Sepúlvedaetal.HeritageScience2014,2:7 http://www.heritagesciencejournal.com/content/2/1/7 RESEARCH ARTICLE Open Access Study of color pigments associated to archaic chinchorro mummies and grave goods in – Northern Chile (7000 3500 B.P.) Marcela Sepúlveda1*, Hélene Rousseliere2, Elsa Van Elslande2, Bernardo Arriaza3, Vivien Standen4, Calogero M Santoro3,5 and Philippe Walter2 Abstract Thisarticle presents theresults of physical–chemical characterization ofthelayers ofcoating covering the bodies, faces and stuffed of four mummified human bodies and seven gravegoods from the Chinchorro Archaic funerary tradition that inhabited theAtacama Desert ofthe far northerncoastof Chile. Using ScanningElectron Microscopy withEnergy-Dispersive X-Rays (SEM-EDX) and X-Ray Diffraction(μDRX), thestudy identified theuse of different pigments including iron, manganese and copper oxides that were combinedinto different preparationsor pictorial recipesforpasteandcoatingbodiesandfaces.TheresultsindicatethattheChinchorro,anarchaichunting,gathering andfishingsociety(7000-3500B.P.)managedacomplexcolortechnologyfortheirmortuarypractices. Keywords:Pigmentidentification,Mummification,Colortechnology,NorthernChile,Chinchorro Introduction whichtheyusedtocreatetheirmummies,thefirsteverin The Chinchorro Tradition extended from 7000 to 3700 humanhistory[5].Themummificationtechniquesvaried, BP, which corresponds to the Archaic Period in the both in terms of how they prepared, emptied and stuffed Andeans Region. These hunting, gathering and fishing the dead bodies, and the application of different external people inhabited Southern Peru and Northern Chile, finishes, and the latter feature has been used to classify living along the coast taking advantage of a wide and the different funerary rites of these archaic coastal groups abundant range of coastal marine resources generated [6-8]. Three main forms of complex mummification by the Pacific Ocean’s cold Humboldt Current [1]. As have been identified among the Chinchorro: the Black they settled on particular coastal spots huge accumula- Mummy Technique, the Red Mummy Technique and tion shell middens have enable us to construct their the Bandaged Mummy Technique. The Black Mummy daily lives, while their extensive cemeteries, found on Technique (ca. 7000–5000 BP) involved removing the the slopes of coastal bluffs and within the domestic organs, cleaning and reinforcing the skeleton then cre- areas, contain multiple overlapping packaged burials. ating a clay mold of the body. Once the body was Studies of the Chinchorro funerary procedures began in sculpted, the skin was replaced and painted over in a the early 20th century and we now know that, in shinyblackpigment.Thefacewascoveredwithaclaymask addition to having a specialized technology well suited painted in the same way. The Red Mummy Technique to extracting coastal resources, these groups developed a (4500–4000BP)differsfromthatoftheblackmummiesin variety of complex methods for treating and preserving that each cavity of the body and head were emptied and the bodies of their deceased, including artificial mummi- then filled with a variety of materials (plant fiber, clay and fication [2-4]. For 5000 years these people developed feathers, among others). Once filled, the body was covered extensiveknowledge of humananatomyand morphology, with red paint. A wig with long black hair was placed on the head of the deceased and kept in place with a black- colored cap. The face was covered with a black mask. The *Correspondence:[email protected] 1DepartamentodeAntropología,LaboratoriodeAnálisiseInvestigaciones bandaged mummies are avariant of the red ones in which Arqueométricas(LAIA),UniversidaddeTarapacá,Arica,Chile theskinwaswrappedonthebodylikebandages[9].Unlike Fulllistofauthorinformationisavailableattheendofthearticle ©2014Sepúlvedaetal.;licenseeChemistryCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsofthe CreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse, distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycredited. Sepúlvedaetal.HeritageScience2014,2:7 Page2of12 http://www.heritagesciencejournal.com/content/2/1/7 other mummification traditions around the globe [10,11], the clay compounds were identified. These results ad- the Chinchorro mummified