Early Archaic Period:
Early Archaic - Willamette:
Long Tom Basin
Coast Fork Basin
Middle Fork Basin
Middle Archaic Period:
Middle Archaic - Willamette:
Long Tom Basin
Coast Fork Basin
Middle Fork Basin
Early Archaic Period
The overall patterns of culture known from the contact period begin to emerge: a mixed gathering and hunting economy based on plants such as camas, bitterroot, wocas, wapato, acorns, and tarweed begin to emerge. This mixed economy is the difference between the earlier hunting based Paleo-American period and the later Archaic traditions. Hunting concentrated on elk, deer and antelope. Fishing is now importance, a pattern that marks the Columbia River groups from then on. Fishing gear such as net sinkers, fish hooks and fish spears appear. Food storage pits are recorded in sites. The fluted point of the Paleo-American period are replaced by lanceolate Cascade and side-notched forms. Metates and manos reflect an increasing use of plant foods. Woven textiles, baskets, nets, mats, and sandals appear in the archaeological record.
The atlatl (see the Pleistocene section for a discussion of atlatl technology) is still the primary hunting weapon, but hunting and gathering tool kits diversify. Regional differences in style begin to appear in the artifacts and economies reflecting the major ecozones in Oregon. The beginnings of the Great Basin Desert-Archaic patterns emerge, as well as the Plateau and its river-centered fish/root pattern, and the Coastal shellfish-fishing-sea mammal hunting way of life. Trade in obsidian starts, but tends to be localized when compared to later cultures. Since the populations at this early time period were small, and since time tends to erase sites through erosion, few sites have been discovered in comparison to later periods. Sites tend to be small, but some favored locations were visited many times, and substantial deposits built up. One very early site is the Stockhoff basalt quarry near La Grande. People started using this quarry for its glassy basalt about 10,000 years ago. It was used until contact. It grew to be perhaps the largest site in Oregon at over 4,000 acres. In places it is about 12 feet thick and it contains artifacts covered by a primary Mt Mazama ash layer. It almost looks like people dropped what they were working on when the eruption turned the day to night and filled the air with thick clouds of choking ash.
Archaeological sites are given a permanent number based on a pattern developed by the Smithsonian Institution in the early days of archaeology. The 48 states were alphabetized and numbered. Oregon was the 35th state. Each county was given a two or three letter code. Lake county became "LK" in Oregon. Then, each new site was given a number, starting with "1". So site 35LK2076 is the 2076th site recorded in Lake county in Oregon. This number is put onto every artifact, so that each object can be traced back to a specific place on the earth and relate to a site form for that site (as well as reports related to the site).
At sites 35LK1881 and 35LK2076, early Holocene rabbit drive sites were found in Christmas Lake Valley. The report by Albert "Chip" Oetting indicated that the sites contained broad shallow roasting pits full of charcoal and charred bone. Fully 98% of the bone (9,825 could be identified) consisted of jackrabbit, cottontail and other members of the hare and rabbit group. At site 1881, C14 dates of 8080, 8880, 8950 and 9120 years ago confirm that large scale rabbit drives, probably with nets, were common practice at this early date in the Great Basin. A similar deposit was found at 2076. Dates of 8780, 8870 and 10020 came back from the charcoal. The bones were again primarily rabbit. A count of pelvic girdles indicated that at least 290 rabbits had been roasted in one pit. The evidence indicates that the bones had been tossed back into the roasting pit after eating.
The historic Western Shoshone and Northern Paiute are know to have organized and performed communal rabbit drives, usually in the fall when the animals and their pelts were in prime condition. Drives varied in scope, and included nets, traps, piled brush, fires and circles of people. The rabbits were frightened into running, and then were funneled into traps or locations where waiting hunters could get many quick kills. The skins were taken to make blankets or robes. The rabbits were usually pit roasted or boiled, as rabbit meat would not keep over two weeks if dried. If the skins were not used, the rabbits were dumped into shallow pits to burn off the fur and roast the animals. The bones were not removed during roasting. In the Fort Rock area, people driving 640 acres (a Section or one square mile) could count on killing between 2000 and 3000 rabbits.
In the Noti-Veneta highway project in Lane County, Brian O'Neill reported on a series of sites containing camas roasting ovens. The majority of these features dated to the Late and Middle Archaic periods, but some of the ovens fell in the Early Archaic and the transition from Paleo-American. A C14 date of 8890 for one oven placed the tradition of camas gathering and roasting in earth ovens this early. Other terminal Early Archaic dates included 4400, 4320, 4190, 4120, 4110 years ago as well as early Middle Archaic dates of 3880 and 3780 years ago. Camas occurs in wet meadows. and was usually harvested between June and October with digging sticks by the women. Large amounts of camas were roasted in shallow earth ovens. The pits were filled with a mix of wood and stones and set on fire, when the fires burned down, they were racked out, and the hot stones left as a layer. Alternating layers of leaves and camas were placed on top of the hot stones and sealed under earth to bake. This process was observed and recorded by pioneers entering the Willamette Valley. The Long Tom site shows that this tradition has a time depth of nearly 9000 years.
