AMISTAD Cultural Resources Study

THE PREHISTORY OF THE LOWER PECOS RIVER REGION

Summarized by Joseph H. Labadie
from Solveig Turpin's
An Overview of the Prehistory in the Lower Pecos River Region

September 1993

INTRODUCTION

The Lower Pecos River region has one of the best documented records of continuous human occupation in the New World. The area surrounding the confluences of the Pecos and Devils Rivers with the Rio Grande also claims a diverse and complex body of Native American rock art, painted on or carved into the hard limestone by successive different ethnic groups. The boundaries of the region are most often defined by the extent of the distinctive Pecos River style pictographs, and by a commonality of items found in dry rockshelter deposits. In terms of modern political divisions, the region includes Val Verde County, and adjacent areas of Terrell, Crockett, and Sutton Counties in Texas.

HABITAT

One measure of civilization is the degree to which humankind devises buffers between itself and the environment. In modern society, technology moderates the effect of climate; creates artificial environments through light, heat, and air conditioning; and accelerates food production to the point where the everyday citizen is remote from nature. Hunters and gatherers, such as those who inhabited the Lower Pecos throughout prehistory, present the opposite extreme. Their relationship to their habitat was more intimate. Rather than drastically modifying their physical world, the people adapted to the available resources, and much of this adaptation was through cultural means. Thus, it seems only reasonable to assume that environment played a significant role in influencing the course of their cultural development.

The natural landscape of the Lower Pecos region is rolling hills and rocky flats dissected by drain-ways that become increasingly deep and entrenched as they near the three major rivers—the Pecos, the Devils, and the Rio Grande. To the casual visitor, these barren hillsides and steep canyons seem bleak and inhospitable. However, the semi-desert of today in part results from a century of overgrazing by domestic livestock, which, in turn, accelerated the effects of an increasingly drier climate. The past environment, reconstructed from plant and pollen studies, animal remains, flood sequences, and the debris left by human populations, was one of deeper soils, more effective moisture, and more plentiful native foodstuffs. Within the living memory of longtime ranchers, erosion has claimed the sediment cover; floods have eradicated stands of oak, walnut, and pecan trees; and thorny shrubs, unpalatable to livestock, have expanded their hold on the remnant rocky soil. In historic times, since weather conditions have been recorded, the climate has vacillated, alternating between wet and dry years and decades. Similar cycles—some of which lasted for centuries—affected the everyday life of the aboriginal people, and the development of their culture.

More than 12,000 years of environmental history can be reconstructed from the plant parts, animal remains, and flood sequences preserved in Lower Pecos archeological sites. At the end of the last Ice Age, the region was a parkland savannah, composed of grasslands dotted with trees and bushes. This habitat in turn supported herds of large game animals, such as bison, camel, horse, and elephant (Dibble and Lorrain 1968; Bement 19861. By 9,000 years ago, dramatic changes in world climate forced the Lower Pecos people to adapt to an increasingly arid environment through the exploitation of desert plants and animals. This drying trend probably experienced a number of ups and downs over the next 6,000 years. The most dramatic change was an interlude of moister climate 3,000 years ago, which permitted grasslands to re-colonize the region (Dibble and Lorrain 1968; Bement 1986). A return to aridity forced the retreat of the prairie, and the herd animals it supported. Semi-desert conditions prevailed until some time shortly before European contact and colonization, when cooler, wetter weather transformed the American West into the "sea of grass" known in frontier history. Extreme droughts are recorded in northern Mexico between 1640 and 1645, and in the Lower Pecos in the 1880s, but short-term climatic fluctuations such as these are usually not discernible in the archeological record.

PREHISTORIC LIFEWAYS IN THE LOWER PECOS REGION

The long, continuous occupation of the Lower Pecos River Region has been reconstructed from the information accumulated over 60 years of formal archeological study (Labadie 1991). Native American occupation of the region may well have begun before 10,000 B.C., and it continued in varying forms until the end of the 19th century—comparatively late in historic times. A considerable amount is known about the technology and diet of the prehistoric hunters and gatherers of the Lower Pecos because the refuse from hunting and gathering activities forms a significant portion of the rockshelter deposits. Factors that motivated the selection of certain localities for open-air camp sites or a preference for one rockshelter above another can be inferred from the distribution of known sites across the landscape. The unwritten rules that made for an orderly society are only assumable by analogy to living or recorded hunting and gathering people who lived in similar environments. The notable rock art affords a view into the esthetic or ideological world of the native people, but the exact translation of their symbols into our frame of reference is probably beyond our reach. Although the art conveys considerable detail about prehistoric lifeways, it is doubtful if modern man can ever completely understand the supernatural world of the aboriginal mind.

Other questions about Lower Pecos prehistory are also no longer answerable. The tribal identity of the aboriginal populations, and what they called themselves are unknown. The common name given them in the popular press—"Pecos Man"—simply refers to their geographic location. Based on similarities in lifeways and material culture, the Lower Pecos Archaic people are often classified as Coahuiltecans, implying only that they probably belonged to the same language group as other hunter gatherers encountered in southern Texas and northeastern Mexico at the time of first European contact. Not until the influx during historic times of Plains Indians, such as Apaches and Comanches, are familiar Native American names mentioned in historic documents.

Similarly, although four major prehistoric pictograph styles have been defined and several minor styles proposed, it is difficult to attribute any of them to a specific group. The styles can be arranged in chronological order, beginning with the elaborate, polychrome Pecos River style, which is generally thought to be between 3,000 and 4,000 years old. The miniature Red Linear pictographs are probably more recent, but their span may overlap with the older style. The two latest prehistoric styles—the Red Monochrome and the Bold Line Geometrics—are believed to be Late Prehistoric in age, dating sometime after A.D. 600 and before A.D. 1600. Some historic pictographs, painted in a style characteristic of the Plains Indians, can be more precisely dated, but many miscellaneous panels cannot be assigned to a defined style, or to even the broadest time range.

For convenience, the long span of prehistory has been divided into four broad time periods. Each is characterized by a variety of traits, ranging from economics to art styles, but projectile point styles are often considered the hallmark of the period. The age of the more common Lower Pecos point styles has been determined by radiocarbon dating (Turpin 1991a), so they can be considered time markers for the sites where they are found. The beginning and ending dates assigned to the various time periods are arbitrary in the sense that the transition from one stage to the next was often slow and gradual.

The Paleolndian occupation (before 7000 B.C.) is set in the habitat as it was at the end of the Pleistocene (Ice Age), when large game animals, now extinct, still roamed the Lower Pecos region. As the climate became more arid, the human populations adapted to a semi-desert environment, developing the cultural traits characteristic of the Archaic Period (7000 B.C. to A.D. 600). Sometime around A.D. 600 to A.D. 1000, changes such as the adoption of the bow and arrow signaled the advent of the Late Prehistoric Period (A.D. 600 to 1600). The coming of the first Europeans at the close of the 16th century marks the end of prehistory, and the beginning of the Historic Period. By 1880, the Native American presence in the Lower Pecos region had been virtually eliminated.

THE PALEOINDIANS: THE EARLIEST PEOPLE

Tentative evidence for the region's earliest human occupation is in the form of scattered and burned bones—fragments of now-extinct animals such as camel, horse, and elephant, found buried at two sites: Cueva Quebrada (41VV162A), and Bonfire Shelter (41VV126A). Both sites were first excavated as part of the National Park Service program of study. No formal tools made of flint or bone, and no features, such as hearths or pits, have been detected with these skeletal fragments. The burning of the bones at Cueva Quebrada (Lundelius 1984) and the cut marks and breakage pattern on specimens at Bonfire Shelter suggest that the animals were butchered and brought into these shelters by man. Radiocarbon assays of charcoal intermingled with the bones of extinct animals date these episodes to the range from 12,000 to 14,000 years ago.

More definitive is a second bone deposit at Bonfire Shelter (Dibble 1970; Dibble and Lorrain 1968). This bone bed is composed of the remains of an estimated 120 bison of the extinct species Bison antiquus, who were stampeded over the cliff, falling to their death on the rocks below. The animals who tumbled over the cliff struck the massive roof fall below, and rolled back into the shelter, where they could be butchered under the protecting overhang. Layering within the mass of bone cast aside after the meat was stripped suggests that at least three separate drives or stampedes over the cliff above took place within a relatively short period of time. Considering that this extinct form of bison was about one-third again as large as the modern species, this method of hunting was certainly effective. Butchering of the carcasses and the recovery of stone tools demonstrate that the drives were the work of humans and not accidents of nature. Among the stone artifacts are the distinctive Folsom and Plainview dart points—time markers of the Paleolndian Period. Radiocarbon dating of charcoal from small hearths in this level places these events at about 10,000 years ago (Bement 1986). Lesser amounts of bone from these extinct bison were recovered from the lowest level of Arenosa Shelter (Dibble 1967) and the intermediate zone at Cueva Quebrada, both approximately the same age as the Bonfire bone bed.