adults, infants, newborns and vanced our understanding of how malleable clays were even fetuses. In short, in practicing mummification or usedtoconstructthemummifiedbodiesandfunctionedto “the materialization of death” [12], these coastal hunter- preservethem.Theabsenceofcolorpigmentsinthesegray gatherers employed complex techniques to prepare the claysfillingandthemud-coatedmummywasalsonoted. body and used a wide variety of materials available in Lastly, continuing the emphasis on this type of com- their surroundings, including wood, plant fiber, animal pound, recent studies have addressed the chemical and skin and hair, ash, clay and pigments. mineralogical composition of the gray clay used as stuff- In this paper we will analyze the coating that covers ing in the black mummies [13]. The results obtained artificially mummified bodies, one sample from stuffed from EDXRF, SEM-EDX and μXRD analyses demon- of mummy, as well as the color that appears on items strate that the clays are complex compositions based on included among the grave goods that accompanied the quartz, albite, sanidine and muscovite occur naturally on bodies from different coastal sites (Figure 1). Specifically, the banks of rivers in the region. The gray clay was used four samples from a single black mummy and two red to completely model the body including the face of the mummieswereanalyzed. Ninesamplesfrom Chinchorro black mummies. The clay was collected and used in grave goods are also analyzed in order to contrast these their natural state, without grinding or the addition of results with those obtained for the mummies. The othercompoundsorcolor pigments. samples were analyzed using Energy Dispersive X-Ray In conclusion, for more than a decade different analyt- Spectroscopy (SEM-EDX) and Micro X-Ray Diffraction icaltechniqueshavecontributedtostudythedifferentele- (μXRD). Taken together, the data provide important ments and compounds used in making the Chinchorro technical information on the pigments used and shed mummies. Despite the color analysis, however, the min- lightontheircolortechnology,understoodasknowledge eralogicalnatureofpigmentfromcoatingsandstuffedhas set and productive processes to obtain the colors they still not been identified. Furthermore, the composition of used. It also contribute to debate symbolism, ritual the coatings must be distinguished from that of the stuff- practices and ceremonialism of the Chinchorro groups. ing. In other words, it is useful to specify whether the coatings correspond to a single layer of paint or several Previousanalysis layers, as someauthorshave proposed before but without Previousstudieshaveusedphysical–chemicaltechniques analytical or empiric proof [19,20]. Lastly, it is necessary to characterize certain elements and compounds used in to compare these compounds with those present in grave the coating and stuffing of the Chinchorro mummies. goods, so we can identify if and how these artifacts were Specifically, Energy-Dispersive-X-Ray-Fluorescence ana- associated with the materialization of the bodies them- lysis (ED-XRF) was used to analyze the elements con- selves. More specifically, we understand the artifacts and tained in the coating used on several mummified bodiestobepartofthesamecolortechnologyproduction Chinchorro individuals [13-15]. In summary, the results sequence, defined as the complete production process, indicated that more manganese was used in coating the from the procurement of the raw materials to their appli- red mummies (than the others), and this finding was cation on a support [21,22]. For this study, we ultimately confirmed by later Atomic Absorption Spectrometry chose to prioritize non-destructive techniques in order to (AAS) [16,17]. While these ED-XRF and AAS analyses preservethesamplesforuseinlateranalyses. helped to recognize the elements present in the coating, the techniques employed did not allow us to identify the Material and methods type of manganese mineral used. Furthermore, the tech- Two samples were extracted from the coating on the niquesusedalsorequiredcrushingofthesamples,which thorax of an adult male black mummy (Body 1) from prevented us from identifying the different minerals the site of Maderas Enco (Maderos Enco C1; Figure 2a). presentinthepaintinthecaseofmixtures.Additionally, The