61 OBSIDIAN DISTRIBUTION AND PROVENIENCE IN THE CENTRAL HIGHLANDS AND COAST OF PERU DURING THE PRECERAMIC PERIOD Richard L. Burger and Frank Asaro Introduction High quality obsidian combines the qualities of chipping predictability and the sharpest known edge of any material. These features were known to the ancient inhabitants of the Old and New World who sought to gain direct access to obsidian or to acquire it in- directly through exchange. However, obsidian suitable for the production of projectile points and other fine edged tools, occurs only in rare instances of volcanic activity when magma of the requisite silica content and acidity cools rapidly enough to prevent the formation of crystals, but not so rapidly as to trap the volcanic gasses. The prime quality obsidian, born of the convergence of these complex factors of igneous tectonics, begins to decline in value as soon as its genesis is completed, due to its absorbtion of water and consequent devetrification. Thus the kind of obsidian valued for tools is often scarce in regions with a history of vulcanism and it is absent elsewhere. Other raw materials, such as basalt, chert, or quartzite, are more commonly encountered in the geological record and provide a convenient alternative for the toolmaker. Moreover, these other materials are harder and less brittle than obsidian. For most ancient communities, the procurement of obsidian represents an extra expenditure of time and energy over and above minimum subsistence needs. Thus it is not surprising to find that obsidian use in most areas fluctuates throughout prehistory along with changes in political, social, and economic organization. This article will discuss the patterns of obsidian exploitation in the Central Highlands and adjacent coast of Peru during the time period of approximately 10, 000 B. C. to 2, 000 B. C., that period preceding the Introduction of ceramics. The domestication of the major Andean plants and animals and their spread into new ecological zones occurred during this time period. The diffusion of domesticated crops from valley to valley, and from the highlands to the coa st implies extra-local communication and movement at a time when populations were small and societies were relatively simple. It has been hypothesized (Lynch 1971; 1973) that the hunting and gathering regimen of pre-agricultural highland populations included transhumance, a system of subsistence scheduling by which groups move from one ecological zone to another in accordance with seasonal climatic fluctuations and concurrent availability of resources. Such a model would partially explain the mechanism by which long distance and Inter-zonal contacts were first established. However, recent research on the coastal lomas (Quilter, personal communication) and highlands (Rick, personal communication) does not support the model of preceramic transhumance. An alternative or complementary model for early contacts, is that inter- regional and inter-zonal travel occurred to procure raw materials. This article will deal only with obsidian, but other trade materials, such as salt, would create the same 62 need to to beyond local boundaries; unfortunately, many of these other items are harder to detect in the archaeological record, and cannot be traced with the same precision. Obsidian distribution at preceramic sites in central Peru 1 Obsidian was commonly used during preceramic times to produce tools in the central highland area from Juaja-Huancayo to Ayacucho. The work of the Ayacucho Archaeological Botanical Project indicates that obsidian was used in a large number of preceramic sites in all of the defined micro-environments: dry thorn forest, tundra, desert, and humid scrub forest (MacNeish, personal communication). Importation of obsidian and manufacture of obsidian tools was already underway by 10, 000 B. C. and continued throughout the Preceramic. D. Browman, in his survey of archaeological sites in the Jauja-Huancayo area, found that the preceramic population of that area frequently used obsidian for projectile points during the middle and late Preceramic (Browman 1970: 89; MacNeish et al. 1975: 20). The highlands north of the Huancayo-Jauja have recently been studied by the Proyecto de Investigaciones Arqueologicas Junin. A summary of their preliminary findings asserts that obsidian first appears in that area in the late Preceramic and even at this late date it is scarce. Obsidian is far less popular in this area than chalcedony, basalt or quartzite (Matos 1975: 55). Obsidian appeared in small quantities in the late Preceramic layers of Uchcumachay, in excavations directed by P. Kaulicke and R. Matos (Wheeler Pierres-Ferreira et al. 1976). At least one obsidian flake was recovered from levels tentatively cross-dated to the Piki/Jaywa phases of the Ayacucho sequence (Kaulicke, personal communication). Sixty kilometers further north, John Rick excavated at the site of Pachamachay, near Lake Junin. In the preceramic layers he found no obsidian tools, and out of approximately six million flakes, only one was obsidian (Rick, personal communication). This picture of obsidian scarcity in the highlands north of Jauja is supported by the research of Danielle Lavallee in the San Pedro de Cajas area. San Pedro de Cajas is located about 35 kms southeast of Pachamachay. In the excavations of preceramic components at four different sites, none of the artifacts were made from obsidian (Levallee and Julien 1975: 109). Further to the north, in the Huanuco drainage, teams from the University of Tokyo excavated extensive late preceramic deposits of the late Preceramic Mito culture at the sites of Kotosh and Shillacoto. No obsidian artifacts were found among the numerous lithics (Izumi and Sono 1963; Izumi and Terrada 1972; Izumi, Cuwliza, and Kano 1972). In the same area, Ravines discovered an earlier site, Ambo, at which he collected tools stylistically resembling middle Preceramic artifacts from Lauricocha and other sites. The artifacts are made of crypto-crystaline quartz; no mention is made of any obsidian (Ravines 1965). Further west in the puna zone, Cardich investigated preceramic remains at Ranracancha and Lauricocha. Apparently, 63 no obsidian was used at these sites In the Preceramic (Cardich 1964; 1973). The northernmost highland preceramic site for which there is good information is Quichqui Puncu in the Callejon de Huaylas (Lynch 1970). The site was occupied for the first time at approximately 7000 B. C. but abundant early ceramics were found mixed with the Preceramic materials (ibid.: 99). Less than a dozen flakes of obsidian were found there and no obsidian points or