Although Bonfire Shelter is the only mass animal kill site known this far south of the Southern Plains, some information on the lifeways of the Paleolndian in the Lower Pecos can be gathered from this 10,000-year-old bone level. The economy obviously relied to some degree upon the hunting of big game, although undoubtedly this highly mobile food supply was supplemented by smaller game and plant foods. Human groups whose economy relies on following migratory herds generally develop cooperative hunting techniques. Although a bison stampede may be easy to start, it is difficult to control unless a pre-arranged plan is followed and each hunter plays his role correctly. Based on the assumption that hunting is primarily a male skill, handed down from father to son, the big-game hunter is seen as a member of a small band, related to the other members through the male line of descent. The question of whether this pattern can be applied to the Lower Pecos Paleolndian is as difficult to answer as the second question, that of whether these people were permanent residents of the region or merely transitory passersby. It is often thought that the scarcity of sites from this time period indicates a sparse population, or one that passed through the region, leaving few remains at their temporary camp sites. Nomadic hunters, constantly on the move in pursuit of game, would leave few perceptible traces behind. In addition, the great length of time that has passed since these people lived has to be taken into account. The environmental changes discussed earlier would have worked to eliminate much of the evidence for their occupation of the region. Finding sites of this stage in prehistory may often be more an accident of preservation than a reflection of the intensity of use. However, the big game hunting strategy and the carefully crafted projectile points are similar to those found across great expanses of North America during this time period. Thus, the Lower Pecos economy may have been part of a widespread tradition of exploiting the soon-to-be-extinct large game animals of the last Ice Age.

THE ARCHAIC PERIOD: CLASSIC FORAGERS IN A CHANGING ENVIRONMENT

Introduction

By about 9,000 years ago, a different way of life began to be reflected in cultural deposits left by the Lower Pecos people. Although the characteristic tools, such as Golondrina and Angostura dart points, are still called Late Paleolndian by many, the term "Pre-Archaic" has been suggested by some researchers as a more appropriate description of the economic adaptation to the changing environment. Reconstruction of past environments, derived from the analyses of pollen and other plant remains, animal bones, flood-deposits, and human excrement, indicates that a warming, drying trend began at the end of the Pleistocene, and continued through time. Interrupted only cyclically by wetter periods, the increasing aridity altered the natural resources, and influenced the lifeways of the Lower Pecos inhabitants. At present, it is difficult to say whether the local population adapted to changing conditions or whether they were replaced by people conditioned to the increasing aridity.

Projectile points characteristic of this transitional time, such as Golondrina, are similar to Paleolndian types, but they are found in contexts little different from those of the later Archaic Period. Sites such as the Devils Mouth (Johnson 1961, 1964; Sorrow 1968) and Baker Cave provide radiocarbon assays and tool types indicative of a 9,000-year-old occupation, but the dietary emphasis is on small game and vegetal foods. One hearth excavated in Baker Cave (Word and Douglas 1970; Chadderdon 1983; Brown 1991), and dated to this time period by both radiocarbon assays and Golondrina projectile points, contained fragments of 12 species of mammals; 23 species of reptiles; six species of fish; and 16 different types of plants, 11 of which are edible. This reliance on the widely varied resources of a semi-desert environment was to characterize the subsistence pattern of the Lower Pecos throughout the remainder of prehistory.

The long span of the Archaic Period is often described as a stable adaptation to the varied and variable resources of a semi-desert environment. Stability, when discussed in the archeological sense, takes on a very broad meaning. Many of the short-term intermittent stresses that affected everyday life are simply not discernible in the physical remains left after thousands of years. Additionally, much of the interpretation of past lifeways must be based on the more permanent artifacts—those that survive in the refuse dumps, burials, and living areas of prehistoric people. The abstract concepts that governed attitudes toward such social practices as marriage, inheritance, line of descent, and religion can only be inferred from analogy to living or recorded groups. Thus, stability in the Lower Pecos Archaic is a generality most appropriately applied to subsistence—the basis of human life; and technology—the means by which subsistence is guaranteed.

The 7,600 years of the Archaic Period are divided into three sub-periods: Early (7000 to 4500 B.C.); Middle (4500 to 1000 B.C.); and Late (1000 B.C. to A.D. 600). (These sub-periods are discussed individually, in more detail, later in this appendix.) The material culture and basic economy remained very much the same throughout the Archaic Period, but significant changes are apparent in other arenas. Two of the defined pictograph styles—the Pecos River and the Red Linear—are of Archaic age. Both are products of a culture economically based in hunting and gathering, but the two styles are markedly different in form and content. Much of the interpretation of preliterate art depends upon placing it in the context of the culture that produced it—in this case, perhaps two different ethnic groups that occupied the same territory, but at different times in the long time span we call the Archaic Period.

Stable Elements in the Archaic Life Style

Housing was provided by the numerous rockshelters common in this region. Generations of fires in many shelters produced the deep, ashy deposits; massive accumulations of burned rock; and blackened ceilings. The discarding of burned rock and other garbage by throwing it out of the shelter mouth resulted in the talus cone—or V-shaped rock slides—which are so visible a trademark of an inhabited shelter. Some evidence suggests that partitions may have been erected to screen parts of the shelter, but domestic life must have been largely communal. The grass-lined pits found in some sites may have been sleeping nests, perhaps made to protect infants from the all-pervasive dust and dirt. The dead were often buried in the shelter deposits left by the living.

Open-air sites representing the Archaic Period are burned rock accumulations or middens on the upland flats and gentler slopes of the canyons. These mounds of burned limestone are the debris from earth ovens, where desert plants were baked. A second type of Archaic upland site is hearth fields, where numerous smaller, sparser rock hearths fan out across the caliche flats. Hearth fields usually contain much more debris from stone tool manufacture and many more utilitarian implements than the burned rock mound sites, suggesting they were open living areas where many varied activities took place. No remains of structures or temporary shelters have been found on these sites, but some sort of rudimentary hut or lean-to may have been constructed of perishable building materials that would not survive in these exposed locations.

By far the majority of information on Archaic lifeways comes from the dry shelter deposits where normally biodegradable materials were preserved. Extensive study of seeds, leaves, pollen, bone, and the contents of human excrement reveals a most varied and eclectic diet. The staple plant foods were pricklypear, lechuguilla, and sotol, augmented by seasonally available nuts, seeds, fruit, and flowers. Although deer and other large mammals were probably prized, the bulk of the meat was supplied by rodents, rabbits, snakes, lizards, and other small animals. Birds, fish, turtles, and insects were eaten. It is noteworthy that many of the protein sources were trapped, netted, or scavenged rather than hunted.

The spear-thrower—atlatl—was the most powerful weapon, used to add force to propelling a dart or lance, this object was not replaced by the bow and arrow until Late Prehistoric times, after A.D. 600. The presence of atlatls in both Pecos River and Red Linear art panels is the major basis for their broad assignment to the Archaic Period.

The large chipped stone projectile points, or dart points, vary in style throughout the Archaic, but their function remained much the same. Other stone tools included scrapers and gouges used in hide and plant processing. Knives, naturally serrated by the flaking technique, were employed in butchering, wood cutting, and plant harvesting. Drills for perforating bones, shell, stone, and leather were finely flaked from flint. The sharp edges of waste flakes—debris from the manufacture of formal tools—were often expediently used for the task at hand. Smooth, fist sized rocks—manos—served as grinding implements to pound nuts, tubers, and seeds into a type of flour, and to powder naturally colored pebbles into pigment for paints. Mortars ranged from the enlargement of natural concavities in the bedrock to slabs of more portable limestone—metates. Often the re-use of bedrock mortars deepened these pits beyond the reach of the normal person. It has been suggested that these extremely deep holes may have been used to store water and food, or to ferment beverages.