laboratory samples were labeled M.26 and M.27. it was not possible to identify layers beneath the surface Two other samples were taken from red mummies: coating, in the sense of differentiating the surface layer Sample M.19 is from the coating on the mummified from possible underlying or priming layers that are body of an infant (0-4 months) from the Morro 1 site macroscopically visible. (M1 Tomb 25, Body 5) (Figure 2b). Sample M.35 came Another study of the gray sediments used to reshape from the head coating of an adult individual from the or remodeled three black Chinchorro bodies and the Playa Miller 8 site (PLM8 Cranium 1) (Figure 2c). A internal filling and external surface of one mud-coated green sample obtained from the stuffingof a mummified mummy was undertaken by [18]. Using a Scanning infant (M.18) from the Estación Sanitaria site whose Electron Microscope (SEM) coupled with an Energy color had not been studied to date, was also analyzed Dispersion Detector (EDS), the size and morphology of (Figure 2d). Lastly, the analysis examined nine samples Sepúlvedaetal.HeritageScience2014,2:7 Page3of12 http://www.heritagesciencejournal.com/content/2/1/7 Figure1MapofLocalisationofsitestudied,northernChile((c)FONDECYT1100354directedbyDr.MarcelaSepulveda). from red- and black-colored grave goods, as pigment in 450 to 490 nm; barrier, 520 nm; dichroic mirror, 505 nm) shell, powder in leather pouch or raw pigment, as pig- and a Scanning Electron Microscope (SEM, Philips XL ment in shell or raw pigment (Table 1), in order to com- 30 CP), using secondary electron (SE) and backscattered pare them with the results obtained from the bodies. All electron(BSE)modes. samples come from important burial sites located on the After being observed front and back under an optical coastline. They correspond to the largest cemeteries of microscope, the samples were immersed in a resin Archaicperiod inAndeanRegion. (Resine Polyester SODY 33) for complementary obser- The samples were collected and packaged, at the vation and analysis. This kind of preparation is com- Archaeological Museum of the Universidad de Tarapacá mon in the identification and analysis of overlapping in San Miguel de Azapa, before being shipped to the layers of paint primers and varnishes on historic paint- LAMSandC2RMFlaboratories(Laboratoired’Archéologie ings [23]. It has also been useful for analyzing parietal Moléculaire et Structurale, Université Pierre et Marie and rock painting [24]. In that state, the resins were Curie in Paris; Centre de Recherche et de Restauration polished on different surface carefully up to the edge des Musées de France in Paris, respectively) in France. of the sample then analyzed using the techniques In the lab, each sample was observed in sequence proposed [25]. Then, we can observe stratigraphy of under a Nikon SMZ-10A binocular microscope and at different samples. We also recognize the form and size different magnifications under a Nikon-Labophot (5X, of inclusions, differentiating big or large ones with size 10X, 20X, 50X, 2- Xenon lamp, B-2A filter; excitation, between 50 and 100 μm, from small under 20 μm. Sepúlvedaetal.HeritageScience2014,2:7 Page4of12 http://www.heritagesciencejournal.com/content/2/1/7 Figure2Samplefromdifferentkindofmummies.a)MaderasEnco;b)Morro1;c)PlayaMiller8;d)EstaciónSanitaria. Afterwards, each sample was analyzed using Energy device consists of a X-ray tube with a copper anode Dispersive Spectrometry (EDX) coupled with the SEM (wavelength of λ=1.54186 Å), a Kirkpatrick-Baez optical in order to obtain a qualitative identification of the ele- system and a Rigaku R-Axis IV 2D detector imaging ments present in the samples. The analyses were per- plate. The X-ray beam is collimated to a diameter of formed in BSE mode (20 KeV, 1000X-3000X), with both 250 μm, while an incident angle of 6 to 10° enlarges the point and zonal identifications. Mapping of the most beam impinging the sample up to one millimeter. The characteristic chemical elements in the sample was also diffracted X-rays are emitted in a 2θ cone and collected performed to assess location and dispersion. To prevent on the imaging plate, from which the diffractogram is any charging effect and obtain images with the highest extracted. The data were processed using