Indian Grinding Stone a/k/a Mortar

By

Andrew J. Myers

INTRODUCTION

This paper examines a small but growing number of bedrock mortars found scattered in various locations throughout the Clarion River valley of northwestern Pennsylvania. The locations of these site types varies. Some have been found in upland rock shelter sites while others have been found in large table like rocks located along stream corridors.

Mortar features are associated with the grinding and crushing of food sources such as corn, nuts, and seeds, and have not been commonly reported from sites found in the region. The most thoroughly studied of the bed rock mortar sites is a rock shelter discovered in 1991 known as Wingate's Rock Shelter (36El82). This site was found by the author while conducting phase I archaeological survey while employed by the Ridgway Ranger District office of the Allegheny National Forest. Recently however, a number of other mortars have been located while digitally photodocumenting known archaeological sites found in the region. These mortars are often found hidden under leaves and moss that accumulate and grow on the rocks. The location of one grouping of mortars was presented to the author by an informant and one appears in a local history of Ridgway, PA.

Since the research is ongoing a brief look at the known sites will be presented along with a regional overview of the Clarion River valley. Large portions of the Clarion River valley are included in the Appalachian Oak Forest and it is acorns that were likely crushed in the mortars rather than corn, due in part to certain climatic considerations. When acorns are processed in a mortar a flour-like substance is formed which creates an edible meal rich in nutrition. A description of Native American acorn management and preparation will also be presented.

Regional Overview/Growing Season

The Clarion River is one of the larger constituents that comprise the Allegheny River watershed of western Pennsylvania. The Clarion basin drains an area of 1,252 square miles (Clarion River Basin Commission) and is bounded on the north by the Tionesta Creek and the upper Allegheny River, on the south by the Redbank Creek, on the east by the headwaters of the West Branch of the Susquehanna, and on the west by the Allegheny River proper. Along with the Redbank Creek, Mahoning Creek and a number of smaller streams and runs, the Clarion River is one of a number of water courses that constitute the Central Allegheny drainage basin. These drainages have origins in the east along the divide separating the Ohio River from the Susquehanna drainage and flow west to empty into the Allegheny River (Ohio River).

The Clarion River originates in the southern portion of McKean County and travels some 101 miles before entering the Allegheny River at Foxburg in Armstrong County. The headwaters are comprised of two branches known as the East and West Branch of the Clarion River. The East branch has origins in the Catherine Swamp located near Clermont, McKean County, while the West Branch originates a short distance south from the town of Kane and the village of Mt Jewett. At Johnsonburg located some 13 miles south of the headwaters, the east and west branches meet to form the Clarion River proper. The vast majority of the river course travels through dense woodlands with the towns of Clarion and Ridgway being the largest populated areas in which the river passes.

All of the land area drained by the Clarion River is unglaciated and is included in portions of the High Plateau and Pittsburgh Low Plateau Sections of the Appalachian Plateaus Physiographic Province. This region is largely characterized by broad expanses of undulating upland plateau at times deeply cut by a myriad of streams and runs. The majority of the region is still covered by vast areas of dense forest. Thick clusters of mountain laurel and rhododendron straddle the rocky landscape in many areas making entry into some locations all but impossible.

Unlike southwestern Pennsylvania and areas along the Lake Erie Plain, portions of Northwestern and North-central Pennsylvania including the Clarion River valley, are located in areas that are only marginally suited for the growing of large amounts of corn. Groups occupying the region would need to supplement their diets with all available hunting and gathering as the option of farming, especially the growing of corn, may not have been an option in many areas.

To grow enough corn to support a large village, as many as 120 frost free days are required to grow eight row northern flint maize (Galinet 1967). Johnson (per comm. 1993) suggests that the number of days required is actually closer to 130 days. In the Clarion River valley, in those areas that average near the required 120-130 day frost free period, the possibility of crop failure is high. The growing of corn would have been somewhat unreliable due to the unpredictable early killing frosts. For example (see Johnson 1999:Table 1) in the Clarion River valley at Ridgway the mean frost free day growing season from 1951-1980 was 126.6. In that time period there were 17 years (57%) in which the frost free day growing season failed to reach the mean average. Similarly at the town of Clarion, further south and west along the Clarion River valley, data compiled there between 1951 and 1980 shows a mean frost free growing season of 130.3 days. During the thirty year study period, in 15 of the years, the growing season failed to reach the mean average. This suggests that crop failure could occur 50% of the time.