Although Pecos River style art seems to show elaborate costumes, only a few items of clothing have been identified in shelter deposits. Rabbit skin robes, woven vegetable fiber loin coverings, fringed skirts, and sandals made of wooden frames padded with fiber make up the entire known wardrobe (Schuetz 1956, 1961, 1963). Plaited or twilled baskets, nets, matting, and cordage were manufactured from fibrous plants (McGregor 1992). Pouches made of split pricklypear may have been carrying containers. No water bags or flasks have been recovered, but it is possible that containers were made of pricklypear pads, the internal organs of animals, leather, or tightly woven baskets. Wooden implements include the atlatl, fending or rabbit sticks, snares and traps, digging sticks and fire tongs, drills, and hearths. A wooden mortar and pestle, cached in a rockshelter near Pandale (Collins and Hester 1968), had been used most recently to grind pricklypear seeds; some were stuck in cracks in the mortar. A second similar mortar from near Dryden was radiocarbon-dated to about A.D. 900. Both of these specimens were of piñon pine—a tree no longer common to the Lower Pecos—suggesting that they may have been hidden because of their rarity and value. Bone and antler were fashioned into ornamental beads, pressure flakers for stone tool manufacture, awls for weaving or sewing, and other tools. Perforated mussel and snail shell, bone, wood and stone beads, deer hoof and mussel shell rattles, and long strings of rattlesnake vertebrae may have been worn as jewelry.

The pictographs—the most spectacular of Archaic-age art forms—are not the only esthetic expressions found in the shelters. Painted pebbles and clay figurines—technically called "art mobileur," or portable art—were produced over a long time span. Smooth, water-worn rocks, usually painted on both sides with black pigment, bear designs ranging from abstract geometrics to reproductions of the human face and body. There are some indications that styles changed over time, but few specimens have been found in contexts that could be accurately dated (Parsons 1986). The most interesting proposition is that painted pebbles were amulets or charms, possibly used by the women as ritual objects (Mock 1987).

Clay figurines are rarer (Shafer 1975), but these simple lumps of pinched or shaped, unfired clay have been found at several sites. Some of the fragments have conical appendages—perhaps breasts—and only those female figures have incised or punctated designs. The object seems to have been to reproduce the human body; rarely is the head shown. These figurines have been interpreted as ritual paraphernalia, perhaps used in fertility rites, witch craft, or curing ceremonies as a substitute for the human subject.

With the possible exception of pictographs, the material culture of the Archaic Period was simple and expedient. The raw material for the most common items—stone tools—was so abundant that a suitable implement could be easily fashioned, used, and discarded. The wooden and bone tools were not only more perishable, but the raw material was less easily acquired, and required more labor to convert to a suitable form. The mobile lifestyle inhibited the size of a tool kit to items easily carried or difficult to replace. Although simple by our standards, this technological level was highly adaptive, in that it was sufficient to maintain the Archaic population for thousands of years. Changes in tool forms, such as projectile points, are one trait used to differentiate between the Archaic sub-periods.

Sub-periods

The Early Archaic Period, 7000 to 4500 B.C.: Deposits representing this time period have been found at several sites. The dietary pattern that was to characterize the majority of the Archaic stage has been well established by studies of plant parts, animal bone, and human feces recovered from these sites. The most informative caves about the overall lifeways of the Early Archaic are Baker, Hinds (Shafer and Bryant 1977), and Eagle (Ross 1965). A heavy reliance on the desert succulents, such as lechuguilla, sotol, and pricklypear, which are high in bulk and low in nutritional values, was supplemented by all manner of fruits, roots, seeds, and leafy plants. Virtually every animal and reptile was consumed, but desert plants were soon the major dietary component.

Early Archaic-age skeletal remains found below the vertical shaft of a deep sinkhole bear some marks of the types of dietary stress experienced by the human population at this time. Several of the individuals interred in this cavern showed extreme tooth wear; bone pitting caused by iron-deficiency; and arrested growth lines indicative of severe nutritional strain during childhood. However, the overall health of the people was better than that of most early agriculturalists and many other hunting and gathering populations.

In the Lower Pecos region, Early Archaic dart point styles are generally lumped under the broad descriptive names "Early Barbed," "Early Stemmed," or "Early Corner Notched." Names borrowed from other regions, such as Martindale, Uvalde, or Gower, are sometimes used; and new type names, such as Baker and Bandy, have been proposed. Few tools, other than the projectile points popular during this time period, set the Early Archaic apart from the general Archaic technology. If the lithic technology remained constant, it can be presumed that many of the more perishable items found in Middle and Late Archaic contexts were also in use in earlier times.

At present, the only features recognized in Early Archaic deposits are refuse pits and burned rock accumulations. Indicative that much of the technology that was the mainstay of Archaic life was developed by Early Archaic times were the few painted pebbles; the surviving fragments of baskets, nets, and sandals; rare wooden and bone artifacts; and the stone tools.

The Middle Archaic Period, ca. 4500 to 1000 B.C.: Although cultural deposits of the Middle Archaic Period predominate in the majority of the rockshelters excavated in the last 30 years, it remains a time the material culture of which is well known, while the underlying social mechanisms are poorly understood. About 4500 B.C., the cultural sphere of the Lower Pecos appears to contract. No longer are the characteristic projectile point styles shared with adjacent cultural areas. The distinctive Pandale point—a beveled dart point with a highly recognizable twisted blade—is considered as the horizon marker for the Middle Archaic. The very names given to the most popular dart points of this time period—Langtry and Val Verde—reflect the core area in which they are found. Although a few specimens turn up in other areas of the State, the widespread shared styles of the earlier and later periods stand in marked contrast to the localized distribution of the Middle Archaic tool forms.

Another purely regional phenomenon—the Pecos River Style pictographs (Grieder 1966a, b; Kirkland and Newcomb 1967; Turpin 1990a)—has been dated to the Middle Archaic Period by experimental radiocarbon assay (Russ, Hyman, and Rowe 1992). These multi-colored cave paintings are considered by some people to be one of the earliest religious art styles of the Americas. The central figures are large, faceless, costumed humans, which are called shamans because they give the distinct impression that they are religious practitioners. The figures are armed with a set of objects, such as fending sticks, atlatls (spear throwers), and pricklypear pouches—all found in the dry shelter deposits.

Smaller figures include other humans, animals, and strange geometric forms. The positioning of many of groups of these forms suggests dominance of the central character over other humans, shamans, or game animals. The one animal in these scenes that exudes power akin to that of the shaman is the "panther," or mountain lion, which still lives in the region today. Deer are a favored topic, as are snakes, fish, turtles, and an occasional bird.

The massive amount of over-painting at some sites, such as Panther Cave and Rattlesnake Canyon (see Zintgraff and Turpin 19911, suggests recurring ceremonial events, perhaps carried out during times when the scattered populace came together for harvest celebrations. Ethnohistorically, such aggregations took place when desert fruits, such as pricklypear, ripened. Congregating for such social events gave small groups the opportunity to exchange information and goods, make political alliances, and marry outside their immediate family.

What motivated the painting of these enigmatic figures and how art served a function in their society can only be inferred from living groups. Often wall painting is part of rituals conducted for events such as puberty celebrations or initiation into select societies. The mystical nature of Pecos River style art has given rise to speculation that it originated as an expression of hallucinogenic visions (Campbell 1958) induced by eating mescal beans—a known practice in historic times. The skill with which many of the Pecos River style scenes were drawn suggests that at least a part-time specialist was at work. In a hunting and gathering society, the artist may well have been the shaman, the magician thereby placating the supernatural to ensure the welfare of the group as a whole. There can be little doubt that the Pecos River style of art was drawn by indigenous hunters and gatherers at a very elemental level of social organization, but with a highly varied and imaginative view of the natural and supernatural world.

The consistency of themes and style in Pecos River style art suggests that a unified belief system prevailed over an area extending some 90 miles north and south of the mouth of the Pecos River, including the mountains of Mexico—visible on a clear day from Lake Amistad. The question then becomes why the Middle Archaic people developed such unique characteristics. One avenue of research now being followed by prehistorians is the scattered and tentative evidence for widespread erosion during the early Middle Archaic Period. Both the Devils Mouth site and Arenosa Shelter show marked discontinuities in strata bearing projectile points of this age. Radiocarbon dates from Black Cave (Turpin 1982)—a major Pecos River style pictograph site—show that a lengthy Early Archaic occupation was almost completely washed away sometime in the Middle Archaic. One explanation for such erosion and massive flooding is an episode of extreme aridity—possibly a widespread drought—that seared the uplands, dried the water-holes, reduced plant cover, and forced human and animal populations to concentrate on the major river courses. There, the people may have found new ways of responding to the difficulties imposed by proximity, elaborating their religious ceremonies, and developing their own brand of material culture.