the FIT2D resolution, each sample was covered with a fine layer of program (http://www.esrf.eu/computing/scientific/FIT2D/), carbon. The INCA 4.15 program was used to acquire while identification of crystalline phases was performed thedata. usingtheEVAprogram(http://www.bruker-axs.de/eva.html). Lastly, the use of a Micro-Diffraction X-Ray device The analyzed thickness of the material depends on the (μDRX), an homemade system in the C2RMF, was used weight on the elements: it is approximately 25 μm to identify the crystalline phases in the pigments. The when it is composed by low Z elements (Al, Si, K) and Table1ResultsofgravegoodsanalysesbySEM-EDXandμXRD Color Sample Site Description SEM-EDX Grainshape X-raydiffraction Black M1 CA15-C Pigmentinashell Mn,Si,Ca,K,Al, Manganesecrystalsnotclearlyobserved Cryptomelane,quartz, Fe,Mg,Cl,S,Na (mixedwiththeothercompounds) anhydrite,calcite,diopside M17 MO-1Tumba25 Pigmentinashell Mn,K,Si,Ca, Manganeseas2-3μmlonggrains Cryptomelane,halite,quartz, Cuerpo5 Cl,Al,P,S,Na calcite,albite Red M1 MASMA-1 RawPigment Fe,Si,Al,Ca, Hematitesphericalgrains,10μmin Hematite,halite,kaolinite? Cl,Na,traceofAs diameterorovoidegrains(1μmthick) M8 CA15-C, Rawpigment Fe,Si,Al,Cl,K, Hematitesphericalgrains, Hematite,goethite,halite, Cuerpo3 S,Na,traceofAs 1μmindiameter quartz,calcite M10 COLON-10 Pigmentinashell Si,Fe,Ca,S,Al, Hematiteovoidgrains,1μmthick Hematite,quartz, Cl,K,Na,traceofAs halite,anhydrite M11 MASMA Pigmentona Fe,Si,Ca,Al, Hematitephericalgrains,10μmin Hematite,halite,kaolinite gridingstone Cl,traceofAs diameterorovoidegrains,1μmthick M13 CA15-C Powderina Si,S,Fe,Ca,Al, Hematitecryslatswithfacets,2-7μm Hematite,quartz, leatherpouch Cl,Na,traceofAs longandovoidgrains(1μmthick) calcite,halite M14 CA15-D Rawpigment Fe,Si,Al,K,Na, Hematitesphericalgrains,10μmin Hematite,halite, Cl,Ca,traceofAs diameterorovoidegrains(1μmthick) diopside?illite? M16 CHIN-4 Rawpigment Fe,Si,Al,S, Hematiteovoidgrains,1μmthick Hematite,gypsum,quarz, P,K,Na,Ca halite,calcite? Sepúlvedaetal.HeritageScience2014,2:7 Page5of12 http://www.heritagesciencejournal.com/content/2/1/7 about 5 μm for heavy ones, like Pb, Hg. The geometry samplesandistheprimarycomponentoftheredsurface of the system is calculated using an alumina powder layer. Other minerals identified included calcite, quartz reference at the beginning of a series of measurements. and a little manganite (MnOOH) (Figure 4). Lastly, the The acquisition time for one measurement is of 5 mi- analysis identified a little cryptomelane (KMn O ), the 8 16 nutes. The spatial resolution of the system does not manganese mineral used in the underlying brown layers allow to characterize each layer on the samples cross- observed under the binocular and optical microscopes. sections because the thickness of the colored materials In some cases there is a higher proportion of cryptome- is less than the size of the beam. To have a higher lane and less hematite. Other peaks of diffractogram spatial resolution, it is necessary to use other instru- were notidentified. ments at synchrotron radiation facilities [26]. Microscopic observation of Sample 27 showed that 5 layers of coating are also present (Table 3). Elemental Results analyses were performed on the different layers of this Analysisofthecoatings sample using SEM-EDX. Subsequently, the elemental Sample 26 of the coating displays, from the outside in, 5 distribution was mapped using the same technique in layers visible under binocular and optical microscopes order to assess co-localizations, such as potential associ- (Table 2; Figure 3). These layers have similar composi- ations between Al and Si, among others. Layer 1, the tions in terms of the type, size and shape of their outermost pigment layer (thin almost 20 μm), is com- components and inclusions, except for Layer 1 and the posed of an aluminosilicate (co-localized Si and Al). interfaces between layers 2-3 and 3-4. Layer 1 corre- Lower layers of the coating (thin 170 μm) were com- sponds to a thin layer of red paint (35 μm). Layer 2 posed of a material rich in manganese, mixed with (thin almost 90 μm) is brown in color, with large black maybe a clay (Si, Al, K, Mg, Na) rich in iron oxides. inclusions within a brown matrix. Layer 3 (thin almost Some grains of calcium (possibly