other tools were encountered (ibid.: 19). It is not clear whether or not the scarce obsidian predates the introduction of ceramics at Quichqui Puncu. To the west of Huancayo, on the upper slopes of the Cordillera Occidental, T. Patterson and his students surveyed archaeological sites in the Huarochiri area. According to a recent summary, obsidian was used there at the end of the early Preceramic continuing in middle Preceramic and intensifying in the late Preceramic (MacNeish, et al. 1975) Unfortunately, the sample of highland preceramic sites in the current archaeological literature, reaches only 20 kms. south of the city of Ayacucho (Ac351) and no information is available about the early occupations of the neighboring regions to the south, such as Andahuaylas or Abancay. Moreover, we are unaware of detailed published information on preceramic sites in the highlands west of the Ayacucho region (eg. Castrovirreyna). One exception to this vacuum of knowledge are the lithic scatters on the Pampas Galeras to the southwest of Ayacucho (Province of Ayacucho, Department of Ica). These assemblages, rich in obsidian artifacts, appear to be the products of hunting activities which, judging from the style of the projectile points, began during preceramic times (M. Neira, personal communication). A systematic survey of these materials has yet to appear, although several surface collections have been made. Piecing together the fragmentary bits of information ennumerated here, a pattern of central highland obsidian distribution can be inferred and presented as a working hypothesis. The area of extensive obsidian use is the Huancayo/Ayacucho region and adjacent highland areas to the west. The neighboring southern region is not included due to the absence of investigation there. In the more northernly highland areas (northern Junin and beyond), obsidian plays a negligible role, appearing sporadically as an exotic material. Obsidian has also been found at preceramic sites on the coast of Peru, particularly to the west and southwest of the highland region where obsidian was commonly used. At the preceramic shellmound of San Nicolas, located on the bay of the same name in the Province of Nasca, W. Strong discovered abundant obsidian flakes and an obsidian "flake knife" (Strong 1957: 10). While he was there he collected more than 400 waste flakes of obsidian along with obsdian cores and nodules "representing workshop debris. " Later investigations by Vescelius and Engel confirmed the prevalence of obsidian at the site, and revealed two points and a scraper made of obsidian (Vescelius 1963). The presence of cotton at the site suggests that it was probably late Preceramic in date (Bonavia and Ravines 1972). 64 Surveys by Engel and Lanning also discovered obsidian at the preceramic shellmounds of Casavilca, at the mouth of the Ica Valley (Engel 1957: 61-62; Lanning 1960: 50; 1963: 368) and at Mound 12 at Otuma, slightly south of the Paracas Peninsula (Engel 1957b: 57-61; Lanning 1960: 49; 1963: 368). Otuma and Casavilca "are charac- terized by an abundance of small projectile points, neatly pressure flaked and made of obsidian... "1 (Lanning 1967: 72). Obsidian points were also discovered in preceramic graves at Asia, a site located in the lower Asia-Omas Valley. At this site there is evidence that hunting was done with spear-throwers using darts tipped with obsidian, worked bone, wood, or chipped quartz crystal (Lanning 1960: 49; Engel 1963: 56, 99, 111). Otuma, Casavilca, and Asia also appear to date to the late Preceramic (Bonavia and Ravines 1972). In the next drainage, that of Chilca, Engel and his collaborators have conducted large scale excavations at several preceramic sites. Recent work by a team from CIZA and the University of Missoari, directed by R. Benfer and F. Engel, concentrated its efforts at the lomas site of Paloma. A surface area of over 2, 000 m2 was opened by Engel and Benfer, 2 a figure which suggests the extensiveness of the sample. Only four obsidian flakes and one obsidian point were found in the excavations; an additional obsidian point fragment was collected from the surface. Lithics are rare at the Paloma site. The dating of these is not definitively resolved but the Paloma site seems to have been occupied during the middle Preceramic. Two of the Paloma flakes were found in early levels of the site (Quilter, personal communication). Engel also excavated at the nearby preceramic site of Chilca 1, from which he illustrates an obsidian point (1966: fig. 12). In the adjacent Rimac Valley, Engel reports "small leaf shape projectile points of obsidian" from the lowest layers of the Chira/Villa site; these apparently have preceramic associations (Engel 1957: 62-65). Further north on the coast, obsidian is not generally found at preceramic sites. For example, none appeared in the preceramic layers at Ancon (Ravines and Mulle 1972), nor at Aspero (Willey and Corbett 1954: 151). It was not found by Wendt at Rio Seco (Wendt 1976: 31-33). Despite the considerable amount of research at preceramic complexes north of Lima, the only evidence of obsidian is a single unique obsidian projectile point found by P. Ossa at La Cumbre in the Moche Valley. This site contains typologically early artifacts such as Paijan points and a fish tail point, as well as extinct megafauna. It has yielded an early C14 measurement of 8585 + 280 B. P. 4, 892 tools and waste flakes were collected, of which the unique obsidian point constituted . 02%. Most of the tools at La Cumbre were made of fine grained grano-diorite or basalt, and less frequently, chert. The obsidian projectile point is not related typologically to the other artifacts (Ossa, personal communication). Since there is no indication of a later occupation at La Cumbre, the stylistic distinctiveness of the obsidian point is best explained as the result of non- local manufacture. The rarity of obsidian and complete absence of obsidian debitage at 65 La Cumbre are consistent with this hypothesis. The exceptional find of obsidian at La Cumbre can be compared to the isolated but statistically insignificant appearance of obsidian at Pachamachay, and the rare obsidian pieces from Paloma may be comparable in importance to the rare fragments of obsidian found at Uchcumachay, Junin. It is unfortunate that quantitative data are not available on the raw materials of lithics at the coastal sites investigated by Engel and Lanning. The impression given in the publications is that obsidian Is frequently found on the coast south of Chilca. Absence of quantitative data prevents comparison of the intensity of obsidian use in the highlands and coast. Inprecise chronological controls on most coastal sites makes diachronic discussion difficult. The extant evidence for obsidian exploitation on the coast is drawn primarily from late Preceramic occupations, but this may partially reflect the bias which exists in the archaeological literature. 