorlsanda fragmentof a finebasalt
bowl,fromthe surface;half of a tubularquartziteartifact,in whichare the
remainsof a caneset withresin,andwhichmayrepresent a spear-thrower weight
(Level2); and half of a "doughnutstone"or perforatedclub head,fromthe
surface.Bone implementsincludepointedawls or sewingtools and bird bone
tubes,fromboththe pre-ceramic andthe ceramiclevels;two mammalbonetubes
fromthe pottery-bearing levels;and a polishedneedle,a spindlewhorl,and a
fragmentof incisedwhalebonefromthe lowerlevels.Shell artifacts,exceptfor
the fishhookblankalreadymentioned,are ornamental. They includeshellsof
Fissurella peruviana,Olivaperuviana, and Tapes (?) sp., withthe apexor spire
 EARLY SITES ON PERUVIAN COAST 65

groundoff;grounddiscsandtrapezoidal ornaments of Mytiluschorus,M. magel-

lanicus, and Pectenpurpuratus; and a roughhexagonalornamentof thickclam
shell.
Woodenimplements areknownprimarilyfromthe surfaceof the site.They
includea seriesof blunt-ended sticks (one fromLevel5), groovedand pointed
wooden"tops"(one fromLevel2), a fragmentof a firedrillhearth(Level4),
and a cut sectionof cane,one end of whichis sealedwith a gourddisc.There
is alsoonerimsherdfroma gourdbowl,fromLevel2.
As mentionedabove,the upperlevelsof the site haveproduceda quantityof
potsherds. A veryfew arealsoknownfromthe lowerlevels,butthesecomefrom
areasnearpocketsof middenintrusivefromLevel2, andprobablyarenot associ-
ated with the cultureof the lowerlevels.The two decoratedsherds,from the
surface,clearlydo not belongwiththe assemblage. Sherdsoccurin Levels1 and
2 in relativelysmallquantities,as comparedwith latercoastsites.The warefits
well withWilley and Corbett's"AnconPebblePolished"type,whichoccurred
unassociated withdecoratedsherdsin the lowerlevelsof the Chavinoidmidden
at Ancon.9It is coarseandpoorlyfired,witha considerable quantityof unevenly
selectedsandtemperanda highcontentof minutemicaflakes.The innersurface
is left rough,the outergenerallypolishedin streakswitha polishera centimeter
or lessin width.The warerangesin thicknessfrom2.0 to 6.6 mm.The onlyforms
areollaswitha characteristically thickenedrim,generallyangularin crosssection;
anda singleopenbowl.
Twinedcottontextilesoccurin somequantity,associatedwith both the pre-
ceramicandthe ceramiccultures.Mostof the textilesaretwined,figure-8looped,
or knitted,somewovenonesbeingpresent,as well as matting.The most com-
mon techniqueinvolveswidelyspacedpairsof wefts, one memberof eachpair
passingovereachwarp,the otherunder.The techniqueis the sameas that used
for makingcanematsin all periodson the Peruviancoast,exceptthat the mats
aremadewithcanewarpsand ropeor cordwefts.
Our excavations at Chira/Villahavebeenmuchmoreextensivethan at any
of the othersitesreportedin thispaper.However,exceptfor ceramics in the upper
levels,and textilesin the upper and lower levels,the artifactyield is verylow,
and we do not yet havea sampleadequatefor the definition
of the culturesin-
volved.Workat thesite willbe resumedin thenearfuture.

SITE 5. PLAYA CULEBRAS

sites can be seen aroundthe narrowplain,some two
Many archaeological
kilometers
wide,which used to be- probablythousands of yearsago- the delta
of theCulebrasRiver.Archaeologically this
speaking, zoneis richandinteresting.
9 WilleyandCorbett,EarlyAnconand EarlySupeCultures,p. 54.
66 SOUTHWESTERNJOURNAL OF ANTHROPOLOGY

Here,we shallconcernourselvesonlywithone of thesesites:the one occupying

thenorthernslopeof the hill closingthe bayof Culebrason its southernside.At
thatpoint,thesandsandbrushof thebeachmeettheslopeof an abruptandrocky
hill.The hill constitutesa perfectlocationfor shelteredandprotecteddwellings:
accessis not easy,and one is safe fromthe southwestwinds.The locationwas
thusexcellentfor the terraceddwellingswhichcanstillbe seen.
Life conditionsherehavechangedcompletely duringthe last centuries.Some
fishermen still liveon the beachof Culebras,but theyfishin deepwatersandthe
shellfoodhasdisappeared. The freshwatersupplywhichmusthavebeenknown
to theancientinhabitants hasnowdisappeared. Fuel,at least,is stillavailablefrom
thetreesof thenearbyCulebrasvalley.
Theslopeof thehillis coveredwithstonework,mainlysquareterracesmarked
byhalf-buried cornerslabs,whichcanbe seento thetopof thehill.Refuseconsists
of sand,organicmatter,shells,ashes,and erodedstonesof the rockysoil. The
occupation areais someeightymetershigh,and formsa semi-circle of somehalf
an hectare.No potteryis to be seenon the surface.
Occupationmust have beendense;test pits at variouspointsalwaysmet graves
or houseslyingone nextto the other.Near the crestof the hill, facingthe sea
towardsthenorthwest, is a rectangular
terracefacedby a rowof erectstonessome
120cmhigh.
Duringthe courseof test excavations, we wereable to cleana few houses.
Theseweresquareor rectangular, and variedin size from 1.5 to 3 meterson a
side.One hada clayfloorfivecentimeters thick,and an emptyfuneralchamber
at its northernend. The fill of the housewas composedof blackash, carbon,
shells(mussels,clams[Mesodesma donacium],anda few Pectenpurpuratus and
landsnails),twigsandsticks,seamammalbones,algae,andcharredhumanbones.
An adjoininghousehad a floormadeof rectangular stoneslabslaid flat on
the bedrockandsealedwithclay.Oneof theseslabswassquareandpolishedfor
ornamental purposes.
PlayaCulebrasis unusualas a pre-ceramic site in that a considerable partof
the site formsa cemetery.We excavatedtwo individualtombs,and threetombs
in a structurewhichcontainsmultiplestone-walled funeralchambers.
Tomb1 lay in a terracethreemeterslong and2.5 in width,facedwithsmall
quadrangular, tabularstones,placederect.Fromjust belowthe surfaceof this
terracecamethe only fragmentof potteryfoundduringthe test excavations, a
poorly-fired,irregularly
polished sherd with abundant sand temper, of the type
described abovefromChira/Villa.
The buriallayat a depthof eightycentimeters in the terrace,in a fill of brown
earth.The upperpartof thetombwaslinedwithlargeflatcobbles,rectangular or
 EARLY SITES ON PERUVIAN COAST 67