calcite?) are at times 180 μm) presents similar black inclusions, as well as co-localized with sulfur (gypsum?). In the layers of coat- large red inclusions in a light brown matrix. Layer 4 dis- ing it is possible to distinguish different compositions plays a more homogeneous composition in terms of the based on the quantity and size of the manganese grains size and color of the inclusions, which are embedded identified. In layer 2, for example, the manganese primar- within a brown matrix similar to the previous layer (thin ilydisplaysaslargeblackgrains,whilelayer3displaysthe of175μm).Lastly,layer5(thinalmost 500μm) alsodis- largest amount of evenly distributed manganese, with plays very large black and grey inclusions in a grayish- grains that are generally smaller, although larger ones are black matrix. One notable feature of this sample, found also present. Thin of Layer 2 is almost 200 μm. Layer 4 in the interfaces between layers 2 and 3 and 3 and 4, is (thin 340 μm) contains manganese but in this case the the presence of very thin layers of yellow-colored pig- grains are small and evenly distributed, and large grains ment, which may be the remains of surface paint later are much less abundant. Lastly, in layer 5 the manganese covered over with the darker brown layers. This sample ismuchlessabundantthanintheotherlayers.Wecannot was only analyzed using μDRX in order to test the dif- obtain precise thinness for this layer. Using μXRD, we fraction technique on this kind of sample. The diffracto- identified the presence of crystalline manganese in the grams obtained during first analyses were difficult to form of cryptomelane and manganite, as well as quartz. read because of the heterogeneity of the sample and the Also identified were albite, calcite and a little kaolinite. variety of minerals present. Despite the difficulties, how- Some elements as Fe, S and Na suggest the presence of a ever, the presence of hematite was confirmed in all smallamountofnatrojarosite(NaFe (SO ) (OH) ))[27]. 3 4 2 6 Sample 19 (M.19)wasdifficulttoprepareinresin. Ob- Table2DescriptiveofidentifiedlayersofM.26(ARI-8), servation under an optical microscope (white and ultra- fromthesurfacetothecoating violet light) showed that the sample consisted essentially Layer Color Thickness(μm) of skin with a thickness of 350 to 450 μm. For the same 1 Red 35 reason the sample could not be analyzed using SEM- EDX. DRX analysis showed the presence of hematite, 2 Darkbrown 90 quartz, calcite, albite and kaolinite, despite the difficulty 3 Lightbrown 180 inobservingthese compounds. Onthesurfaceofthislayerwecanobserve Under microscopic observation, Sample 35 (M.35) did bigyellowgrainmaybeasanolddecoration not reveal a layer of surface paint, but a heterogeneous 4 Lightborwn 175 layer approximately 200 μm thick was observed, com- Onthesurfaceofthislayerwecanobserve posed of large black grains approximately 100 μm long thesameyellowgrainsfromLayer3 contained in a brown mixture of smaller sized black, 5 Grey-black 500 red, white and translucent minerals. Also, no layer of Sepúlvedaetal.HeritageScience2014,2:7 Page6of12 http://www.heritagesciencejournal.com/content/2/1/7 Figure3TransversalcutofM26.ObservationbyOpticMicroscope. paint was visible on the surface under SEM-EDX, and it as observed in the sample slices and as indicated in the is therefore likely that this individual’s mask was not results previously obtained through EDXRF and AAS. painted, but rather obtained its blackcolor from the pig- One notable finding is that samples M.26 and M.27 ment contained inthe moldingplaced overthe face. The (coming from a black mummy coating) empirically prove spectrum obtained by EDX indicates the co-localized the presence of several overlapping layers of coating. presence of silica, aluminum and potassium, indicating These coatings are composed of a clay material (with that the sample may be a type of clay. Also observed quartz, calcite, albite, kaolinite and calcium sulfate) and were large grains of silica, possibly sand, and a com- a compound rich in manganese (cryptomelane and pound rich in iron and sulfur. μDRX analysis confirmed manganite). The relative proportions of these two main the presence of cryptomelane, hematite, quartz, calcite, constituents varies from layer to layer, as does