3 The sources of preceramic obsidian Lanning, in 1967, speculated that obsidianfound on the early sites of the south and south central coast had been imported from the southern highlands (1967: 72). R. Ravines later discovered an obsidian flow in the barren heights of San Genaro, Huancavelica (Petersen 1970: 50). This source, located near the snow line, is strategically located, being near the headwaters of the Pisco, Mantaro, and Pampas drainages. These river valleys. provide natural routes of transportation of the obsidian to coastal and highland areas. Ravines speculated that this mine could be the source of the obsidian artifacts encountered on the coast and highlands. Subsequently, the Ayacucho Archaeological Botanical Project located another natural obsidian deposit near Tukumachay, to the south of the city of Ayacucho. Although the samples recovered were of low quality obsidian, the possibility remained that there exists, or existed, a better outcrop of the same geological deposit which could have served as a source for early tool makers of Ayacucho and neighboring regions. The source of obsidian artifacts can be determined by matching the trace element composition of source samples with the same measurements for the artifacts in question. The underlying assumptions of the technique are that 1) each source of obsidian has a unique, and therefore, diagnostic trace element composition; and 2) the trace element composition of a single source is relatively homogeneous. These assump- tions have been rigorously tested and generally confirmed, with the amendation that in exceptional cases, homogeneity does not occur. However, the variation which does occur in the trace element composition of rare sources is patterned, and therefore recognizable, and equally diagnostic of the source (Bowman, Asaro, and Perlman 1973). Moreover, archaeological obsidian is usually more chemically homogeneous than the flow itself because only selected parts of most deposits were mined in antiquity. Since coincidental similarities do exist in the amounts of some trace elements in any two flows, the provenience analysis is most convincing when a large number of trace elements are 66 measured. Since low precision measurements blur the compositional differences between different types of obsidian, a high level of measurement precision is desirable. When low precision techniques are used on a small number of trace elements, the result may be failure to distinguish between different sources. In 1973, Burger began work on the trace element analysis of Andean obsidian using a method of rapid scan X ray fluorescence at the Department of Geology, University of California, Berkeley. This technique had been applied earlier to Mesoamerican and Californian obsidians (eg. Jack and Heizer 1968; Jack and Carmichael 1969). The technique provides semi-quantitative measurements for Sr. Rb, and Zr. This procedure proved inadequate in destinguishing between the different types of Peruvian obsidian (Burger and Asaro, in press). It was therefore decided to apply a technique which provided quantitative measurements for a larger number of trace elements. This goal led to the collaboration of the authors at the Lawrence Radiation Berkeley Laboratory. The first step was to subject the two sets of source samples and a selected group of preceramic obsidian artifacts to a complete neutron activation analysis. This produced quantitative measurements for 26 trace elements. The analytical procedure is described elsewhere (Burger and Asaro, in press). The sample tested from the flow in Ayacucho (845 X) did not match any of the artifacts from preceramic or later archaeological sites. The three samples collected by Ravines from the Quispisisa mine, Huanacvelica (861 H, 861 J, and 861 K) had a homogeneous chemical composition which matched many of the archaeological samples tested from Ayacucho, the South Coast and other areas. It was demonstrated by neturon activation that ten samples from preceramic sites near Ayacucho, three samples from the preceramic site of San Nicolas (Nasca), and six samples from preceramic strata at Uchcumachay (Junin) came from the Quispisisa mine. Many other samples were analyzed by neutron activation, including pieces from sites throughout Peru and northern Bolivia. An attempt was made to sample diachronically as well as geographically for each region (Burger and Asaro, in press). The neutron activation analysis of these samples initially characterized six of the most important sources of obsidian used in ancient Peru. This work permitted us to design a method of X ray fluorescence which could successfully distinguish between these types of obsidian by providing quantitative measure- ments on the amounts of Ba, Ce, Rb, Sr, Zr and other trace elements. A detailed description of this procedure will soon be published (Burger and Asaro, in press). The X ray fluorescence technique has the advantage of being rapid, non-destructive, and inexpensive. Using this method, we analyzed a larger sample of obsidian artifacts in order to identify their sources. Any X ray fluorescence results which did not fit the established groupings were selected for further analysis by neutron activation. Through this process, we isolated a second type of obsidian used at preceramic sites in the Ayacucho area. The chemical composition of this type of obsidian differs from both the Quispisisa source material and the sample taken from the natural outcrop near Tukumachay, 67 Ayacucho. This new type of obsidian must come from an unlocated flow. Since obsidian of this composition did not occur in our 1974/5 sample of over 900 pieces at archaeological sites outside of the Ayacucho area, we hypothesized that the obsidian source is in the Ayacucho region. We christened obsidian with this distinctive trace element composition as the Ayacucho Type obsidian, a tern which will be replaced when the actual source of the obsidian is discovered. More recent research (1977), in which an additional 90 obsidian samples were analyzed by X ray fluorescence, indicates that four of the five fragments tested from the Early Horizon site of Chupas, Ayacucho, are of the Ayacucho Type; the fifth fragment is from the Quispisisa mine. No obsidian artifacts tested from outside the Ayacucho area were of the Ayacucho Type. These new findings are in accord with the