square,someerectand otherslaid flat Similarlargecobblesappearsporadically

in the wallsof the lowerpartof the tomb.The structureis set in a cut in the
rockyhillside.The skeletonwasthatof a youngmale,lyingon his backandwith
the head orientedwest.The skull showsoccipitalflattening.Badly carbonized
mattingof the reedknownlocallyas juncolay aroundthe bones,andmayonce
havecoveredor underlain the body.Otherassociated objectsaretwo cylindrically
drilledrayvertebrae; twomusselshells,containing scrapsof twinedclothof totora
fiber; anda wadded fragment of twined cotton cloth.
Tomb2 lay in the samesmallterrace,fifty centimeters deeperthanTomb 1.
It is a smalloval chamber,70 cm by 40 cm by 30 cm, cut into the rockof the
hill. The burialwasthat of an elderlymale,flexedand lying on the rightside,
with the skull orientedwest.This skull was also occipitallyflattened.The ac-
companying artifactswerea pebblechopper,a scraper,a percussion-flaked pick,
anda twinedfiberheadband.
The necropolis,mentionedabove,is facedwith a wall of largestoneslabs,
3.6 meterslong,runningeast-west,some15 metersabovesea leveland 30 meters
distantfromtheriverbed.The endsof thiswallcurvein towardthenorthto form
shortlateralwallsfrom1 to 1.5meterslong.Two largereinforcing slabs,30 cm.
thickandcementedwithclay,seemmadefor thepurposeof diminishing thepres-
sureof the hill slopeon the centralpartof the east-westwall.In addition,all of
the stonesof thewallarecementedwithclayto the hillside.
Therearethreenicheson the insideof the long wall,andone on eachlateral
walLAll aresquare,30 cm on a side,and20 cm deep.Eachis surmounted by a
lintelstone,andbackedby a cementedstoneslab.
In front of this long wall are manygravesof whichwe excavatedthree.
GraveA containedtwo skeletons,respectively 1.65and 1.8 metersfromthe top
of the wall, lying one abovethe otherand separatedby a layerof clay. Both
wereextended,face down,orientedwest. The lowerwas an adult female,the
uppera child.The femalehada twinedjuncomatwrappedaroundthe head,and
a gourdfull of a chalkywhitesubstance,probablypowderedlime,lying beside
theskeleton.Remnants of severaltwinedmatsandbasketslayaroundbothbodies.
GraveB, locateda little northof GraveA, containeda singledecomposed
skeleton;a curvedPectenshellornament, ground,polished,andbiconically drilled
at one end;an Olivaperuanashellwiththe spirecut off to makea bead;anda
fiberstringknottedto holda bolastoneor fishlinesinker.
GraveC, west of graveA and nearthe westernlateralwall, containedtwo
burials.The upperof thesewasextendedon its left side and orientedwest,and
was accompanied by a spire-lopped Olivashell, a cut gourd,a hammerstone, a
pebblechopper, and by a necklace of ninelittleclam shellswiththe beak abraded
68 SOUTHWESTERN JOURNAL OF ANTHROPOLOGY

off.The lowerskeletonlayin thesameposition.Next to thebodywerefourpebble

choppers,a scraper;two water-polished pebbles,a gourdcontainingpowdered
lime,a mussel-shell
fishhook,twelveshellsof MytiluschorusandM. magellanicus,
threeof whichcontainedpowdered lime,a fragmentof mattingpaintedwithred
pigment,probablyocher,andtwojuncobaskets,oneof whichcontaineda mussel
shell.
As alreadyindicated,workon thesesitesis still in progress.The authorplans
notonlyto extendhissurveyto thenorthandsouthof theareaherereported, but
alsoto continuetestingthe sitesmentionedin this reportandto undertakemore
extensiveexcavationsat someof them.Meanwhile, this information
is offeredas
recordof pre-ceramicsiteson the Peruviancoast,andas an indication of research
in an
possibilities important and littleknown fieldof archaeology.
LIMA,PErU