While there are some upland areas where corn could apparently have been grown these endeavors likely represent casual or supplemental farming and could not support a sizeable population. Roberts (1990) comparative study in Pottery County, PA, in which he examined both upland and riverine settings, suggests that at least in certain areas the uplands could have as many as 30 to 40 days of growing season longer than the river valleys. This suggests that corn could be grown in these areas due to the extended length of the growing season. Recent research conducted by Hasentab and Johnson (2001: 11) has demonstrated that maize horticulture would have been risky in the valley bottoms and that hilltops found throughout western Pennsylvania and southwestern New York State were preferred microenvironments for Late Prehistoric horticulturalists. This was due to a weather phenomena known as cold air drainage (NOAA 1974:259, 284, 320, 399, 423) in which the air layer near the ground on cold nights grows colder and moves topographically lower (Yoshino 1975:407) into the valley floors. This leaves certain areas located along the mountain slopes, above the valley floor and below the summit, known as thermal belts (Cox 1923:2) the preferred areas for the growing of crops. These are areas in which frosts are never observed. These thermal belts may support vulnerable vegetation long after frost has killed all green above and below (NOAA 1974: 284-285).

Still, many upland areas freeze well into early spring and are hit with late summer and early autumnal frosts that kill vegetation many weeks in advance of the river valleys below. The river valleys are aided by late summer and early autumnal morning fogs that help protect vegetation from freezing.

Regardless of location, if a major aboriginal maize horticultural subsistence was practiced in the Clarion River valley the evidence is currently lacking, although it must be noted that very few archaeological investigations have ever been conducted there. Unlike the Allegheny River in which the growing season in many areas exceeds 130 days, there are currently no known large horticultural village sites located in the Clarion River valley suggesting seasonal and temporary inhabitance.

Another important limiting factor dictating the growth of corn during Late Woodland/Late Prehistoric times, was the global cooling episode known as the Little Ice age. During the Little Ice Age or Neo-Boreal the temperatures dropped one to two degrees cooler than present. This was the lowest decline in temperature since the end of the Late Pleistocene (Campbell and Campbell 1984). This factor would have all but eliminated a viable horticultural maize economy from being practiced throughout much of western Pennsylvania including the Clarion River valley. It must be noted that there are a handful of sites from the interior of the Allegheny Plateau that have produced small amounts of corn including the Elk County earthworks (LaBar 1969); (Smith and Herbstritt 1976) and Dutch Hill (36JE132) rockshelter. These remains are small in number and could not have been grown in large enough quantities to supplement a large village.

Along with hunting and gathering other cultigens would need to be exploited. Of particular importance to this paper is the number of oak species available in the Appalachian Oak forest. The Clarion River valley is located in two general forest types. The majority of the river passes through Kuchler’s (1964) Appalachian Oak Forest while there are also areas with similar forest composition as the Northern Hardwoods Forest.

Following Kuchler (1964), species typical of the Appalachian Oak Forest include white oak (Quercus alba) and red oak, (Quercus rubra) as the dominant species. Associated species include: red maple (Acer rubrum), sugar maple (Acer saccharum), sweet birch (Betula lenta), bitternut hickory (Carya cordiformis), pignut hickory (Carya glabra), mockernut (Carya (tomentosa) alba), chestnut (Castanea dentate), American beech (Fagus grandiflia), tulip tree (Liriodendron tulipifera), Weymouth pine (Pinus strobes), scarlet oak (Quercus coccinea), bear oak (Quercus ilicifolia), chinkapin oak (Quercus muhlenbergii), chestnut oak (Quercus prinus), black oak (Quercus velutina), and hemlock (Tsuga canadensis).

Species typical of the Northern Hardwoods Forest (after Kuchler 1964) include, sugar maple (Acer saccharum), yellow birch (Betula allegheniensis), American beech (Fagus grandiflia) and hemlock (Tsuga Canadensis) as the dominant species. Associated species include: striped maple (Acer pennsylvanicum, red maple (Acer rubrum), mountain maple (Acer spicatum), white ash (Fraxinus Americana), mountain-laurel (Kalmia latifolia), Weymouth pine (Pinus strobes), black cherry (Prunus serotina), American yew (Taxus Canadensis), American basswood (Tilia Americana), and white elm (Ulmus Americana).