original hypothesis concerning the general source area of the Ayacucho Type obsidian. In order to fully characterize the chemical composition of the Ayacucho Type obsidian, five samples were analyzed by neutron activation. These came from the sites of Puente (Ac158), lomachay (Ac102), and Ac500. These results, as well as those for preceramic artifacts and source samples from the Quispisisa mine, are reproduced in Table 1. Also included in this table is the composition of a unique fragment found on the surface of Jaywamachay and that of the nonutilized outcrop near Tukumachay. The compositional differences between the Ayacucho Type obsidian and the obsidian from the Quispisisa source which are detected by X ray fluroescence are seen in Table 2. The provenience results of both the neutron activation and X ray fluroescence analysis for the preceramic artifacts tested are summarized in Tables 3 and 4. The two types of obsidian just discussed constitute more than 99% of the materials of the preceramic artifacts tested. However, the single piece of obsidian re- covered by Rick at Pachamachay in his preceramic layers is of a unique chemical compo- sition. A second fragment, collected from the surface of Jaywamachay, has a different- and equally rare composition. This sample may date to the Initial Period/Early Horizon occupation of the site. The occasional appearance of unique obsidians when large samples are tested is an interesting phenomenon which has received little attention in the literature. Such anomolies may be produced in the contact zones between the obsidian and the *adjacent non-igneous geologic formations. Alternatively, they may come from igneous formations with small patches or inclusions of obsidian. Such flows might provide material for an occasional tool but not a reliable source for long-term tool production. A third alternative is that such rarities will prove to be long distance imports from unsampled regions. In the case of Pachamachay or Uchcumachay, the first two hypotheses are considered more probable than the latter. Discussion It has been shown that obsidian mined at two different natural deposits was used In central Peru during preceramic times and that the most popular source of obsidian was the Quispisisa quarry near Castrovirreyna, Huancavelica. This quarry 68 was already being exploited by 10, 000 B. C. and its obsidian transported to sites in Ayacucho, approximately 110 kms. to the east. Inportation of this obsidian apparently extended to highland areas in the northeast, more than 200 kms. from the Quispisisa quarry. The evidence for obsidian use in these regions has already been described. Although we have been unable to analyze preceramic artifacts from the Juaja-Huancyo area, obsidian from later occupations there have been shown to come from the Quispisisa source (Burger and Asaro, in press). The movement of Quispisisa obsidian to distant areas included the south and south central coast. Forty-nine samples tested from San Nicolas all were shown to be Quispisisa obsidian. The Quispisisa mine is located 270 kms. by air from San Nicolas. Also identified as Quispisisa obsidian were artifacts from Jaywamachay, Puente, lomachay, Tukmachay, Ac500, Ac300, Pikimachay, Pampas Galeras, Uchcumachay, and La Cumbre (see Table 3). Three alternative models can be presented to describe the mechanism by which Quispisisa obsidian reached preceramic archaeological sites outside the Castro- virreyna area. The first model is one of direct exploitation in which representatives of distant communities journey to the source area, mine the raw material themselves, and then return to their communities. A second model would be the existence of mobile non-specialized traders, possibly local pastoralists from near the mine, who would mine and transport the obsidian to the areas which utilized it. A third model would postulate multiple small exchanges from group to group as part of a larger exchange system, so that by the time the obsidian reached a distant site like La Cumbre, it would have passed through the hands of many communities. All three of these models have the advantage of not requiring complex economic organizations absent among small scale societies of hunter-gatherers, littoral collectors and fishermen, or incipient agriculturalists. Markets, ports of trade, self-sufficiency through archipellagos may have all played a role in trade in later times, but they are difficult to reconcile with the present under- standing of the preceramic societies of Peru. The first model, that of direct acquisition, could only function if there was unrestricted access to the source of obsidian. The neutrality and multi-ethnic exploita- tion of essential raw materials is a pattern wh-ich occurs in the Central Andes before and after the arrival of the Spaniards. For example, the salt deposits of the jungle and highlands were the destinations of long distance journeys by numerous outside groups (Conchas Contreras 1975: 74-76; Oberem 1974: 350). These groups arrived, camped, mined the salt, and then returned to their homes. Such trips are often so long that they imply some sort of trade and social contact during the journey, if only to procure food and drink and to avoid hostilities. Contact would have occurred even if these other groups do not play an active role in gaining the obsidian. For example, a trip to San Nicolas from the Quispisisa mine crosses numerous ecological boundaries and must have crossed many cultural boundaries as well. The trip to the mine and back must have taken over a month, if we use a figure of 15 to 20 kms per day as the average distance covered by a llama caravan (Flores 1968: 130). This is probably a very conservative estimate of trip length because it is based on distance by air rather than the actual distance than must be traversed along the winding paths of the convoluted 69 central highlands. The trip from Chavin to Olleros, Ancash, for example, measures only 28 air kilometers, which would be a trip of two days using Flores' figure. But even the most fit resident cannot make the journey by foot in less than 14 hours; it would be even slower with cargo animals or humans carrying a heavy load. Thus, the time to reach the Quispisisa mine and to return to San Nicolas might actually take almost two months, especially if one includes several days for the mining activities. However, unlimited access to resources, even during the Preceramic, cannot be presumed. Recently Moseley has argued that jural rights over coastal resources began on the central Peruvian coast during the Preceramic (1975: 51-52). The second model, that of long distance trading