The Late Archaic Period, ca. 1000 B.C. to A.D. 600: The beginning of the Late Archaic Period coincides with a brief interval of wetter, cooler climate, which in turn promoted an increase in grassland vegetation. This environmental change is most clearly shown in the archeological record by the re-use of Bonfire Shelter, after a period of about 7,500 years, for mass bison kills in which herds of bison were stampeded over the cliff above, probably on several different occasions. Their carcasses fell and were butchered; then, portions were dragged into the shelter and stripped of meat, and the bones were broken and discarded. The remains of an estimated 800 animals form a massive bone layer 3 feet thick. The rotting meat scraps, bone, and grease apparently ignited spontaneously (much as oily rags left in a pile may), thereby reducing much of the deposit to brittle fragments and ash, and obscuring any possibility of detecting layers that might reflect different hunting episodes. Radiocarbon dates were derived from this burned bone and charcoal taken from two small hearths, remnants of fires lit during the butchering process. The most reliable assays—those run on charcoal samples—average 2,645 years ago (Turpin 1991).

The large dart points found in this bone bed are similar to those defined as Montell, Castroville, and Marshall—types that are more common on Late Archaic sites in Central Texas. Interestingly, these specimens are also very much like forms found on some Archaic bison kills in the Texas Panhandle. However, similar styles are found in quantity at the Devils Mouth Site and Arenosa Shelter, in numerous other excavations, and on open sites. In all of these contexts, they are considered as time markers for the period. Because Bonfire Shelter is the only site where mass hunting techniques were clearly employed, the excavator came to the conclusion that for a very brief time the bison of the Great Plains expanded their range to include the Lower Pecos. There they were followed by their human predators, bearing their characteristic weapons. When the trend to aridity was reinstated, the herds and their attendant hunters retreated with the grasslands, leaving only these slightly perceptible traces of their intrusion.

A second change possibly coincides with the advent of the bison herds and hunters. In the miniature monochrome art style called Red Linear (Turpin 1984), scenes of combat, hunting, and ritual include one representation of a herd of bison racing toward a crack in the shelter wall. Although no concrete radiocarbon dates or associated cultural deposits can be related to this art style, the uncanny resemblance of this scene to the events at Bonfire Shelter seems more than coincidental.

The Red Linear style pictographs differ from the earlier Pecos River style in size, color, location, and themes. Despite their almost-cartoon-like simplicity, these stick figure drawings are much more representative of everyday life among hunter-gatherers. Scenes of hunting, warfare, ritual, and procession brim with a vivacious energy foreign to the other Lower Pecos pictograph styles. However, the underlying themes are more solemn, with their concern for basic survival in the form of conquest, hunting success, and human reproduction. Red Linear is the one style that deals with sexuality and gender. The emphasis on pregnancy and erect phalluses in this style hint at fertility as a dominant factor in ritual painting. The secluded location of some Red Linear panels and the male-oriented activities portrayed suggest ritual painting, perhaps for puberty ceremonies or initiation rites.

As was the case in the Pecos River Style, the weapons pictured are of Archaic age, but their relative age is demonstrated in the few sites that contain examples of both types. Red Linear artists often incorporated elements of the older paintings into their compositions, just as they often used natural features of the shelter walls as a backdrop for their vignettes. The consistency in Red Linear art and the lack of a developmental sequence may result from its origin in another medium, such as hide painting, and it may be that it was brought into the region by the bison hunters, who mimicked the earlier occupants by leaving their own artistic tradition on the walls.

Although the bison and their hunters were probably not in the Lower Pecos for more than a few decades or centuries, their very presence suggests that a different way of life may have been practiced for this short period, at least by the followers of the herds. The driving and mass slaughter of large, untamed herd animals depend on cooperation between the hunters. Accounts of Plains Indian communal hunts prior to their acquisition of the horse describe methods of positioning the herds for stampede. In one approach, a V-shaped drive line was built of piled stones, arranged much like a funnel whose opening was the cliff or bluff selected for the jump. Hunters lay behind the piled stones hidden beneath brush or hides. A human decoy dressed in a bison costume wandered between the animals and the jump location. Curious bison would drift toward this runner until he had attracted them to the head of the drive line. Then, at the appropriate moment, hidden hunters would leap up waving brush or hides to start the stampede. The decoy would jump into a hole dug for that purpose, or dart out of the way as the herd began to run. If the lead animals saw the danger and tried to stop, the mass of panicked animals behind pushed them over the edge. This technique reportedly required considerable coordination and timing so that the herd did not turn from the chosen path. Discipline was imposed so that an over-anxious hunter did not prematurely start the stampede or frighten the animals before they were in the proper position. The shooting of individual bison was discouraged because it was detrimental to the success of the mass kill. This, and the efficiency of this killing technique, perhaps explain why so few projectile points are found in many kill sites. Another possible technique, known from historic times, was the judicious use of grass fires to force the herds to run in the chosen direction. It is not known what methods were used to stampede the bison of Bonfire Shelter, but it is certain that they were successful.

If we can continue analogy with historic American Indians here, we can assume that the bison was highly prized for a number of uses. The large quantities of meat obtained were probably dried on racks in the sun, or over low fires. The mixture of powdered dried meat with fat or bone marrow and various berries or fruits formed a storable staple. Bones were shattered to get to the nutritious marrow, and boiled for bone grease. Tools were made from some of the bones. Bison hides were preferred for robes, and in later times were used to make war shields. Bison horn headdresses were also prized. Very few parts of the animal could not be used to advantage. However, at Bonfire Shelter, waste is apparent in the number of skeletal parts that were not dismembered, and in the paucity of butchering tools recovered. It seems that the hunters were too successful, and that the overkill exceeded the labor force available to fully process all the carcasses.

The meat and hides garnered from the mass slaughter seemingly would require immediate processing at a nearby camp. There are several likely shelters and open camp sites close by, but only Eagle Cave, half a mile down the canyon, shows any evidence of bison or the characteristic projectile point styles. In addition, bison bone has been found in deposits of this time period at Arenosa Shelter (Dibble 1967), Skyline Shelter (Turpin 1990b), and at the Devils Rockshelter. Piedra del Diablo, an open burned rock midden site, reflects heavy use during this time, based on the quantity of Marshall points and one radiocarbon date of 3,000 years ago. The primary feature associated with these characteristic points was a large accumulation of burned limestone. This site, and the similar projectile points found on other open burned rock middens suggest that the bison hunters used rock ovens to process meat or desert vegetation, much as did their predecessors and the following Late Prehistoric residents.

With the return to arid conditions came the retreat of the bison and their hunters, and the apparent revival of the old way of life. Whether or not the resident Archaic people were displaced, exterminated, or co-existed with the bison hunters during the short span of this intrusion is not clearly shown in the archeological record.

Ensor and Frio dart points are the hallmarks of this late phase of the Late Archaic. These time markers are found in upper shelter deposits, and on open sites. They are the most common styles associated with the circular or crescent-shaped burned rock middens described in the following section. Their use continues well into the Late Prehistoric Period, after the first signs of adoption of the bow and arrow.

By far the majority of the burials recovered from shelter deposits are of Late Archaic age (Turpin 1988a, 1990b; Turpin, Henneberg, and Riskand 1986; Powell 1991). Thus, our knowledge of Archaic burial practices is greatest for this time period (Turpin 1992). The general mortuary pattern consists of flexed or semi-flexed burial—usually close to the shelter wall. These burials are often wrapped in matting or overlain with basketry, and the pit covered with limestone slabs. Very rarely, the grave holds remnants of clothing, ornaments made of shell and bone, or tools used by the person during a lifetime. Infants, wrapped in deer or rabbit skin robes and matting, are interred with their broken cradle boards (Banks and Rutenberg 1982). That some ceremony or ritual for the dead was performed is indicated by the traces of ochre (powdered hematite) found in some graves. The magical properties of ochre are suggested by the projectile points stained with red pigment, and the use of ochre as paint for the pictographs. Skeletons have been found interred under rock falls, in smaller clefts, and in sinkholes. Based on the number of individuals represented by skeletal remains and the absence of household debris, some small shelters may have been specifically set aside as burial locations. Cremation, with the burned bones and ashes buried in pouches, is another approach to disposal of the dead. It has been suggested that cremation was afforded only to individuals who had some special status in the group, such as shamans. However, considering the long human occupation of the region, burials reported from formal excavations are relatively few.

The number of upland sites bearing Late Archaic points, coupled with an increase in the proportionate number of unifaces—usually considered as vegetal material processing tools—has led to speculation concerning an escalating reliance on desert plants. At the same time, a higher recovery of fish bones and scales in shelter deposits suggests exploitation of food sources that were previously less important. The analysis of the stomach contents of one adult male who died about 1,150 years ago, late in the transition from the Late Archaic to the Late Prehistoric Period, shows a remarkably eclectic last meal or meals: represented were parts of bat, snake, fledgling birds, small whole fish, bone elements, fur of white-tail mice, pocket gopher, and pricklypear, grass, and other plants—but the main course was grasshopper parts. It has been proposed that either a population increase or a deteriorating environment forced the people to rely upon food sources they had not fully utilized previously. This point must be reconciled, in order that the transition from the Late Archaic to the Late Prehistoric period can be understood.