It is proposed that in the Clarion River valley where short growing season’s limited corn production and untimely killer frosts made agricultural production even less viable, Native American groups would be forced to supplement their diets with any and all food sources while occupying the region. The Appalachian Oak forest provided acorns which could be a viable supplemental food source to complement hunting and gathering, and minor agricultural endeavors.

Acorn Management and Preparation

The gathering and preparation of acorns for use as a food source involves a six step process. The labor involved in this process is conducted entirely by the female members of the group from beginning (gathering) to end (production of food).

The information presented herein was extracted from the book entitled "It Will Live Forever. Traditional Yosemite Indian Acorn Production" published in 1991 by Beverly Ortiz as told by Julia F. Parker. The book focuses on one way to cook acorns, that of, Yosemite Miwok/Paiute elder Lucy Tom Parker Telles as handed down to Julia Parker.

In California the Yosemite Miwok/Paiute preferred the black oak acorn due to availability, taste, color, flavor, storage capability, and thickening properties. By percent black oak (quercus kelloggii) acorns contain 31.4% water, 3.44% protein, 13.55% fats, 8.60% fiber and 41.81 % carbohydrate (Ortiz 1991:39), although depending on species, can contain up to 18% fat, 6% protein, and 68% carbohydrate which compares favorably with modern grains (Martin 1996).

Black oaks grow throughout much of California including in the mountains and hills of Northern California, the Coast Ranges as far south as San Diego, and throughout the Sierra Nevada foothills (Ortiz 1991: 39).

Black oak acorns take two years to fully develop and in good years each tree can produce between 200 and 300 pounds. Black oak acorns have always been the most preferred by the California Indians (Heizer and Elsasser 1981).

The first step involved in acorn production is gathering. When the leaves turn yellow, Grandmother Lucy taught that it was acorn gathering time (Ortiz 1991:41). Acorns fall from the trees twice each season and are gathered from the ground. The first fall consists of unhealthy, worm and insect infested acorns, and is left alone (Ortiz 1991:41). Once the acorns are gathered they must be dried. First a careful inspection of the nut occurs to see if there is any insect damage. Bad nuts are removed from the bunch and returned to the earth (Ortiz 1991:45).

The acorns were dried by placing them on the ground on a quilt or canvas cloth (Ortiz 1991:45). New "fresh" or "green" (newly gathered) acorns are too soft to shell and pound properly, acorns are usually stored for at least a year before used. Unshelled black acorns may be stored for ten to twelve years because of their high tannic acid content (Ortiz 1991:47). Acorn storage took place off the ground. Storage used to occur in a granary (Chuckah). They were built off the ground above a boulder, stump, or tree round (Ortiz 1991:47).

Once dried, the next step was cracking and shelling of the acorn. One by one each acorn is held between thumbs and index finger with its pointy end stabilized against a flat rough stone. The stone provides a firm foundation for cracking while its rough surface provides a place to secure the pointed end (Ortiz 1991:49). The flat end is then struck with the end of a small elongated rock (hammerstone) to crack its shell (Ortiz 1991:49). The cracked hulls were then removed by hand (Ortiz 1991:51).

The next step involves a process known as winnowing. This process removes a skin from the nut. This would be the same type of skin found on peanuts such as Spanish Redskins. The acorns were placed in a winnowing basket and then tossed in the air which removed the skin. At times a knife was used to aid in the removal of skins (Ortiz 1991:54).

The clean kernels were then pounded with a pestle. This is the process that forms the mortar. Like the acorn tree, the mortars are owned by an individual or family, each one handed down through the generations, every woman has her own special rock (Ortiz 1991:59).

Shallower mortars appear to be preferred over deeper mortars. The bedrock mortar holes in which Julia pounds are shallow, usually less than half an inch in depth. Deeper holes are used to pound acorns in many areas, but the flour that results from pounding the comparatively oily black oak acorn will build up at the bottom of deep holes like peanut butter making pounding ineffective (Ortiz 1991:63). The removal of the residue requires vigorous scrubbing, so different bedrock mortars are usually reserved for different foods (Ortiz 1991:63).

Smaller mortars and pestles were used to grind grass and forb seed into flour (pinole) typically were not as massive as those required from acorns (Anderson ). Many of these were portable in nature.