expeditions, is appealing since it is derived from the widespread tradition of lengthy trips by puna dwellers into the lower agricultural valleys of the sierra and coast in order to obtain products which cannot be produced within the puna environment. Such journeys are scheduled to coincide with the dry season when pastrues are reduced and the temperatures are the coldest on the puna. This same type of pattern may have existed even preceding the domestication of camelids since herds of wild camelids may have migrated naturally to lower zones in search of pasture. If this was the case, this model might apply to the early as well as the middle and late Preceramic. In the second model the direction of movement would have been reversed, as compared with the first model. However, the overall result would have been quite similar: contact between peoples of different ecological zones and cultures, and the exchange of goods. Both systems would be capable of providing dependable supplies of quantities of obsidian. Either model would have been adequate to supply the inhabitants of San Nicolas, Otuma, or Jaywamachay with the large quantities of obsidian that they needed for their tools. The third model, one of multiple exchanges, would have had a different impact since only people In adjacent communities would have been brought into contact. This model best accounts for the rare occurrence of obsidian in distant areas such as the North Coast. It may also apply to nearer areas, on the outer edge of the obsidian utilization sphere, such as northern Junin (Uchcumachay) or the area around Lima (Paloma). Ayacucho Type obsidian has only been found at preceramic sites in the Ayacucho area, where it made up 12% of the preceramic artifacts tested. The earliest concrete evidence of its use is from Puente (Ac158), from a "zone" with an estimated age of 7100 B. C. Obsidian of this type was was found in later preceramic layers at Puente, Ac500, and lomachay (Ac102). Its use apparently continued throughout the preceramic, Early Horizon, and possibly Early Intermediate Period. A relatively local source for this obsidian was hypothesized because of its apparent absence from archaeological sites elsewhere. Such an explanation, however, does not account for the greater popularity of the Quispisisa source obsidian at preceramic sites in Ayacucho (87% of the sample). Perhaps this unlocated source will be found in some of the yet unsampled and somewhat distant areas of Ayacucho, such as the Provinces of Cangallo or Victor Fajardo. Other 70 considerations might include differences in the quality of the two obsidians or more limited access to the Ayacucho Type source as the result of jural restrictions. If the Ayacucho Type source is local or semi-local, it could have been distributed by either direct exploitation or by exchange. During the Peruvian Preceramic, society underwent major social and economic transformations, and one would expect that the use and distribution of obsidian would reflect these radical changes. It may be, for example, that small quantities of obsidian reached the coast in early preceramic times through multiple exchanges, but its use intensified in late Preceramic times, at which time it was procured through mining expeditions or through trade with the members of pastoral caravans. The study of such diachronic shifts must await future archaeological investigation. Conclusions 1. Long distance distribution of obsidian in the central highlands of Peru predates the advent of agriculture, and was underway by 10, 000 B. C. 2. Direct contact between the coast and highlands is domonstrated by the presence of obsidian from the Quispisisa mine, over 4,,600 meters above sea level, at preceramic sites on the south and south central coast. This situation existed by the late Preceramic (approximately 2500 B. C.) and probably earlier. 3. Journeys of one to two months were made during preceramic times in order to distribute Quispisisa obsidian to the coast and distant highland areas. 4. At least two obsidian flows were already being exploited in central Peru during the early Preceramic. One is the Quispisisa source near Castrovirreyna, Huancavelica and the other is an unlocated deposit probably in the Ayacucho region. 5. Beginning in the early Preceramic, small amounts of obsidian reached distant areas, up to 900 kms. away from the original source. This probably occurred through multiple exchanges. Acknowledgements: Special thanks are due to Julie R. Jones, Rogger Ravines, Richard S. MacNeish, Paul Ossa, John H. Rowe, and Robert Vierra. This work was made possible by financial support provided by the Atomic Energy Commission, and the University of California Archaeological Research Facility. We wish to thank Mrs. Helen V. Michel for her assistance in the neutron activation measurements and Mr. Duane Mosier for maintaining the necessary electronic equipment. We are grateful to Mr. Lim Tek and the staff of the University of California Berkeley Triga Research Reactor for the nuclear irradiations used in this work. The development of the measurement system was funded by the US Department of Energy. 71 Footnotes 1. This article is confined to the Preceramic in central Peru because of the paucity of data for other regions. Work by M. Neira and others shows that obsidian is abundant at preceramic sites in Arequipa. Preliminary XRF results on samples from Sumbay, Arequipa and elsewhere indicate that neither Quispisisa nor Ayacucho Type obsidians were utilized. 2. Some of this work was begun by Engel and C. I. Z. A. before the collaboration of R. Benfer and the University of Missouri team. 3. At present, there is no adequate chronological framework for dealing with the Preceramic Period on a pan-Peruvian basis. The terms used in this paper are broad chronological divisions based on C14 measurements and the artifactual inventories of the sites. The coarseness of the divisions are appropriate to the vagaries of the current ,diachronic controls prior to the introduction of ceramics, and for this reason were preferred over the system utilized by Lanning and others. The terms used will be employed as follows: approx. absolute dates Lanning (1967) early Preceramic ? to 6000 B. C. Preceramic I to m middle Preceramic 6000 B. C. to 2500 B. C. Preceramic IV to V 2500 B. C. to 1800 B. C. Preceramic VI late Preceramic 72 Bibliography Bonavia, D. and R. Ravines 1972 El preceramico andino: evaluacion y problems. Revista del Museo Nacional XXXVIII: 23-60. Lima. Bowman, H. E., F. Asaro, and I. Perlman 1973 Composition variations in obsidian sources and the archaeological implications. Archaeometry 15(1): 123-137. Browman, D. 1970 Early Peruvian Peasants: the Culture History of a Central Highlands Valley. Ph. D. dissertation, Department of Anthropology, Harvard University. Burger, R. and F. Asaro in press Analisis de los artefactos de obsidiana de los Andes Centrales a traves de sus elementos trazos; neuvas perspectivas sobre la la interaccion economica prehispanica en Peru y Bolivia. Revista del Museo Nacional XLHI, Lima. Cardich, A. 1959/ Ranracancha: un sltio prehistorico en el Departamento de Pasco, 1960 Peru. Acta Prehistorica III/IV, Centro Argentino de Estudios Prehistoricos, pp. 35-48. Buenos.Aires. 