THE LATE PREHISTORIC PERIOD: THE BEGINNINGS OF CHANGE

The Late Prehistoric is the period most marked by changes that affected many of the basic patterns established during the long Archaic stage. The first sign of change—and the one used as a time marker for the period—is the adoption of the bow and arrow. Arrow points are found in the upper level of sites mixed with dart points of Late Archaic age. The earliest radiocarbon date associated with arrow points—A.D. 510 to 630—comes from the top stratum of Arenosa Shelter (Dibble 1967), now under Lake Amistad. This date is difficult to interpret because this level was highly mixed, containing items ranging in age from Late Archaic Ensor and Frio dart points to historic trash. Arrow point types include Scallorn, Sabinal, Perdiz, Toyah, Cliffton, and Harrell. Few good dates have been obtained to clarify the sequence of arrow point types, but evidence from outside the region suggests that the earlier expanding-stem Scallorn was replaced by the contracting-stem Perdiz. Toyah and Harrell arrow points are considered to be later.

The drawing of human figures with bow in hand in the Red Monochrome pictographs places this style in the Late Prehistoric Period. The contrast between the realistic humans and animals characteristic of Red Monochrome art and the mythic qualities of the earlier Pecos River pictographs is so strong that the Red Monochrome is recognized as intrusive, brought into the region as a fully developed art form. The close similarity to rock art of the Big Bend region and the western Plains points to a possible origin with the nomadic hunters who traveled the Southern Plains in late prehistory and early history.

The most recent prehistoric pictographs—Red Monochrome—are definitely of Late Prehistoric age, painted after A.D. 600 and before European contact in 1590 (Turpin 1986a). This can be assumed from the frequent drawing of the bow and arrow, which were introduced into the Lower Pecos at that time. This style is also markedly different from the preceding styles, which indicates a break in art, and possibly cultural, traditions. Life-size and life-like human figures are posed frontally, with arms and legs outstretched. Realistic animals of many economically useful species such as deer, turkey, rabbit, and catfish—as well as panther and coyote—are clearly shown. Hostility is prominently displayed in figures riddled with arrows. However, this characteristic is not limited to this style. The rarity of Red Monochrome pictographs and the strong similarities between the sites lead to the conclusion that this style was brought into the region in fully developed form but lasted only a short time. The strongest resemblance to other art styles is with certain figures in the Big Bend region of Texas.

A fourth prehistoric style, the Bold Line Geometrics (Turpin 1986b), are abstract designs, usually painted in thick, glossy, red pigment. The artists used various combinations of straight lines and blank spaces to form zigzag, triangular, and diamond-shaped patterns. The only variation on this theme is an occasional four-legged, insect-like figure. Geometric designs are the most common form of prehistoric art, and their abstract nature makes them very difficult to interpret. However, the Bold Line Geometrics are very similar to pictographs found in northern Mexico, and may be another example of radical change brought about by the movement of people across the Rio Grande.

A new form of mortuary custom—burial under low mounds of rock—may also have come into use during the Late Prehistoric Period (Turpin 1992). The only cairn excavated to date contained four projectile points: two dart points similar to Late Archaic forms—a Perdiz, and a Scallorn arrow point; and one smooth stone in the heart of the feature. No skeletal material was recovered, but chemical tests of the fill show a high phosphate concentration, indicating the decomposition of large quantities of bone or other organic material. Similar features are common higher on the Pecos River and along the margins of the Southern Plains, where their function as burial cairns has long been known.

Among the upland sites attributable to this time period are several locations with the structural remains of a different type of housing. Stone circles, popularly called "tipi rings," or "wickiup rings," made of six to eight paired blocks, probably held supple pole frameworks for grass- or hide-covered huts much like those used by the historic Comanche. Some of these structural remnants are found adjacent to burned rock middens, and bear a mixture of Archaic and Late Prehistoric stone tools. On the later ring sites, artifacts such as small, triangular arrow points, steeply bitted end scrapers, and a few fragments of plain brown pottery are more characteristic of sites on the fringes of the Southern Plains than any earlier Lower Pecos tool kits. Similar ceramics are informally called "Abilene Brown," indicative of their common occurrence in that area. A few potsherds in the collection from the Lipan Apache mission of San Lorenzo de la Cruz indicate the late date of this pottery type. The stone circle sites and their characteristic artifact assemblage are probably of very late or proto-historic age, and could very well represent early Apache occupation of the Lower Pecos. Alternatively, Spanish documents refer to the practice of several tribes along the Rio Grande gathering at the mouth of the Pecos for winter bison hunting. These upland camps would be appropriately suited for such a temporary occupation.

No change in the basic subsistence during the Late Prehistoric Period has yet been detected, but because most of the analyses have centered on rockshelters with very sparse late deposits, this may be simply a matter for re-direction of study. Based on the number of Ensor and Frio dart points found on ring- or crescent-shaped burned rock middens, these sites have generally been considered as Late Archaic. However, radiocarbon assays of charcoal from many of these features indicate that they were used from A.D. 700 until historic times. This places them with in the Late Prehistoric Period, as defined here. Stone circles found adjacent to burned rock middens contain both arrow points and Late Archaic points.

The distinctive shape of the ring middens has been explained by drawing analogy with documented historic Apache baking ovens. A circular pit is lined with stones, filled with dry wood, and ignited. When the coals burn down, the hearts of sotol or agave are thrown in, covered with wet grass and twigs, and blanketed with earth; and a second fire is lit above. The tubers bake for 1 or 2 days, making them suitable for storage and later consumption. The burned rock, when scraped away to expose the pit oven, forms the crescent or circle characteristic of these sites. These features reflect a continued reliance on desert plants such as sotol and agave, but the shape of the remnant rock pile is distinctive.

Mussel shell is often found on the burned rock midden sites, indicative of a secondary food source. Faunal remains have not been recovered from these upland sites in quantities sufficient to determine the full range of game animals. Species recovered from shelter deposits of this time period do not differ from those of earlier times. The animals shown in the rock art are those known from Archaic deposits, and all are still found in the region today. The spread of bison late in prehistory is known from many sites on the Southern Plains, in central and south Texas. A bison-hide-lined pit in one shelter, a few very late pictographs showing bison, and comments in Spanish documents from Northern Mexico suggest that bison were available to the Lower Pecos residents; however, no strong evidence for a heavy reliance on these large herd animals has been found.

The Late Prehistoric Period has only recently come under intensive study, and much remains to be learned. These dramatic changes in fundamental cultural patterns, such as burial customs, art, weaponry, and structures, are usually attributed to an influx of new people, but until an accurate assessment of the age of each of these traits can be gained, their sequence and origin will remain debatable.

THE HISTORIC INDIANS: THE LATECOMERS

Technically, the historic period begins with the first Spanish expedition to cross the Lower Pecos: Gaspar Castaño de Sosa's journey from Monclova, Mexico, to the Pecos Pueblo in 1590-1591. In reality, repercussions from Spanish colonization probably touched the lives of the native population before this date. One possible source of information about the approaching Europeans was the native tribesmen of northern Mexico, who reportedly retreated across the Rio Grande, eluding slavery and epidemic diseases. Spanish documents also refer to the northern Mexican practice of traveling to the Rio Grande for winter bison hunts and trade. Here they met with the nomadic hunters of the Southern Plains, called the "Cibolos" and "Jumanos" in the Spanish records. The arid lands and rugged terrain presented no significant barrier to these people long before the horse granted them mobility. The Spanish considered their desert opponents savage and ferocious—a natural response to the collision of two such very different cultures. To what extent the natives of the Lower Pecos warred among themselves is unknown, but the Spanish threat to their traditional lifeways was clear. It seems probable that some form of alliance made it possible for the northern and southern native groups to co-exist at a temporary meeting ground along the Rio Grande. At least, by 1693, La Junta—a Spanish campaign to the Conchos-Rio Grande area—found that the camps of allies of the southern tribes lined the Rio Grande to the mouth of the Pecos. Thus, the Lower Pecos may have served as a refuge long before the Spanish scribes recorded the movements of the 17th and 18th centuries.