A method was used to best accomplish the pounding of the acorns. A few handfuls of "starter" were placed on the mortar. This keeps the whole nut meats from bouncing around during pounding. As the pounding progresses and the whole kernels begin to break up, the pestle is balanced in the meal and another handful of "starter" or whole nuts is added. This continues for awhile and eventually between four to six handfuls of nuts will be crushed in the same way with the "starter" being used to keep the nuts in place. Eventually though the pounding begins in ernest until the nuts are crushed into powder (Ortiz 1991:70).

The pounded meal is then sifted in order to get the finest flour possible. This is done in a sifting basket. Larger chunks are again pounded in the mortar (Ortiz 1991: 85).

The final step before cooking is leaching. Acorns contain bitter tasting tannic acid which needs to be flushed from the flour. The powder was placed in a sand basin and rinsed with cold water (Ortiz 1991:95) which removes the tannic acid. The flour would then be ready to cook.

The importance of the acorn as a supplemental food source cannot be underestimated. In some regions of the United States various forms of food preparation such as the grinding of nuts overshadowed large scale agriculture. For example in California according to (Ortiz 1991:ix) farming and harvesting of staple crops such as corn did not play a large role in the subsistence pattern as in other areas of North America. In heavily-wooded northwest California, the people sometimes set out in redwood dugout boats to fish rivers for salmon or in the ocean for marine animals. In the eastern deserts people lived in nomadic family groups, harvesting plant foods such as pine nuts, and hunting game such as rabbits, with great skill. In the deserts along the Colorado River, squash, beans, and corn were farmed, while in Central California, acorn was a staple food, like rice, flour, and potatoes today (Ortiz 1991: viii-ix).

Even with only small groups dabbling in farming the population of California was large. According to estimates, the number of Indian people living here 250 years ago exceeded 300,000 making California the most populous of any comparably-sized area north of Mexico (Ortiz 1991:ix). If nut gathering could support a large population in California it could surely help supplement the diet of small groups living in the Clarion River valley. In many areas throughout the Clarion River valley and in northwestern, Pennsylvania oaks and related species currently exist in large numbers. This is an interesting concept especially considering the short frost free growing season of western and west-central Pennsylvania. Many of these oak forests have likely been in place for many hundreds if not thousands of years.

Evidence suggests the Native Americans were adept at managing forest environments. At the time of the first European settlement, natives regularly burned the woods to drive game, improve visibility, facilitate travel, and provide browse for deer (Day 1953). The effects of burning are such that certain heat tolerant tree species could be influenced or even maintained through the use of fire. According to (Black et al. 2006: 1267) any increases in anthropogenic disturbance, especially fire, could have a dramatic impact on forest composition, with expected decreases in sugar maple and hemlock and increases in oak (Quercus spp.), hickory (Carya spp.), and chestnut. In a study of Native American sites in the Allegheny River drainage (Black et al. 2006) in areas with high Native American activity oak, beech, hemlock, chestnut pine, and maple dominated. In areas with low Native American activity beech, hemlock, and maple dominated.

Similarly, the Native groups living in California, were adept in the use of fire as well. According to (Martin 1996) regular low level fires in California encourage oaks. Without burning, those species with low fire resistance, such as shade tolerant conifers and brush dominate. In California burning has been largely stopped for the past 70 years and the forest looks entirely different from those witnessed by European explorers (Martin 1996). In 1844 when John C. Fremont led an expedition to the Sacramento Valley, he described the north state foothills as smooth and grassy; [the woodlands] had no undergrowth; and in the open valleys of rivulets, or around spring heads, the low groves of oak give the appearance of orchards in an old cultivated country (Martin 1996).

MORTAR SITES

Wingate's Rock Shelter

During the summer of 1991 while conducting Phase I archaeological reconnaissance for the Ridgway Ranger District office of the Allegheny National Forest the discovery of a small rockshelter was made, located high in the hills above the Clarion River valley in Elk County, Pennsylvania. While preparing to photograph and subsequently test the site for evidence of prehistoric occupation a rather unique feature was discovered within the rockshelter. Found carved into a table like rock was an intact prehistoric grinding stone a/k/a mortar.

The site was subsequently recorded by the author into the PASS files at the Carnegie Museum of Natural History in Pittsburgh as Wingates Rockshelter (36EL82) named for a forester that worked at the Ridgway Ranger District office at the time.