1964/ Lauricocha. Fundamentos para una Prehistoria de los Andes 1966 Centrales. Studia Prehistorica VII/X, Centro Argentino de Estudios Prehistoricos. Buenos Aires. Conchas Contreras, J. 1975 Relacion entre pastores y agricultores, Allpanchis Phuturinga vIII 67-101. Lima. Engel, F. 1957 Early Sites in the Pisco Valley of Peru.. American Antiquity 23: 34-45. 1963 A preceramic settlement on the Central Coast: Asia, Unit 1. Transactions of the American Philosophical Society 53 (3): 3-139. Philadelphia. 1966 Geografia Humana Prehistorica y Agricultura Precolombina de la Quebrada de Chilca 1. Lima. 73 Flores Ochoa, J. A. 1968 Izumi, S. , Los Pastores de Paratla: una introduccion a su estudio. Antropologia Social 10. Instituto Indigenista Interamericano, Mexico. P. J. Cuwliza, and C. Kano 1972 Excavations at Shillacoto, Huanuco, Peru. Tokyo University Museum, University of Tokyo Press, Tokyo. Izumi, S. and T. Sono 1963 Andes 2: Excavations at Kotosh, Peru, 1960. Kodokawa Publishing Co., Tokyo. Izumi, S. and K. Terrada 1972 Andes 4: Excavations at Kotosh, Peru, 1963 and 1966. University of Tokyo Press. Jack, R. N. Jack, R.N. and I. S. E. Carmichael 1969 The chemical "fingerprinting" of acid volcanic rocks. California Journal of Mines and Geology, Short Contributions 100: 17-31. and R. F. Heizer 1-968 Fingerprinting of some mesoamerican obsidian artifacts. Contributions of the University of California Archaeological Research Facility 5: 81-100. Lanning, E. P. 1960 1963 1967 Lavalle, D. and M. 1975 Lynch, T. F. 1970 Chronological and cultural relationships of early pottery styles in ancient Peru. Ph.D. dissertation. Department of Anthropology, University of California, Berkeley. A pre-agricultural occupation on the central coast of Peru. American Antiquity 28: 360-371. Peru Before the Incas. Prentice Hall, Englewood Cliffs, N. J. Julien. El habitat prehistorico en la zona de San Pedro de Cajas, Junin. Revista del Museo Nacional XI: 81-119. Lima. Excavations at Quishqui Puncu in the Callejon de Huaylas, Peru. Occasional Papers of the Idaho State University Museum No. 26. Pocatello. Preceramic Transhumance in the Callejon de Huaylas, Peru. 1971 American Antiquity 36: 139-148. Harvest Timing, Transhumance, and the Process of Domestication. American Anthropologist 75 (5): 1254-1259. MacNeish, R. S., A. Nelken-Terner, and A. G. Cook 1970 Second Annual Report of the Ayacucho Archaeological-Botanical Project. Robert S. Peabody Foundation for Archaeology, Andover. MacNeish, R.S., T.C. Patterson and D.L. Browman 1975 The Central Peruvian Prehistoric Interaction Sphere. Phillips Academy, Andover. Matos Mendieta, R. 1975 Moseley, M. E. 1975 Muelle, J. C. y R. Ra 1973 Oberem, U. 1974 Petersen, G. 1970 Ravines, R.. 1965 Strong, W.D. 1957 Vescelius, G.S. 1963 Prehistoria y ecologia humana en las punas de Junin. Revista del Museo Nacional XLI: 37-74. Lima. The Maritime Foundations of Andean Civilization. Cummings Publishing Co., Menlo Park. vines Los estratos preceramicos de Ancon. Revista del Museo Nacional XXXIX: 49-70. Lima. Trade and trade goods in the Ecuadorian montainia. In: Native American Indians, edited by P. Lyon, pp. 346-357. Little, Brown, and Co., Boston. Mineria y metalurgia en el antiguo Peru. Arqueologicas 12. Museo Nacional de Antropologia y Arqueologia, Lima. Ambo: a new preceramic site in Peru. American Antiquity 31: 104-105. Paracas, Nazca, and Tiahuanacoid Cultural Relationships in South Coastal Peru. Society for American Archaeology Memoir 13. Salt Lake City. New finds at San Nicolas. Nawpa Pacha 1: 43-46. Berkeley. 74 1973 75 Wendt, W.E. 1976 El asentamiento preceramico en Rio Seco, Peru. Lecturas en Arqueologia 3, translated by P. Kaulicke. Universidad Nacional Mayor de San Marcos. Lima. Wheeler Pierres-Ferreira, J., E. Pierres-Ferriera, and P. Kaulicke 1976 Preceramic animal utilization In the Central Peruvian Andes. Science 194: 483-490. Willey, G. R. and J. 1954 Corbett Early Ancon and Supe Cultures. Columbia University Studies in Archaeology and Ethnology 3. New York. 76 Key to Figure 1 1. La Cumbre 2. Shillacoto 3. Kotosh 4. Ambo 5. Lauricocha 6. Ranracancha 7. Rio Seco 8. Aspero 9. Pachamachay 10. San Pedro de Cajas 11. Ushoumachay 12. Ancon 13. Huarochiri 14. Chiravilla 15. Huancayo preceramic sites 16. Chilca 1 17. Paloma 18. Asia 19. Quispisisa obsidian source 20. Ac 100 (Pikimachay); Ac 102 (lomachay) 21. Ac 158 (Puente) 22. Ac 500, Ac 351 (Tukumachay) 23. Otuma 24. Pampas Galeras 25. Casavilca 26. San Nicolas 77 ECUADOR BRAZIL Obsidian source Obsidian artifacts abundant Obsidian artifacts rare No obsidian artifacts km 0I200 O 200 XBL 7711-11310 Figure 1. Map of Peru indicating archaeological sites and geological deposits discussed in this article. -~o IV C 0 'V <6' 0 78 Table 1. Results of Neutron Activation Analysis Element Abundances in Representative Peruvian Obsidians (Values in ppm or percent if indicated by % after chemical symbol) Al% Dy Mn Na% OBSIDIAN FROM QUISPISISA SOURCE Sr SOURCE MATERIAL FROM QUISPISISA, HUANCAVELICA 1 861 H BURG-66 5.86 +/- .14 1.59 +/- .09 363 */- 3 2.973 +/- .018 3.75 +/- .23 95 4/- 67 2 861 J BUPG-67 1.01 +/- .17 1.77 */- *09 367 +/- 3 2.951 4/- .018 4.06 4/- .24 115 */- 69 3 861 K BURG-68 o.66 +/- .a8 1.83 4/- .08 361 +/- 3 2.874 */- .018 4.07 4/- .23 112 +/- 70 ARTIFACTS FROM SAN NICOLAS 4 822 G BURG-3 6.95 4/- 5 822 N SURG-8 6.99 4/- 6 822 W BURG-18 6.90 4/- ARTIFACTS FROM JAYWAMACHAY 7 861 R BURG-74 6.96 4/- a 871 J BURG-92 6.86 .1- 9 890 0 SUR-129 1.06 4/- 10 890 P SUR-130 1.08 4/- 11 893 F 8UR-108 6.97 4/- 12 893 G BUR-109 6.80 4/- 13 893 H BUR-110 6.45 4/- ARTIFACTS FROM UCHCUMACHAY 14 861 W 15 871 ? 16 890 E 17 890 G 15 890 U 1 9 893 0 8URG-79 BUR-106 BUR-120 8UR-123 BUR-135 BUR-1lS 6.89 +/- 6.94 +/- 6.97 4/- 6.88 +/- 6.71 4/- 6.93 +/- .12 .12 .11 .18 .20 .16 .20 .14 .16 .21 .10 .21 .1Z .16 .17 .14 ARTIFACT FROM IOMACHAY 20 890 Q BUR-131 6.73 +/- .08 ARTIFACT FROM PUENTE 21 890 R SUR-132 6.67 4/- .10 1.64 4/- .09 1.52 */- .10 1. 80 /- .09 1.69 4/- .08 1.69 0/- 09 1.70 */- .09 1. 82 4/- .10 1.64 4/- .08 1.68 4/- .09 1.63 +/- .09 1. 