the size halite, albite and kaolinite. A jarosite mineral, probably the inclusions, and it therefore follows that they are dif- natrojarosite, was also identified, but should be inter- ferent preparations made with the same raw materials. preted cautiouslyuntilitcan be confirmed. NotableinthecaseofSampleM.26arethetwointerfaces In general, the samples analyzed display similar char- betweenlayers2-3and3-4,whichcouldbeinterpretedas acteristics in terms of shape and chemicals composition. decorations applied before repainting. The layer of red The size and relative proportion of the minerals do vary, pigment on the surface corresponds to the decoration Figure4DiffractogramofM.26fromMaderasEnco. Sepúlvedaetal.HeritageScience2014,2:7 Page7of12 http://www.heritagesciencejournal.com/content/2/1/7 Table3DescriptiveofidentifiedlayersofM.27(ARI-9), Analysisofthegravegoods fromthesurfacetothecoating Nine samples from a variety of black (n=2) and red Layer Color Thickness(μm) (n=7) objects from different sites as Morro 1 (Mo-1), 1 White-yellow 20 Colon 10, Chinchorro 4 (CHIN-4) and Camarones 15 (CA15-C and CA15-D) were analyzed (Table 1). Black 2 Darkbrown 170 was less frequently present on the objects, and found in 3 Lightbrown 200 the interior of shells (Figure 6a and 6b). Red was found 4 Lightbrown 340 as a raw material inside leather pouches, or as a raw 5 Layernotveryattached,difficulttoanalyse - material (Figure 6c and 6d). It was also identified inside mortars and shells that were used as containers, or currently visible; but this layer is very fragile and comes possibly for mixing. offeasily. Under SEM-EDX, the black color is primarily indica- Sample M.19 confirms that the surface coatings with tive of the presence of manganese moreover, other ele- paint might have been applied directly on the skin, as ments were found such as Si, Ca, Al and Na. One usually observedamong the red mummies. Nevertheless, distinctive element identified was Fe, which was found the layer is very fragile and most of it seems to have in Sample 1 (M.1) at CA15-C. A difference was found in come off already. Sample M.35 shows that, despite the Sample M.17 from Mo-1, which doesn’t contain Fe but change in the mummification style from black to red, present P. Another difference observed was the identifi- the technological tradition that produced the black- cation of a manganese compound in the form of 2 to colored paint is indistinguishable between these two 3 μm needles embedded in layers of clay (Table 1). forms of mummification. These are pictorial mixtures μDRX analysis determined that these elements corres- that contain the same minerals and share the same pro- pond to cryptomelane, along with quartz, calcite and a duction process, and they seem to have remained the variety of salts (anhydrite and halite, among others). The same over time despite the fact that the black color, two samples correspond to paint adhered to the walls of which initially was applied over the entire body, eventu- a shell used as a container, and thus they are understood allywaslimited tothehead alone. to be components of different compounds; however, some of the quartz and salts may correspond to sedi- mentsfrom theplacewheretheobjectwasfound. Analysisofthestuffing In regard to the red color, SEM-EDX analyses indicate The only stuffing (M.18) analyzed in this study is green- the presence of iron above all. In certain cases other ele- colored sand. Observation of the sample displays the ments such as silica, aluminum, potassium, magnesium, presence of grains of different colors and sizes, despite sulfur and calcium can also be observed. In all cases our characterization of the grains as “green.” SEM-EDX except CHIN-4, there was a minor amount of arsenic, analysis demonstrates the presence of different elements an element naturally occurring in watercourses in the such as sulfur, chlorine, copper, calcium and potassium region [31]. More interesting still is the presence of iron (Figure 5). μDRX analysis indicates the presence of with different morphologies. A series of samples (M1 quartz, possibly sand and atacamite. This identification from MASMA, M11 from MASMA-1 and M14 from is particularly important, as the use of copper mineral CA15-D) were identified as containing iron in large has not