The second possible source of unrecorded contact may have come in the form of illicit commerce in slaves. As Lieutenant Governor of Nueva Leon, Gaspar Castaño de Sosa was allegedly involved in widespread slave trade, sending out expeditions to garner labor for the mines and fields of Northern Mexico. His ill-prepared and hasty expedition in 1590 was to be financed in part by capture of Indian slaves along the route. His knowledge of the location of the mouth of the Pecos may well have come from slaves or slavers. His ignorance of the rough terrain beyond would suggest that the Rio Grande may have been the upper limit of his information. However, he was aware that the Pecos was the same river that flowed by the pueblos described by the Espejo expedition of 1582. Slavers kept no records of their illegal operations, and no tales of the Indians are preserved by a written language. Thus, in that sense, Castaño's travels provide us with the first sketchy description of the Lower Pecos in historic times. When his wagon train—composed of 160 to 170 people, eight or 10 two-wheeled ox-carts, cattle, goats, horses, and dogs—left Monclova without permission from the Spanish crown, he became a fugitive, eventually apprehended and returned to Mexico. Exiled to the Philippines, he died in a revolt by his Chinese galley slaves. The court reversal of the verdict of exile came too late. His later adventures are a topic for historians; only the casual facts noted as he passed through the Lower Pecos are of interest here.

His caravan crossed the Rio Grande beneath present-day Del Rio, and headed west to a camping ground on Seminole Canyon. From there, with an ultimate destination of the Pecos Pueblo, his wagon train wandered between the Pecos and Devils Rivers for 26 days, searching for a crossing, and for water sufficient for the needs of such a large group. The stony terrain reportedly cost them 25 dozen pairs of horseshoes. In the midst of all the complaints about difficult terrain and the inaccessibility of the rivers, it is perhaps significant that at no time was a shortage of feed for the numerous domestic animals mentioned. This, coupled with other ethnographic references to bison hunting at the mouth of the Pecos, gives the impression that the Lower Pecos was marginally included in the great expansion of grasslands that turned much of the American West into a savannah parkland for a few centuries.

Yet another clue—this time to the nature of the native population—is the complete absence of any reference to encounters with Indians from the Rio Grande until the wagon train finally came to the Pecos River below the present-day ruins of Fort Lancaster, several weeks into the journey. A scouting party reported that they came upon a great number of people of the Tepelguan (or Depesguan) nation, who gave them buffalo and antelope meat and skins, and offered to take them to a place of settlements and abundant maize. "Tepelguan" is probably derived from the Nahuatl words "tepetl" (mountain) and "guan" (at the junction of), implying the geographic origin of this group rather than their tribal identity. The translators of Casta&ntildde;o's journal suggest these may have been members of the poorly known Jumano—buffalo hunters of the Southern Plains who had wide-ranging contacts with settled tribes to the east and west. In some reports, the Jumano are settled agriculturalists of the Rio Grande; in others, they are nomadic hunters of the Plains, or traders with the Caddo of East Texas.

Such mobility is not uncommon among hunters and gatherers, but the discrepancies in these reports lead some authors to suggest the name "Jumano" was applied indiscriminately to any tattooed or body-painted Indians. This difficulty in determining exactly who the Spanish encountered is only the first in a long series of ambiguities. Dozens of tribes are named by various expeditions to other parts of the Southwest, but whether the names are those by which the people called themselves or those stuck on by the explorers because of some characteristic is generally impossible to determine. The names of individual leaders were often applied to the group as a whole, and any one group could be called by several names. Even if the tribal identification of these people is correct, the Jumano remain one of the least known of Texas tribes. Thus, this identification does little to clarify the true identity of the "Tepelguan" met by Castaño's scouts.

The apparent de-population of the area immediately north of the Rio Grande could be more easily understood if the natives had experienced the effects of contact with the Spanish slavers, or if Castaño was, in fact, capturing them for shipment to Nuevo Leon. If the former were true, at the first sight and sound of Europeans, the people would be well advised to take refuge in the many hiding places afforded by the rugged terrain. In the latter case, no records would be kept as evidence. Thus, it is possible that Castaño's first recorded encounter with Indians, north of the Rio Grande, high on the Pecos River near present-day Sheffield, more accurately reflects the range of their experience with the white man than any population distribution.

Eighty years passed before Spanish missionary expeditions again penetrated the canyon country north of the Rio Grande. There, they encountered native groups with such little-known, exotic names as the Boboles, Guyquachales, Tiltilqui, Yoricas, Bacoras, and Mayhuam—peoples who were moving north ahead of the Spanish tide. The Great Pueblo Revolts of 1680 turned the brunt of Apache force southeastward, catching the remnants of earlier populations between them and the Spanish Colonial frontier. By this time, smallpox, measles, and other imported contagious diseases had afflicted the indigenous populations, who had no natural immunity. Whether the few survivors became stragglers in lower Rio Grande missions or refugees in northern Mexico, the fate of these people is unknown.

During the 18th century, Spanish exploration of the Lower Pecos gave way to retaliatory expeditions responding to Apache depredations. Military communiques record the demise of the Cibolos and Jumanos—long-time allies of the Spanish—at the hands of the Apaches, who reportedly dominated the Rio Grande by 1729. The Spanish line of defense depended upon small, poorly equipped garrisons that were ineffective against the mobile guerrilla warfare characteristic of the Plains Indians. The base closest to the Lower Pecos—the Presidio Aguaverde, on the San Diego River, in the modern town of Jimenez— lasted from 1773-1780—a bare 7 years. The Apache threat was only extinguished by the arrival of their traditional enemies, the Comanches, in a classic example of a cure that was worse than the illness.

The general inability of the colonial power to pacify the frontier was disrupted by revolution and wars on both continents. Distracted by the internal throes of independence, the various Mexican governments were forced to abandon their northern colonists to the depredations of the united Comanche Kiowa raiders who annually came south to garner slaves and booty.

When the Mexican American war ended with the Treaty of Guadalupe Hidalgo in 1848, the hostile Indians were quick to recognize the advantages of an international boundary, crossing the Rio Grande—and political jurisdiction—with impunity. The American government's immediate concern was linking east and west. In 1849, the famous Texas Ranger, Jack Hays, was charged with exploring the road that came to be known as the Lower Route to El Paso and on to California. Crossing the river that the Spanish called the San Pedro, Hays declared that it more likely belonged to the Devil—a name it bears to this day. In 1852, Fort Clark was established squarely athwart the Comanche Trace, one of the major Indian trails that led from the Red River to the settled communities of Coahuila. A year later, a major crossing of the Devils River was selected as the site of Camp Hudson, a garrison charged with protecting the road between Fort Clark and Fort Lancaster, higher on the Pecos River. A number of notable battles were fought on the bluffs of the Devils River before the Civil War intervened, re directing the efforts of the military, and allowing the Comanches and Kiowas to grow in power.

The Civil War also served as a pretext for the Massacre of Dove Creek, near San Angelo—an event that was to have far-reaching implications. When the Texas Rangers ambushed a camp of peaceful Kickapoo Indians who were in the process of moving to Mexico, they started an undeclared war that was to ravage south Texas, supply the Mexican army with horses and cattle, raise animosity between Mexico and the U.S., and cause the U.S. military to violate international law. The surviving Kickapoos and their Mexican relatives protected Mexican towns and citizens from border raids, and were, in turn, provided with a haven from which they wreaked revenge on Texas. Allied with the Apaches, the Kickapoos were formidable warriors, feared by even the most ferocious of their enemies, the Comanches. Finally, after negotiations with the Mexican government reached stalemate, in 1873, the U.S. Army crossed the Rio Grande and attacked the Kickapoo village at Rey Molino in the mountains north of modern-day Musquiz. Capturing women and children and holding them hostage succeeded where open combat had not. When the Kickapoo warriors surrendered, they and their families were transported back to the Middle West, taking a route that circumvented the mountains south of Del Rio to avoid encountering outraged citizens of Texas.

Fort Clark, in Brackettville, was the headquarters of the Seminole Indian Black Scouts, ably led by John Bullis—romantically known as "the Lightning Bolt of the Frontier." Bullis and his men were famous for their bravery, a reputation earned by the four Congressional Medal of Honor winners buried in the Seminole cemetery just outside Brackettville. Bullis also blazed the military road west, dynamiting a passage across the Pecos River canyon not far from the modern bridge. Always on the move in search of their elusive targets, the Seminole scouts are credited with digging wells and building camps at sites that were later to become famous: Howard's Well, for one of the bloodiest massacres in Texas history; and Meyers Springs, the largest historic rock art site on record.

Under pressure from the military, their routes to Mexico cut off, their population dwindling, and the bison herds decimated, the Comanches and Kiowas retreated north, consolidating their forces around their home territory. The last major battle fought in the Lower Pecos region took place at Kickapoo Springs, in Edwards County north of Fort Clark, in 1872.