Further work studying the alluvial stratigraphy of the Long Tom basin by Dorothy Freidel, Lynn Peterson, Patricia McDowell and Tom Connolly has indicated that plant gathering and processing was in effect 9700 years ago. They recovered C14 dates of 9660, 9485, 9130, 8890, and 7690 years ago. The 9000 dates were 240-250 cm below the present surface and the 7000-8000 year old dates 95-165 cm deep. Dates in the 3700-4600 range fell between 85 and 100 cm deep. The burial of sites as the valley slowly fills with alluvium explains the relatively low density of surface site from surveys. The majority of Willamette Valley sites are simply buried, and the early sites nearly nine feet below current surfaces.
The Wildcat Canyon site (35GM9) contains two phases of occupation that would fall into the Early Archaic: Philippi and Canyon Phases. The earlier phase was characterized by a series of simple campsites. The later phase contained many wells dug to penetrate the aquafer of Wildcat Creek, which may have become intermittent as the climate became more arid. This digging of wells is an interesting side light on Oregon prehistory.
Early Archaic (9000-6000 BP)In the Willamette Valley.
This period is not represented in the Portland Basin data to date.
Lebow (1984) reported on the testing of the Ripple Site (35CL55), a 31 acre site on the Mt Hood National Forest at an elevation of 1500 feet. A total of 36,543 lithic items were recovered, which included 377 artifacts. Nine of the fourteen projectile points were "Cascade" style lanceolates similar to those recovered from Cascadia Cave (see Santiam basin below). Thirty-two 1x1 meter units were excavated into the site, some of which were contiguous blocks. Initial testing had indicated the site was only 20 cm deep, but the extensive testing found areas 30-40 cm and deeper. The soils had been mixed by tree falls. Based on the style of the points, he concluded the site was occupied some time around 6000 BP, at the end of the Early Archaic. "It is proposed here that the lanceolate projectile point style from Cascadia cave (and the Ripple Site, Medicine Creek Site, and Baby Rock Shelter) dating from 8000 to 6000 B.P. be termed the Cascadia Point rather than the Cascade Point" (Lebow 1984: 58). This was based on visual and measurement dissimilarities between the type defining collection of Cascade points and the materials from the Oregon sites.
The Ruby site (35CL42) on the Mt Hood National Forest was tested by Burtchard (1991). The site produced a number of foliate Cascadia style dart points, five of which were collected in 1975 and later lost. Burtchard excavated 167 three-and-a-half gallon buckets passed through 1/8 inch screens. He also dug two trenches to get a look at the stratigraphy. He recovered 1094 pieces of debitage, 27 tools (3 unifaces, 8 bifaces, 7 cores and 2 end-modified cobbles) and 16 bird gizzard stones. Birds eat small stones that grind up seeds, they become highly polished. A high percentage of the CCS had been heat treated ranging from 20% to 56% from the test units. Heat treating consists of baking CCS in earth ovens to drive out part of the water and to make it waxy in feel. This improves the flaking qualities of CCS. The author was surprised by the density of artifacts, indicating many small and perhaps intensive reuse episodes.
The data is scarce, indicating a hunting lifeway, but sampling in other areas demonstrates that plant gathering was the staple source of food throughout known human occupation in temperate North America. Populations were relatively low, and thus human groups had access to the best of the patches of available productivity. Competition for resources would have been very low. Groups could move from abundant resource to abundant resource, picking and choosing those that fit their movement strategies and leaving behind those that were less desired or harder to obtain. Choice would have been high and conflict over choice very low.
No archaeological data available at this date.
No archaeological data available at this time.
Cascadia Cave was placed on the National Register in 1990 after additional testing was done by Heritage Research Associates (Baxter et al. 1989). It contains both Early and Middle Archaic cultural deposits. It is the type-site for the Cascadia Phase developed by Beckham, Minor and Toepel (1981).
Luther Cressman sampled the cave in 1937 and concluded the site had been looted to the point of loss of scientific value. In a 1963 Screenings article, amateurs (Howard & Howard 1963) dug a pit to a depth of 12 feet. In 1964, Newman, under an NPS contract, began excavations. Near the bottom a charcoal sample gave a date of 7910 ±280 radio-carbon years ago. Newman recovered leaf-shaped dart points throughout the deposit. The dart points are Cascade style. Newman concluded that the cave represented an earlier generalized culture exploiting a broad ecological base in a forested environment (Newman 1966: 27). In 1988, HRA tested the site for the nomination. Radio-carbon dates of 4920 ±120, 6360 ±210 and 7910 ±280 were obtained.