Wingate's Rockshelter (36EL82) is located on a ridgetop overlooking the Clarion River valley in the remote southwest portion of Elk County, PA. The nearest village of any size is the ghostown of Arroyo located some 1.07 kilometers to the southwest. Arroyo was once the setting of a sizeable tannery and now consists of only a handful of permanent residents and seasonal hunting camps.

The topography of this portion of the Clarion River valley is included in the Pittsburgh Low Plateau Section of the Appalachian Plateaus Province. Bedrock at the elevation of the rock shelter corresponds to the Pennsylvania Pottsville Formation Pottsville Sandstone. The site is positioned at the 1529 ft. elevation along a narrow ridge spine positioned well above the Clarion River valley below. While the location of the rockshelter seems remote by today’s standards it was likely one of a number of rockshelters used along well defined trail systems. The canoeable Clarion River road passed directly below the site and the rockshelter is ideally positioned and included in a proposed trail described by the author (see Myers 1997:51; Trail 1) that has antecedants well north of the site. This proposed trail originated at Buckaloons (and points northward) and traveled through the Tionesta Creek drainage before entering the Clarion River drainage on it’s way to the West Branch of the Susquehanna passing directly below Wingate's rockshelter.

36EL82 Wingates Rockshelter looking southeast

Wingate's rock shelter is quite small by local standards. General measurements are 7 meters north south by about 3 meters east to west. The space from backwall to dripline measures around 2 meters. The height of the rockshelter ranges from around 1.83 meters in the northern end to perhaps as much as 3 meters along the southern end. The rockshelter faces approximately 40-60 degrees northeast and is open 140 degrees SE and 320 SW. The habitation floor of the site is small with only some 10 square meters of clean living space being found underneath of the overhang. The majority of the northern end on the overhang is filled with talus including in the general vicinity of the mortar.

Field Sketch Map Wingates Rockshelter

Water sources are unknown from the immediate vicinity of the site. Springs are likely located along the base of the slope found below the site. Cole Run is the nearest drainage located some 213 meters northwest of the site while the Clarion River is located some 244 meters to the southeast. To reach either drainage a steep descent downward to the valley floor is required with the slope down into Cole Run being slightly more gradual that the steep drop off to the Clarion River. Relief from the rockshelter to the valley floor is some 73 meters.

Profile view of the site looking northwest

The mortar was carved in a rock found located under the protective portion of the overhang. The rock most likely did not shock from the ceiling but was one of a number of rocks found strewn as talus slope on the extremely rocky eroded hillside. The flat and soft nature of the sandstone slabs were probably key variables in selecting the particular rock for a mortar.

The mortar while located near the old dripline was not formed by running water. The main feature is too deep and too perfectly formed to have been done through natural weathering.

Dimensions

The mortar is made up of of two and perhaps as many as three individual grinding areas, two of which are now broken. The main mortar is circular in shape measuring some 17 cm in diameter. The mortar measure 18 cm north-south by 16.5 cm east- west. The mortar measures 9.5 cm in depth and is deepest near the center. Located just 2 cm sw of the main mortar is another now broken mortar. This mortar is 6 cm north to south by 11 cm east to west and now measures 6.5 cm deep. The remaining portion of this mortar is missing. A third mortar may also have been carved into the rock and is now only represented by a rounded polished arc. The rock apparently cracked on occasion resulting in the locating of new milling station periodically.

OTHER MORTARS

One mortar is mentioned in a local history book. Harry Hill, a former police chief of Ridgway, Pennsylvania, reported in his history of Ridgway entitled “The Story of Ridgway” that a mortar was discovered in a rock found along Big Mill Creek (Hill 1964: 1) located just west of Ridgway. He speculated that corn was ground on the rock. Big Mill Creek is located only a few miles east from Wingate's Rockshelter. This author has not seen the mortar described by Hill and knows of no one who could locate the feature. Mr. Hill has long since passed so this mortar has not been verified.