Historic Indian encampments are the rarest of Lower Pecos archeological sites. Forrest Kirkland, the noted recorder of the rock art, published an account of a rockshelter (Kirkland 1942) near Pandale where glass beads, glass, metal bells, scrap metal, scissors, and brass buttons were found. A local resident claimed that the buttons were from soldiers stationed nearby in 1879 following a request to Fort Clark for help in controlling the Apaches, but this date seems contradictory in the light of regional history. The only other recorded historic Indian camp site consists of a series of tipi rings adjacent to a historic pictograph site that produced an arrow point and a ceramic fragment, both commonly attributed to the Mission era. The scarcity of historic Indian remains is understandable when we take into account the short duration of their occupation of the Lower Pecos and the mobility of their lifestyle. If the Comanche camp is typical, the meager possessions would leave little trace once camp was broken and moved.

Fourteen pictographs and one petroglyph panel convey some of the world view of the transient groups that sequentially dominated the Lower Pecos River, protected by the white man's discomfort with the terrain and his temporary contempt for settling the region. Until such time as more occupational sites are located, these murals provide the best evidence of the historic American Indian presence in the region. Changes in subject matter and style chronicle the attitude of indigenous people to their new experience with European culture. Curiosity gives way to overt hostility until the warrior ethic of the Plains becomes the dominant theme.

The Spanish era is represented by detailed pictures of mission churches, missionaries, horses, and domestic animals (Turpin 1988b). Creeping hostility appears at Missionary Shelter (Turpin 1989), where the composite figure of a church and priest is impaled by a lance. Bison, tipis, guns, white riders pierced by lances, and mounted warriors reflect both the warrior ethic and the biographical style of the Plains Indians, and were probably painted by Comanche or Kiowa artists during the 19th century (Turpin 1988c).

Who, then, were the native people of the Lower Pecos region? Although the Spanish called the arid lands of southwestern Texas and northern Mexico the "despoblado," which means unpopulated, their documents provide the names of a number of groups who were apparently already displaced persons who had moved north ahead of the conquest. "Despoblado" was probably partly a value judgement, indicating that the area was unfit for Christian Europeans, and also partly a reflection of the decimation of the local people by disease and warfare. The distinction is of consequence only in the myth currently circulated that "Pecos Man" vanished into the air, leaving only a mysterious legacy of art and artifacts. In fact, the original inhabitants were the victims of hostile expansion, which began before the advent of the white man, and their displacers met the same fate as they. No chroniclers traveled with the native peoples, and few white men ever penetrated their rugged refuge. However, if the history of other Indian nations can be extended to this area, the few hundred years between contact and eradication were centuries of displacement by other native cultures, retreat, bloodshed, starvation, and disease. Thus, a 12,000-year history of native life in the Lower Pecos ebbed away, leaving only scattered traces in the genetic pool of northern Mexico and lower Texas.

All the later people—the ones with recognizable names—were intruders. The Apaches moved into the Lower Pecos after the Great Pueblo Wars of 1680, only to be caught in a vise between the Spanish and Comanches, their traditional enemies. The "Norteños"—as the Comanches, Kiowas, and their allies were known—maintained their camps on the Red River, riding south through the Lower Pecos to raid the settled communities of Coahuila, but returning home with their captives and booty. The Kickapoos migrated from the U.S. to the mountains of Mexico, from whence they ravaged south Texas. The Seminole scouts stationed at Fort Clark were Black-Indian imports from Florida, who achieved hero status in the frontier war. All these people, friends and foes alike, fell victim to the Anglo-American—the bearer of civilization, whose iron horse was the railroad. Completed in 1881, the railroad linked east and west and opened the area to settlement and commerce.

THE SOCIAL STRUCTURE OF LOWER PECOS ABORIGINAL LIFE

Each of the broad time periods in Lower Pecos prehistory is characterized by a variety of cultural traits, most obviously those reflected in the material objects that have survived over time. However, despite shifts in emphasis—for example from hunting to gathering—much of the basic lifeways remained the same. By comparing the information on Lower Pecos technology and subsistence with living or documented groups in similar arid environments, some inferences about the social structure of the people can be made.

Although the emphasis may have ranged from mass killing of herd animals to trapping of rodents or roasting of pricklypear pads, hunting and foraging were always the mainstay of the Lower Pecos economy. Slight differences in social attitudes probably prevailed during times when hunting had priority, but those are also the periods about which we know the least.

The diversity and distribution of food resources called for a nomadic life, in which moving from place to place was dictated by the availability of food, water, and other resources. Sedentism—that is, living in permanent settlements—is usually prompted by a surplus food supply that can be stored for use throughout the year. Although some hunting and gathering groups are granted such a surplus by an abundant nature, all current information indicates that the Lower Pecos people were nomadic throughout prehistory.

As another form of human adaptation, the structure of society can be inferred from the conditions imposed by a hunting-gathering lifeway in an arid environment. The population must remain sparse and thinly distributed in order to exploit resources without seriously depleting them. Thus, the basic social unit would be the extended family—a small, highly mobile group drawn together by ties of kinship. The familial relationships served to further the unity of the group, increasing cooperation through familiarity and affection. Other members may have been attracted by the success of some individual, or simple compatibility. Group composition was fluid, permitting any friction to be resolved by moving to another band. Practicing a nomadic life, these small groups could make seasonal rounds, harvesting plant foods as they ripened, or moving when local supplies were diminished. In anthropological terms, the social structure of the Lower Pecos remained at the band level of organization.

During times of relative plenty, for example, when desert fruits such as pricklypear ripened, the dispersed families congregated for communal harvests and celebrations. The need for exogamy—the rule of marrying outside the group—could be satisfied at these meetings. A forum for the exchange of information was provided, and political ties were formed or cemented. It seems probable that much of the most elaborate rock art was painted as part of rituals enacted during these periods of aggregation.

Marriages contracted between members of different groups served a most useful function by providing ties with a larger segment of the population. The more links an individual could forge, the more options would be available during times of stress. Thus, marriage was a political alliance, with implications beyond that of the individuals concerned. However, the union was probably highly informal by modern standards.

By necessity, some form of population control—sexual abstinence, or infanticide—would augment the high infant mortality rate to keep the nuclear family small, and within the limits imposed by the food supply and the need for mobility. Life expectancy was short, and children probably made a rapid transition to adulthood, learning to perform their mature roles at an early age. Miniature weapons and domestic utensils found in shelter deposits would serve the dual purpose of amusement and education.

Women, in an economy that emphasizes the gathering of plant foods, would be responsible for the bulk of the foodstuffs. Girls would learn to harvest and process plant material for both consumption and the manufacture of clothing and utensils. Traditionally, men would control the hunting of larger animals and the manufacture of the more complex stone tools. The production of the finest projectile points involves labor that goes far beyond that needed to produce an efficient weapon. Such an expenditure of time and effort to produce esthetically pleasing tools must have carried psychological or social rewards beyond that of sheer utility.

In all human societies, some able members of the group fulfill the need for leadership, although the degree to which power can be exercised corresponds to the population density and the resources to be controlled. In an area such as the Lower Pecos, where people and basic commodities were sparse and dispersed, a leader would usually rise to the top during times when organization was needed for a specific task. Examples might include the direction of a communal hunt by the most able hunter, or the leading of a raid or defense of the camp by the bravest warrior. Many Texas tribes had both war and peace leaders, on the principle that the one most qualified to fight may not have been the best advocate of reconciliation. Because the headman's power was simply a matter of acknowledged ability, high rank was not hereditary or endowed with much control over the rest of society. Often such a leader functioned only during his prime, to be later replaced by another, more competitive and able man. The elderly were not without merit, with their experience and wisdom often contributing to endurance through times of stress.

The oldest and most widespread form of religious thought—shamanism—is found among hunter-gatherer societies throughout the world. The most obvious evidence for its role in Lower Pecos ideology is Pecos River style pictographs. These complex and elaborate artworks are considered to be shamanic works, perhaps visions achieved in trance and made permanent on the walls of the canyons. Traditionally, the shaman, or medicine man, was charged with the general welfare of the group. Through their special relationship with the supernatural world, shamans were able to ensure the success of a hunt or battle, cast spells to enchant man and animal, and communicate with the spirits. In addition to the shaman's role as magician or conjurer, as the most learned person in the group the shaman also served as a storehouse of oral tradition, stories, myths, and facts about the group's history, and about the order of the universe. In many respects, the shaman would have been a teacher, relating his accumulated knowledge to the youth. On an even more practical level, the shaman was an herbalist, curing disease through a combination of real and magical treatments. The shaman was probably the most powerful single individual in society; one can only speculate upon his fate if he were ineffectual in performing these many tasks.