59 4/- .08 1.71 4/- .09 1.77 4/- .08 1.67 */- .08 1.59 l/- .10 1.55 4/- .09 1.68 +/- .08 1.66 */- .09 370 +/- 5 3.021 +/- .029 3.65 4/- .28 265 +/- 126 373 +/- 5 3.051 +/- .032 3.97 +/- .32 138 4/- 109 369 +/- 4 2.952 4/- .031 4.26 4/- .37 132 */- 95 366 */- 359 */- 364 +/- 364 +/- 360 +/- 364 +/- 358 * /- 370 +/- 358 +/- 365 +/- 363 4/- 363 4/- 365 4/- 3 2 3 3 2 3 3 3 3 2 2 3 3 363 */- 364 */- 2.897 +/- .018 3.000 4/- *018 2.652 +/- .017 2.859 +/- .018 2.709 +/- .017 3.011 */- .018 2.905 4/- .018 2.974 4/- .018 3.005 4/- .018 3.037 4/- .018 3.031 4/- .018 2.992 4/- .018 2.990 4/- 9018 3.88 +/- 3.78 */- 4.45 4/- 4.32 */- 4. 01 +/- 4.00 4/- 3.67 +/- 3.57 +/- 3.66 +/- 3.67 4/- 3.64 +/- 4.14 4/- 3;77 4/- .23 .22 .23 .23 .22 .22 .21 .22 .22 .22 .22 .22 .22 130 149 102 9 96 110 239 4/ 4/ 4/ 4/ 141 +/- 140 4/- 56 +/- 110 +/- 111 +/- 95 +/- 62 68 72 75 68 71 74 72 72 63 66 73 72 2.577 +/- .016 4.76 4/- .23 160 */- 65 2.919 +/- .018 3.88 +/- .24 107 4/- 73 ARTIFACTS FROM AYACUCHO SiTE AC300 22 890 X BUR-138 6.56 4/- .12 1.73 4/- .10 23 890 1 BUR-140 6*90 +/- *17 1.64 */- .09 361 4/- 3 3.002 */- .019 363 +/- 3 2.344 4/- .015 3.87 */- .25 214 +/- 82 5.06 +/- .22 237 4/- 72 AYACUCHO TYPE OBSIDIAN ARTIFACTS FROM IOMACHAY 24 893 J 8UR-111 7.2 25 893 K BUR-112 6.8 26 893 M BUR-113 6.8 !0 ./- .37 1.48 4/- .10 1 +/- .08 1.37 +/- .08 12 ./- .10 1.57 +/- .09 449 4/- 3 443 */- 3 452 +/- 3 2.980 4/- .018 2.878 4/- .018 3.033 */- .018 4.51 4/- .23 111 +/- 70 4.19 */- .23 109 4/- 73 ARTIFACT FROM AYACUCHO SITE Ac 500 27 893 N BUR-114 6.97 #/- .12 1.34 4/- .09 ARTIFACT FROM PUENTE 28 893 E BUR-lO7 o.65 #/- .12 1.52 #/- .09 453 +/- 3 3.414 4/- .020 509 +/- 3 3.294 4/- .020 3.78 4/- .23 38 */- 76 3.64 4/- .24 196 4/- 77 RARE TYPE 2 OBSIDIAN ARTIFACT FROM JAYWAMACHAY 29 828 V BURG-38 7.10 +/- .13 1.59 +/- .10 508 +/- 7 3.109 4/- .023 3.64 4/- .23 0 +/- 75 OBSIDIAN FROM AYACUCHO SOURCE (NEAR TUKUMACHAY) 393 */- 3 2.165 */- .014 4.47 +/- .22 156 4/- 95 30 045 X SLAG-60 6.60 */- - 1 2 1. 27 +/- .1 0 As *U Ba Sm La Co OBSIDIAN FROM QUISPISISA SOURCE SOURCE MATERIAL FROM QUISPISISA, HUANCAVELICA 1 861 H BURG-66 13-7 +/- 1.8 8.376 ./- .102 2 861 J BURG-61 15.8 +/- 1.8 8.374 4/- .102 3 861 K 8URG-68 13.9 ./- 1.8 8.269 */- .101 689 #/- 22 2.382 4/- .015 26.88 4/- .92 .55 4/- .06 706 4/- 22 2.401 4/- .015 24.34 4/- .91 .53 +/- .06 675 4/- 22 2.402 4/- .015 26.21 4/- .92 .47 */- .06 ARTIFACTS FROM SAN NICOLAS 4 822 G OURG-3 15.1 *i- 1.1 5 822 N BJRG-8 14. S/- 1.1 6 822 W BURG-18 14.9 */- 1.1 ARTIFACTS FROM JAYWAMACHAY 7 861 R 8URG-74 14.2 4/- 1.9 8 871 J SURG-92 11.3 4/- 1.7 9 890 0 BUR-129 L1.8 4/- 1.8 10 890 P WUR-130 16.7 4/- 1.8 11 893 F BUR-t0S 15.2 4/- 1.4 12 893 G BUR-109 16.0 4/- 1.4 l; Rol " 311R-110 IS.6 4/- 1.4 ARTIFACTS FROM UCHCUMACHAY 14 861 W 8URG-79 15.4 4/- 2.0 15 871 z BUR-106 1 .7 4/- 2.0 16 890 E BUR-120 14.5 4/- 1.7 17 890 G WR-123 17.0 */- 1.8 18 890 U BUR-135 17.8 4/- 1.7 19 893 0 BUt-115 1S.6 /- 1.4 ARTIFACTS FROM IOMACHAY 20 890 Q BUR-131 14.7 ?/- 1.6 ARTIFACT FROM PUENTE 21 890 R BUR-132 d.6 4/- 6.0 8.518 ./- .053 8.374 4/- .052 8.627 +/- .053 8.642 */- .111 8.879 #/- .072 8.789 4/- .074 8.731 4/- .074 8.723 4/- .062 8.648 +/- . 062 3.642 4/- .062 8.627 4/- .111 8.753 4/- .069 8.701 4/- .073 8.741 +/- .073 8.772 4/- .071 8.729 +/- .063 8.797 +/- .071 8.649 +/- .101 711 4/- 21 2.482 +/- .009 27.59 +/- .63 731 4/- 21 2.*440 +/- .009 26.89 ./- .62 736 4/- 20 2.476 +/- .009 27.82 4/- .64 749 +/- 748 4/- 747 +/- 738 +/- 737 4/- T39 4/- 750 +/- 696 4/- 749 4/- 710 +/- 730 */- 761 */- 774 +/- 24 19 22 23 21 21 21 23 18 22 23 22 20 2.422 4/- . 016 2.520 4/- .013 2.543 */- .013 2.532 4/- .013 2.479 4/- .010 2.*485 /- .011 2.477 4/- .011 2.420 4/- .016 2.522 4/- .013 2.S45 */- .013 2.513 4/- .013 2.524 ./- .013 2.484 */- .011 27.47 */- .98 28.05 4/- .80 26.62 4/- .81 29.46 4/- .85 26.*49 4/- * 70 27.89 +/- .72 28.23 4/- .71 26.33 4/- *99 28.79 4/- .82 28.42 */- .82 27.66 4/- .83 27.48 4/- .80 26.44 +/- .71 .61 4/- .05 .46 4/- .05 .48 4/- .05 .40 4/- .53 */- .56 4/- .54 +/- .66 4/- .65 4/- .54 4/- .50 4/- .54 +/- .69 4/- .45 +/- .47 4/- .50 */- .06 .05 .06 .06 .05 .05 .05 .06 .05 .06 .06 .06 .05 748 */- 21 2.525 +/- .012 27.10 4/- . 79 .60 */- .06 686 */- 19 2.528 4/- .025 26.75 ./-1.57 .56 4/- .07 ARTIFACTS FROM AYACUCHO SITE AC300 22 890 X 8UR-138 1.2 */- 1.9 8.956 +/- .077 23 890 Z 8UR-140 14.4 +/- 1.6 8.884 ./- .072 735 4/- 23 2.585 4/- .015 28.63 */- .91 743 */- 21 2.546 4/- .013 27.76 4/- .78 AYACUCHO TYPE OBSIDIAN ARTIFACTS FROM IOMACHAY 24 893 J 3UR-111 4.7 1/- 1.1 4.604 */- .042 25 893 K SUR-112 3.0 4/- 1.1 4.663 *J- .042 26 893 M BUR-113 3.1 +/- 1.1 4.705 +/- .042 ARTIFACT FROM AYACUCHO SITE Ac 500 27 893 N BUR-114 3.0 +/- 1.2 4.720 +/- .043 ARTIFACT FROM PUENTE 28 893 E BUR-10 1.3 4/- 1.1 4.595 4/- .043 231 */- 16 1.975 +/- .009 21.50 4/- .64 .20 4/- .04 259 4/- 16 1.956 */- .009 22.54 */- .64 .43 */- .04 227 4/- 16 1.979 +/- * 009 23.16 */- .66 .28 4/- .04 212 ./- 1T 1.996 */- .009 23.01 4/- .68 .26 4/- .04 208 4/- 17 2.217 4/- .009 18.82 */- .64 .48 +/- .05 RARE TYPE 2 OBSIDIAN ARTIFACT FROM JAYWAMACHAY 29 828 V 8URG-38 3.2 */- 1.1 4.537 +/- .039 312 4/- 14 2.341 4/- .009 21.69 4/- .63 OBSIDIAN FROM AYACUCHO SOURCE (NEAR TUKUMACHAY) 4.500 +/- .038 288 */- 14 1.955 +/- .008 27.93 4/- .61 79 .49 ./- .06 .55 +/- .06 .19 */- .03 30 845 X WURG-0 3.1 +/- 1.0 .16 */- .03 Sc Fe% Cs OBSIDIAN FROM QUISPISISA SOURCE SOURCE MATERIAL FROM QUISPISISA, HUANCAVELICA 1 861 H BURG-66 1.68 */- .02 2 861 J BURG-67 1.47 +/- *02 3 861 K BURG-68 1.44 +/- .02 ARTIFACTS FROM SAN NICOLAS 4 822 G BURG-3 1.52 +/- .02 5 822 M BURG-S 1.*49 +/- .02 6 822 W BURG-IS 1.50 +/- .02 ARTIFACTS FROM JAYWAMACHAY 7 861 R BURG-74 1. 51 4/- 8 871 J BURG-92 1.52 .1- 9 890 0 BUR-129 i.52 4/- 10 890 P BUR-130 1.52 4/- 11 893 F BUR-108 1.55 4/- 12 893 G BUR-109 1.55 /- 13 893 H BUR-lLO 1.53 4/- ARTIFACTS FROM UCHCUMACHAY 14 861 W BURG-79 1.48 4/- 15 871 Z BUR-106 1.51 4/- 16 890 E BUR-120 1.52 4/- 17 890 G SUR-123 1.51 4/- 18 890 U SUR-135 1.53 4/- 19 893 0 BUR-115 1.50 4/- .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 ARTIFACTS FROM IOMACHAY 20 890 0 BUR-131 1.*47 +/- 02 ARTIFACT FROM PUENTE 21 890 R BUR-132 1.49 */- .02 ARTIFACTS FROM AYACUCHO SITE AC300 22 890 X SUR-138 1.52 +/- .02 23 890 z BUR-140 1.SO */- .02 .55 +/- .02 11.55 ./- .31 .55 4/- .02 11.57 +/- .31 .55 */- .02 11.64 +/- .31 .60 */- .01 11.51 4/- .24 .57 +/- .01 11.27 +/- .23 .57 +/- .01 11.30 +/- *23 .61 */- .02 .55 +/- .02 .55 */- .02 .58 +/- .02 .57 +/- .01 .59 +/- .01 .56 +/- .01 .59 +/- .02 .62 +/- * 02 .57 4/- .02 .60 +/- .02 .56 +/- .01 .55 +/- .01 l1.36 +/- .32 10.96 +/- .23 11.65 +/- .30 11.74 +/- .30 11.53 +/- .25 11.24 +/- :25 11.42 +/- .25 11.46 4/- .32 11.47 +/- .24 11.61 +/- .29 11.51 +/- .29 11.33 +/- .27 11.51 +/- .25 .58 +/- .01 11.52 +/- .28 .56 +/- .02 11.07 +/- .27 .56 +/- .02 11.29 +/- .28 .56 */- .01 1I. 28 +/- .27 1.25 */- .14 1.37 +/- .15 1.38 4/- . 