Mortar pictured (bottom left)by Harry Hill in The Story of Ridgway and found along Big Mill Creek

A significant milling station was brought to the attention of the author by a local informant Mr. Desmond. The site known as Millstone Creek Milling Station located on Allegheny National Forest property, has recently been recorded into the Pennsylvania Archaeological Site Survey (PASS)files at the Carnegie Museum in Pittsburgh and with the Bureau for Historic Preservation in Harrisburg as (36El209). This particualr site is located just east of the Lamonaville cross roads in the Millstone Creek drainage of Elk County. The site consists of several mortars (40+) that were carved in a grouping of three flat table like rocks located well above the present drainage course of the East Branch of Millstone Creek. They do not appear to be natural potholes caused by hydraulic drilling but rather purposefully made. The author has subsequently photographed the site and some of these will be made available at this website.

Mortar east of the village of Lamonaville in Elk County, PA

Another possible rockshelter mortar was discovered while photographing the Maple Creek Rockshelter (36FO68)located on the hillslope above the Clarion River near Clarington in Forest County, PA.

Possible mortar feature found at the Maple Creek Rockshelter located near Clarington in Forest County, PA

The Maple Creek rockshelter is very small with only about 6 meters of living space accounting for the entire site area. The ceiling of the overhang is approximately 1 meter high. Basically there was enough room for one or two few people who built a hearth and then possible ground a mortar into a rock found underneath of the overhang. The mortar itself is approximately 6 cm in diameter and 2 cm deep.

Maple Creek Rockshelter(36FO68)

Recently another mortar was discovered while digitally photographing a rock shelter in Elk County. The mortar was found in a flat rock near the dripline of the rock shelter. There are actually the remains of three seperate grinding areas in this rock. They would apparently use one mortar until it was ready to break or had ground completely through the table like rock and then shift the focus of the grinding to a new location beside the earlier version. The last one used is likely the best physical specimen. Mortar No. 1 is a whole mortar that measures 10 cm in diameter. Mortar No. 2 is approximately 8.5 cm in diameter while Mortar No. 3 is an arc of an earlier mortar which would have measured approximately 8.0 cm in diameter. The two largely intact mortars measure approximately 3 cm in depth.

Mortar found in Elk County rock shelter in the Toby Creek drainage

Elk county rock shelter housing mortar. Note: rock in foreground near dripline contains the mortar

CONCLUSIONS

There is a small but growing number of bedrock mortars found in portions of the Clarion river drainage of western Pennsylvania. These sites are found in association with oak species typical of the Appalachian Oak Forest. Due to the fact that the growing season is too short to produce adequate amounts of corn these mortar sites were likely used to grind acorns found in the oak forests. There is currently little evidence that suggests major aboriginal maize horticulture was practised in the Clarion River valley. Groups occupying the region in Late Prehistoric times were likley hunting and gathering and had arrived from horticultural village sites located in distant locations.

The importance of acorns as a subsistence base cannot be under estimated. In California numerous groups followed a subsitence strategy based on gathering and procuring of acorns. Estimates of population during the late 17th and 18th centuries suggest that California had a population nearly double in size of the entire Great Plains which extends from mid Canada to the Gulf of Mexico (Martin 1996). The grinding of acorns produces a flour which can then be transformed into a number of food items including soups and/or hard tack biscuits (Ortiz 1991:35-36). Depending on the species of acorn they compare favorably with the nutritional values of modern grains (Martin 1996).

Native American groups were adept at manageing forest by the use of low level fires. Regular use of fire encourages oaks (Martin 1996). When the fires were stopped plants with low fire resistence such as shade tolerant conifers and brush began to dominate (Martin 1996) Local groups may have used fires to influence oak species.

While the age of the mortars is unknown they may have been used seasonally by members of the McFate tradition (A.D. 1400-1575) that occupied village sites on the Glaciated Allegheny Plateau. Numerous rock shelters and upland stockaded villages were occupied by McFate people while they hunted and gathered in the Clarion River valley and surrounds. It is likely that they supplimented their diets with acorn flour and it's derivatives while travelling abroad.

Additional information on bedrock mortars will be presented as it becomes available.

References

Hill, Harry M.
1964 The Story of Ridgway. Ridgway Publishing Company. Ridgway, PA.

Myers, Andrew J.
1997 An Examination of Late Prehistoric McFate Trail Locations Pennsylvania Archaeologist 67(1):45-53.

Vento, F. J. et al.
1980 Excavations at Dameron Rockshelter (15JO23A)Johnson County, Kentucky. Ethnology Monographs Number 4, Department of Anthropology University of Pittsburgh.