If we take the dictionary interpretation of the word primitive as meaning "relating to the earliest age or period," the Lower Pecos aborigines can be considered primitive. However, in no way can their lifestyle be considered as inefficient or impoverished for all time. The pictographs for which the area is most noted testify to a complex ideological or mythological world far more varied than their economic and technological remains suggest. Furthermore, by adapting to their environment, learning to utilize rather than drastically alter the natural resources, the native peoples were able to endure for over 10,000 years. Only 400 years have passed since the first European set foot in the Lower Pecos, and only time will tell how long our culture will survive.

REFERENCES CITED

Banks, Kimball and Gary Rutenberg
1982. A Child Bundle Burial from Val Verde County, Texas. Texas Archaeological Foundation. Midland, Texas.

Bement, Leland C.
1986. Excavations of the Late Pleistocene Deposits of Bonfire Shelter, Val Verde County, Texas. Archeology Series 1. Texas Archaeological Survey, The University of Texas at Austin.

Brown, Kenneth M.
1991. Prehistoric Economics at Baker Cave: A Plan for Research. In, Papers on Lower Pecos Prehistory, edited by Solveig Turpin, pp. 87-140. Studies in Archeology 8, Texas Archeological Research Laboratory, The University of Texas at Austin.

Campbell, Thomas N.
1958. Origin of the Mescal Bean Cult. American Anthropologist 60:156-160.

Chadderdon, Mary F.
1984. Baker Cave, Val Verde County, Texas: The 1976 Excavations. Center for Archaeological Research, The University of Texas at San Antonio Special Report 13.

Collins, Michael B. and Thomas R. Hester
1968. A Wooden Mortar and Pestle from Val Verde County, Texas. Bulletin of the Texas Archeological Society 39:1-8.

Dibble, David S.
1967. Excavations at Arenosa Shelter 1965-1966. Texas Archeological Salvage Project Report, The University of Texas at Austin. Report submitted to the National Park Service.

1970. On the Significance of Additional Radiocarbon Dates From Bonfire Shelter, Texas. Plains Anthropologist 15(50):251-254.

Dibble, David S. and Dessamae Lorrain
1968. Bonfire Shelter: A Stratified Bison Kill Site, Val Verde County, Texas. Texas Memorial Museum Miscellaneous Papers 1. The University of Texas at Austin.

Greider, Terence
1966a. Speculations on Some Pictograph Sites in the Region of Amistad Reservoir. Report submitted to the National Park Service by the Texas Archeological Salvage Project, The University of Texas at Austin.

1966b. Periods in Pecos River Style Pictographs. American Antiquity 31(5):710-720.

Johnson, LeRoy, Jr.
1961. The Devils Mouth Site: A River Terrace Midden, Diablo Reservoir, Texas. Bulletin of the Texas Archeological Society 30 (for 1959):253-285.

1964. The Devils Mouth Site: A Stratified Campsite at Amistad Reservoir, Texas. Archeology Series 6, Department of Anthropology, The University of Texas at Austin.

Kirkland, Forrest
1942. Historical Material from Fielder Canyon Cave. Bulletin of the Texas Archeological and Paleontological Society 10:11-014.

Kirkland, Forrest and W. W. Newcomb, Jr.
1967. The Rock Art of Texas Indians. University of Texas Press, Austin.

Labadie, Joseph H.
1991. The Amistad Preservation Project. In, Rock Art Papers Vol. 8, edited by Ken Hedges, pp 111-130. Museum Papers No. 27, San Diego Museum of Man.

Lundelius, Ernest L., Jr.
1984. A Late Pleistocene Mammalian Fauna From Cueva Quebrada, Val Verde County, Texas. In, Special Publications of the Carnegie Museum of Natural History 8:456-481.

McGregor, Roberta
1992. Prehistoric Basketry of the Lower Pecos, Texas. Monographs in World Archaeology No. 6. Prehistory Press. Madison, Wisconsin.

Mock, Shirley B.
1987. The Painted Pebbles of the Lower Pecos: A Study of Medium, Form, and Content. Masters Thesis, Department of Anthropology, University of Texas at San Antonio.

Parsons, Mark L.
1986. Painted Pebbles: Styles and Chronology. In, Ancient Texans, by Harry J. Shafer, pp. 180-185. Gulf Publishing Company. Houston.

Powell, Joseph P.
1991. Human Skeletal Remains from Skyline Shelter (41VV930), Val Verde County, Texas. In, Papers on Lower Pecos Prehistory, edited by Solveig Turpin, pp. 149-174. Studies in Archeology 8, Texas Archeological Research Laboratory, The University of Texas at Austin.

Ross, Richard E.
1965. The Archeology of Eagle Cave. Miscellaneous Papers of the Texas Archeological Salvage Project 7. University of Texas at Austin.

Russ, Jon, Marian Hyman, and Marvin W. Rowe
1992. Direct Radiocarbon Dating of Ancient Rock Art. In, Rock Art Papers Vol. 9, edited by Ken Hedges, pp. 121-128. San Diego Museum Papers No. 28, San Diego Museum of Man.

Schuetz, Mardith K.
1956. Analysis of Val Verde County Cave Material. Bulletin of the Texas Archeological and Paleontological Society 27:129-160.

1961. An Analysis of Val Verde County Cave Materials: Part II. Bulletin of the Texas Archeological Society 31:167-205.

1963. An Analysis of Val Verde County Cave Materials: Part III. Bulletin of the Texas Archeological Society 33:131-165.

Shafer, Harry J.
1975. Clay Figurines from the Lower Pecos Region, Texas. American Antiquity 40(2):148-158.

Shafer, Harry J. and Vaughn M. Bryant, Jr. 1977. Archeological and Botanical Studies at Hinds Cave, Val Verde County, Texas. Annual Report of Research to the National Science Foundation. Texas A&M University Anthropology Laboratory, Special Series 1.

Sorrow, William B.
1968. The Devil's Mouth Site: Third Season. Miscellaneous Papers of the Texas Archeological Salvage Project 14, The University of Texas at Austin.

Turpin, Solveig
1982. Seminole Canyon: The Art and Archeology. Texas Archeological Survey Research Report 83. The University of Texas at Austin.

1984. The Red Linear Style Pictographs of the Lower Pecos River Region, Texas. Plains Anthropologist 29(105):181-198.

1986a. Pictographs in the Red Monochrome Style in the Lower Pecos River Region. Texas Archeological Society Bulletin 55:123-144 (for 1984). Austin.

1986b. Toward a Definition of a Pictograph Style: The Lower Pecos Bold Line Geometrics. Plains Anthropologist 31(1121.

1988a. Seminole Sink: Excavation of a Vertical Shaft Tomb, Val verde County, Texas. Plains Anthropologist 33(122, part 2), Memoir 22.

1988b. Rock Art of the Despoblado. Archeology 41(5):50-55.

1988c. Arroyo de Los Indios: A Historic Pictograph in Northern Coahuila, Mexico. Plains Anthropologist 33(120):279-284.

1989. The Iconography of Contact: Spanish Influences in the Rock Art of the Middle Rio Grande. In: The Columbian Consequences, Vol. 1, edited by D.H. Thomas, pp. 277-299. Smithsonian Institution Press, Washington D. C.

1990a. Speculations on the Age and Origin of the Pecos River Style, Southwest Texas. In, American Indian Rock Art Vol. 16, edited by Solveig Turpin. Proceedings of the International Rock Art Conference and 16th Annual Meeting of the American Rock Art Research Association. Austin.

1990b. Skyline Shelter: Test Excavations on the Devils River, Val Verde County, Texas. Report to Texas Historical Commission, Austin.

1991. Time Out of Mind: The Radiocarbon Chronology of the Lower Pecos River Region. In Papers on Lower Pecos Prehistory, edited by Solveig Turpin. Studies in Archeology 8, Texas Archeological Research Laboratory, The University of Texas at Austin.

1992. More about Mortuary Practices in the Lower Pecos River Region of Southwest Texas. Plains Anthropologist 37(138):7-17.

Turpin, Solveig, Maciej Henneberg, and David S. Riskand
1986. Late Archaic Mortuary Practices of the Lower Pecos River Region, Southwest Texas. Plains Anthropologist 31(114):295-316.

Word, James H. and Charles L. Douglas
1970. Excavations at Baker Cave, Val Verde County, Texas. Texas Memorial Museum Bulletin No. 16. The University of Texas at Austin.

Zintgraff, Jim and Solveig A. Turpin
1991. Pecos River Rock Art: A Photographic Essay. McPherson Publishing Company. San Antonio.