15 1.49 4/- .12 1.24 4/- .11 1.20 4/- .10 1.16 ./- .14 1.45 /- . 15 1.29 */- .12 1.22 4/- . 12 1.24 I/- 11 1.30 o/- .12 1.35 L/- .12 1.31 */- .15 1.38 4/- .13 1.35 4/- .12 1.39 4/- *13 1.15 L /- .11 1.32 4/- .12 1.19 4/- *11 1.34 4/- . 12 1.30 4/- .13 1.22 I/- 11 20.93 +/- Ole 21.05 4/- .18 20.96 +/- .18 21.08 4/- .13 20.87 4/- .13 20.82 L/- .12 21.02 4/- .19 21.26 4/- .18 19.98 4/- .13 20.08 4/- .13 20.96 o/- .14 20.74 4/- .14 20.88 4/- .14 20.96 4/- .19 20.89 */- .17 19.86 4/- .13 19.93 ./- .13 20.00 4/- .13 20.86 4/- .14 19.94 4/- .13 19.97 L/- .13 19.91 +/- .13 20.00 4/- .13 AYACUCHO TYPE OBSIDIAN ARTIFACTS FROM IOMACHAY 24 893 J BUR-1l1 1.71 +/- .02 25 893 K SUfR-112 1.72 +/- .02 26 893 M BUR-113 1.73 +/- .02 ARTIFACT FROM AYACUCHO SITE Ac 500 27 893 N BUR-114 1.72 +/- .02 ARTIFACT FROM PUENTE 28 893 E BUR-107 2.03 +/- .02 .51 +/- .01 4.07 +/- .12 .48 +/- .01 3.86 +/- .12 .49 +/- .01 4.00 +/- .12 .48 +/- *01 4*01 +/- *12 .49 4/- .01 3.89 +/- .13 .25 +/- .05 16.20 +/- .11 .32 +/- .05 16.06 +/- .11 .27 ./- .05 16.24 +/- .11 .24 +/- .05 16.62 +/- .12 .28 +/- .05 15.97 +/- .11 RARE TYPE 2 OBSIDIAN ARTIFACT FROM JAYWAMACHAY 29 828 v BURG-38 1.87 +/- .02 .54 +/- .02 3.80 +/- *11 .24 4/- .05 15.44 +/- .17 OBSIDIAN FROM AYACUCHO SOURCE (NEAR TUKUMACHAY) 30 845 X SURG-60 L.56 +/- .02 80 Sb Th o48 +f- UL 4o 89 +/- o 13 027 +/- AS l6oOO +/- .11 Ce Hf OBSIDIAN FROM QUISPISISA SOURCE SOURCE MATERIAL FROM QUISPISISA, HUANCAVELICA 1 861 H BURG-66 2 861 J BURG-67 3 861 K BURG-68 .422 4/- .009 .420 4/- .009 .415 +/- .009 52.*48 0/- .81 51.13 +/- .80 50.77 +/- .80 3.39 +/- .08 3.33 +/- .08 3.31 4/- .08 1.104 +/-.008 1.101 +/-.008 1.098 */-.008 1.129 +/- .028 1.168 4/- .029 1.121 */- .028 ARTIFACTS FROM SAN NICOLAS 4 822 G BURG-3 .421 4/- .007 5 822 N BURG-8 .424 4/. .007 6 822 W BURG-18 .438 4/- .007 ARTIFACTS FROM JAYWAMACHAY 7 861 R SURG-74 .440 /- .009 8 871 J BURG-92 .442 4/- .008 9 890 0 BUR-129 .427 4/- . 009 10 890 P BUR-130 .431 /- .009 11 893 F BUR-108 .440 4/- .008 12 893 G BUR-109 .416 4/- .008 13 893 H BUR-110 .426 */- .008 ARTIFACTS FROM UCHCUMACHAY 14 861 W BURG-79 .431 */- .009 15 871 Z BUR-106 .440 /- .007 16 890-E BUR-120 .434 4/- .009 17 890 G BUR-123 .426 */- .009 18 890 U BUR-135 .445 4/ .009 19 893 0 BUR-115 .431 4/- .008 S1.73 +/- .60 52.32 4/- .61 51.94 +/- .61 52.23 4/- .83 51.06 4/- .81 52.99 4/- .74 53.26 +4- .74 50.96 4/- .69 51.05 4/- .68 51.89 +/- .69 52.01 +/- .83 53.46 4/- .83 52.78 4/- .73 52.33 */- .73 51.53 +/- .71 52.03 4/- .69 ARTIFACTS FROM IOMACHAY 20 890 0 BUR-131 .415 +/- .009 50.78 +/- .71 ARTIFACT FROM PUENTE 21 890 R BUR-132 .428 +/- .009 51.83 +/- .76 ARTIFACTS FROM AYACUCHO SITE AC300 22 890 X BUR-138 .417 4/- .010 52.92 4/- .81 23 890 Z BUR-140 .436 +/- .009 51.58 4/- .72 3.25 +/- .06 3-39 */- .06 3.22 4/- .06 3.31 4/- 3.27 4/- 3.27 4/- 3.32 4/- 3.19 +/- 3.30 +/- 3.24 +/- 3.23 +/- 3.31 */- 3.20 +/- 3.34 +/- 3.20 +/- 3.24 +/- .08 .08 .08 .08 .07 .07 .07 .09 .08 .08 .08 .08 .07 3.19 4/- .08 3.36 4/- 08 3.28 4/- .09 3.28 4/- . 08 1.144 +/-.006 1.134 4/-.006 1.103 4/-.006 1.111 4/-.008 1.124 +/-.007 1.114 +/-.007 1.122 +/-.007 1.113 +/-.006 1.104 +/-.006 1.116 +/-.006 1.112 +/-.008 1.135 */-.007 1.117 4/-.007 1.119 */-.007 1.098 +/-.007 1.119 4/-.306 1.106 */-.007 1.106 +/-.007 1.121 +/-.008 1.114 +/-.007 1.187 */- .021 1.156 4/- .021 1.169 +/- .021 1.096 4/- .029 1.144 +/- .030 1.149 +/- .025 1.154 +/- .025 1.141 +/- .024 1.156 */- .024 1.140 4/- .024 1.213 1.192 1.170 1.123 1.118 1.160 +/- .031 +/- .029 +/- .025 ,/- .025 +/- .025 +/- .024 1.154 +/- .025 1.162 +/- .027 1.135 4/- .028 1.176 +/- .025 AYACUCHO TYPE OBSIDIAN ARTIFACTS FROM IOMACHAY 24 893 J BUR-111 .287 +/- .006 25 893 K 8UR-112 .291 4/- .006 26 893 N BUR-113 .300 +/- .006 ARTIFACT FROM AYACUCHO SITE Ac 500 27 893 N BUR-114 .288 */- .006 40.49 +/- .59 40.73 +/- .59 40.97 +/- .60 40.70 4/- .59 3.79 +/- . 07 3.84 4/- .07 3.67 +/- .07 3.76 +/- .07 ARTIFACT FROM PUENTE 28 893 E BUR-107 .332 ./- .007 38.96 +/- .58 3.85 4/- .07 1.951 +/-.009 .968 +/- .023 RARE TYPE 2 OBSIDIAN 29 828 V BURG-38 .344 +/- .006 42.92 +/- .82 3.86 +/- . 09 1.936 +/-.011 .953 +/- .032 OBSIDIAN FROM AYACUCHO SOURCE (NEAR TUKUMACHAY) 30 845 X BURG-60 .307 +/- .006 46.25 4/- .60 81 Eu Ta Yb 1.770 */-.008 1.781 4/-.008 1.780 +/-.008 1.791 4/-.008 .847 4/- .022 .876 +/- .022 .826 */- .022 .815 +/- .023 .744 */- .020 3 o 71 +/- *07 1.641 +/--007 82 Table 2. Selected trace element composition of Andean obsidian: x-ray fluorescence analysis made at Lawrence Berkeley Laboratory (in parts-per-million) obsidian type Ba Ce Rb Sr Zr Quispisisa Ayacucho Rare 9 (Pachamachay) 735 220 844 52 43 195 138 150 139 56 171 99 85 60 Table 3. Summary of source identifications of preceramic obsidian artifacts analyzed by neutron activation (NAA) and x-ray fluorescence (XRF) Archaeological Site Quispisisa Source Ayacucho Type Rare Method San Nicolas, Dept. Ica Ac335, Dept. Ayacucho (Jaywamachay) Ac158, Dept. Ayacucho (Puente) Ac351, Dept. Ayacucho (Tukumachay) Ac500, Dept. Ayacucho Ac300, Dept. Ayacucho AclOO, Dept. Ayacucho (Pikimachay) Pampas Galeras, Dept. Ica Uchcumachay, Dept. Junin Pachamachay, Dept. Junin La Cumbre, Dept. La Libertad 30 9 3 4 10 1 1 NAA 8 XRF 23 3 2 NAA2 XERF 10 XRF 3 NAA 1 XRF 6 NAA 2 XRF 10 XRF 1 XRF 5 5 8 NAA 6 XRF 2 1 XRF 1 1R%F 1 49 NAA 3 XRF 46 1 83 Table 4. Analysis of obsidian from the excavations of the Ayacucho Archaeological- Botanical Project Sites Sampled AclOO AclO2 Ac158 Ac300 Ac335 Ac351 Ac5OO (Pikimachay) (Iomachay) (Puente) Microenvironmental Zone* dry thorn forest dry thorn forest desert tundra humid scrub forest tundra humid scrub forest (Jaywamachay) (Tukumachay) Site Zone Phase * Tentative dates* Quispisisa Source Ayacucho Type AclOO f-2 Ac102 8 7 6 Ac158 XIII XII XI VIII VII VI V I AC300 C-north C-south Ac33 5 M-N K J-2 H G F E D C Ac351 C-2 C-1 Ac300 F E D-1 * data from MacNeish et al. 1970; and MacNeish, personal communication. 1 1 2 3 1 1 Puente Puente Jaywa Chihua Puente Puente Jaywa Piki Piki Piki Piki Piki Chihua Jaywa Chihua Ayacucho Huanta Puente Puente Puente Jaywa Jaywa Jaywa Jaywa Cachi Cachi Puente Piki Chihua 7000 BC 7300 BC 5600 BC 3600 BC 7100 BC 6900 BC 5900 BC 5300 BC 5200 BC 4900 BC 4720 BC 4520 BC 4000 BC 5500 BC 3100 BC 10000 BC 9020 BC 7940 BC 7610 BC 7030 BC 6850 BC 6550 BC 6490 BC 6410 BC 6300 BC 2500 BC 1800 BC 7300 BC 4600 BC 3000 BC 2 1 1 2 1 1 1 6 4 1 3 3 2 2 2 5 1 5 4 2 1 2 1 1 1 2 - -.F --- - - j ic- -