previously been identified in the region for spheres approximately 10 μm in diameter and ovoid Archaic period. We cannot rule out that it may have spheres of 1 μm (Figure 7). Other samples (M8 from been extracted from mines unknown to date, as copper CA15-C,M10from COLÓN 10 andM16from CHIN-4) comes from specific places and this mineral is not displayed iron in 1 μm spheres. Lastly, in a single case presentinalltheAtacamaDesert.Itmayalsohavecome (M13 from CA15-C) angular grains 2 to 7 μm long and from exchanges with people having access to mines lo- spheres 1 μm in diameter were also observed. Through cated in what is now southern Peru, or those located μDRX analysis it was determined that in all cases the 300 km South of our area of study, where metallurgical mineral in question was hematite, combined with dif- productions indicate that copper mines were operating ferent types of other minerals: halite, calcite and gyp- in the 15th and 16th centuries [28]. While the identifica- sum (Figure 8). Other peaks of diffractogram were tion of copper minerals on Chinchorro mummies has not identified. been recognized previously [29], the analytical identifica- tion presented herein is something new, as it demon- Discussion strates the existence of the search for copper-based Results obtained from the analysis of the coatings minerals during the Archaic period, much earlier than applied to the mummies indicate that the mummies what haspreviously beenthought[30]. were painted several times, as we observed different Sepúlvedaetal.HeritageScience2014,2:7 Page8of12 http://www.heritagesciencejournal.com/content/2/1/7 Figure5SpectrumofM.19bySEM-EDX. Figure6Samplefromdifferentgravegoods.a)Camarones15C;b)Morro1;c)Camarones15C;d)Masma. Sepúlvedaetal.HeritageScience2014,2:7 Page9of12 http://www.heritagesciencejournal.com/content/2/1/7 Figure7ImagebySEM. painting layers. This empirical evidence could be inter- among the living, or whether subsequent repainting was preted in one of two ways: either the mummified bodies part of the funerary trajectory that required exhumation. were not buried until they were fully materialized and We think that constant action of exhumation would several layers of paints were applied or they were buried have affected the condition of these fragile effigies, andexhumedandpaintedfromtimetotime.Theuncer- which is not the situation. Whatever the case, the action tainty arises from the difficulty in determining whether of repainting reaffirms the special attention that was paid the repainting occurred at the time the individual was tothesematerializedbodies,andthissameattitudeisalso mummified and while the body remained for some time observedintheintentionalrepairofothermummies.The Figure8DiffractogramofM.1fromMasma1. Sepúlvedaetal.HeritageScience2014,2:7 Page10of12 http://www.heritagesciencejournal.com/content/2/1/7 identification of the raw materials sheds light on the of color, from its procurement to application and even Chinchorro people’s knowledge of how to procure and preservation. But we still do not know why, in the entire prepare these materials.Despite the changes in the use of Andeanarea,onlytheChinchorropeopleemployedsuch black color on the mummified bodies, from all the body color technology for mummifying their dead or to in Black Mummies to head only in Red Mummies, the materialize bodies during the archaic period. To date, mixturesusedremainedforalongtime. thereisalsonoconsensusabouttheoriginof Chinchorro Pigment analysis of the grave goods identified minerals mummification practices or the origin of these coastal similar to those specified on the mummified bodies, peoplesthemselves[8,32]. hematite and cryptomelane which had not be identified In terms of funerary rites, the analysis confirmed that until now as previous studies had been only elemental the mummified bodies were not buried or abandoned [13-15]. As the hematite of the coast are different from once the mummification process was complete. On the those found in the highlands of this region [31], these contrary, this long and laborious effort was part of a may have come from different supply sites. Nevertheless, more prolonged ritual use and consumption, which theidentificationofcryptomelaneisparticularlyinterest- would have involved the deterioration of the outermost ing, as the sources of this mineral are found above 4500 layer of the coating. In social terms, the ‘materialization meters of altitude [22]. This finding allows us to formu- of death’ through the mummification of the bodies late interesting hypotheses about how the archaic hunt- shows a degree of specific knowledge and careful, time- ing and fishing people of the coast may have journeyed consuming preparation indicative of the presence of to the highlands themselves or traded with highland “mummification experts”. The practice seems to have hunter-gatherers to obtain this material [22]. But we developed primarily in the Late Archaic period when the need to know more about the other hunter gatherer coastal population increased, perhaps in association with tradition development of archaic period to understand climate changes and increasing water ressources that theirpossibleexchange.Pigmentanalysisopenthepossi- occurred at the end of the Middle Archaic [1,4]. Also, bility to study economic side of this hunter gatherer Chinchorro artificial mummification was triggered by from thisAndeanregion. arsenic poisoning and grief [7]. Arseniasis causes prema- From a technological perspective, the results obtained turebirth, stillbirth,spontaneousabortion andhighperi- confirm that the coastal Chinchorro people had specific natal dead, so Chinchorro parents facing high mortality knowledge of pictorial recipes. The fact that pigments rate of their babies (fetuses and newborns) started to and clays were combined to make the paint that was preserve the tiny bodies to assuage their emotional grief. then used to coat the mummies also points to specific Mummification was a synergy between environmental knowledge of the raw materials used, particularly the constraints,grief, ideology andtheafterlife. plastic or additive qualities of the clays and the coloring Moreover, after the beginning of mummification strength of the manganese and iron minerals. Effectively, process, the identification and use of different pigments theChinchorroartisansselectedandcarefullymixedthese may point to incipient social distinctions among the raw materials to obtain the desired colors, knowing how Chinchorro. Reviewing the chronology of the different much of each material to use, how finely to grind them, types of mummification shows that one kind did not and how to combine them. The results also suggest that replace another, but rather, the black mummies were they knew the source of the raw materials used in their somewhat contemporary with the red ones, and the red preparations, some of which were located hundreds of ones with the natural mummies, although in general the kilometers away, as for example for cryptomelane. black ones tend to be older than the red ones [33]. Lim- In productive terms, the objects analyzed show the iting the application of black color to the heads of the pigments in stages prior to its application on the bodies. mummies, which occurred later, may have been related Once the colored raw material was obtained, it was no to the availability of the raw material, in this case manga- doubt ground up and then stored as a powder or mixed nese, although the same may be said about the different with other compounds. The powder present inside the forms of hematite and cooper minerals. Nevertheless, leather pouch corresponds to such a preliminary stage, changes in the Chinchorro’s mummification techniques while the lumps found correspond to another way of canbealsorelatedtovariationinsocialdimensionofthis storing the pigment—mixed with clay compounds. The group,specificallywithideologicaland/orsymbolicsphere pigment or mixture could then be mixed again with withaestheticconnotations.Thismaymeanthataccessto water, for example, to liquefy it into a paint that could anduseofcertainpigmentswasrestrictedtocertainindi- then be applied as a coating on the mummified bodies. viduals, including fetus and children, especially if these Thus, the objects and the bodies analyzed display differ- minerals were difficult to obtain. Therefore, it is possible ent stages of the same color production sequence. These thattheknowledgeofcolortechnologyandofhowtopre- results tell us something of the Chinchorro’s knowledge pare the mummified bodies was in the hands of experts