The latter testing concluded the site was primarily an elk and deer hunting camp. The rock art, however shows male and female symbols, possible plants and bear tracks associated with what may be curvilinear representations of the Santiam River.
I have interpreted the east panel to perhaps show a family group consisting of a male and three females. Two of the females may be wives or relatives as they are connected to the male while the third female symbol is detached. The group is connected to a ladder-like complex that may reflect either a group of plants or may be a fish weir?
The Santiam River appears to be symbolized above the grouping and the presence of bears in, or near, the river may suggest the bears are fishing, which may validate the interpretation that the structure is a weir? Fish remains are difficult to find in an archaeological context. There is evidence that areas were specific to certain mystical powers. Rock art studies have shown certain symbols are more common in some areas. Because our rock art samples for the valley are so few, this thesis cannot be tested at the present time in the Willamette Valley. The large number of bear tracks at Cascadia Cave may simply be a visual statement that "bear power" was expected or common during a shamanistic trance in this basin.
The (Lindberg-Muir: 1983) work by students from Linn-Benton Community College field school under Marty Rosenson, recovered Early Archaic cascade and Middle Archaic large corner-notched points from the Horn Ridge site (35LIN252) at an elevation of 3680 feet.
Winthrop and Gray (1985) explored the Moose Molalla One site (35LIN139) at an elevation of 3950 feet. The site was in a saddle along a ridge line used as a travel route. A mix of Cascade, Side-notched and corner-removed Middle Archaic points were recovered, placing the occupation in the Baby Rock Phase... but they cautioned that the usage of broad-necked points into later periods in the uplands was still an open question.
Flenniken and Ozbun (1988) tested the Cougar Ridge Way Trail #4 site (35LIN116), where two Cascade-like dart points were recovered. Twenty samples were sourced, all to Obsidian Cliffs. They noted "Experiments have demonstrated that approximately 3500 flakes are produced in the manufacture of one corner notched dart point from a prepared flake blank".
Long Tom Basin
During the construction of the Noti-Veneta section of Highway 126 in Lane County, a series of sites were discovered and tested by OSMA. O'Neill (1987) tested the Long Tom Site (35LA439) and discovered fire pits and acorn/camas ovens. Radiocarbon dates for two of the ovens were 3880 ±90 and 4110 ±70. While these dates are Middle Archaic, later work gave the following dates: 3780 ±110, 4120 ±70, 4190 ±199, 4320 ±100, 4400 ±75, and 8890 ±120. The last date places the earliest levels of the site in the Early Archaic. In the early phase of testing, 20 features were found, then 8 in 1980 and 12 in 1985.
The series of tested sites were found on old point bars, levees and the flood plain. The point bars were younger than the levees and flood plain, and site on the bars were Late Archaic in date, while Middle and Early Archaic sites were found in the latter landforms. The Late Archaic occupation was found in the upper 40 cm and included a scatter of flakes and tools, including four small narrow-necked arrowheads. The lower component was Middle Archaic in date and consisted of ovens, fire pits, post molds, and storage pits. Also associated were hammerstones, anvils, pestles, and mortars which were not found in the upper component. The ovens averaged 130 cm in diameter and were basin-shaped and contained fire-cracked rock, charcoal and a carbonized hazelnut. The fire pits contained a single carbonized hazel nut and one of the two storage pits also continued a carbonized hazelnut shell.
The Hannavan Creek Site (35LA647) was tested by Cheatham (1988). He found a series of hearths and ovens and C14 dates of 7750 ±100 and 6830 ±100 from the over 300 charred camas bulbs found in the ovens. He also recovered cherry seeds and charred hazel nut fragments. The site was located on the edge of the grassland prairie against a gallery forest.
Flenniken, Ozbun and Fulkerson (1989b) tested the Warehouse site (35LA822) that may be Early Archaic in date, based on the hydration rates and lanceolate dart points. This site would be in Molala territory near the Mohawk Kalapuya territory. They recovered two biface fragments, one bifacial blank fragment, three preforms, three exhausted lanceolate darts and five projectile point fragments. Almost all of the tested obsidian (47) came from Obsidian Cliffs and two from the Inman source near Eugene. They felt the site was a logistic hunting camp associated with rejuvenation of projectile points.
Coast Fork Basin
No data available at this time.
Middle Fork Basin
No archaeological data available at this time.
The Early Archaic represents the initial adaptation of groups to the Holocene biological environment. Populations would have still been relatively low. The best places where resources were concentrated, would still be available with little on no competition. Cooperative groups could move from place to place and seldom run into non-cooperating other human beings. Natural biotic resources were still relatively abundant in comparison to populations exploiting the resources. Scarcity was probably not much of an issue. Competition was also not much of an issue. The kinds of social interactions under those circumstances tend to be loose and free of political rules. Economics (resource extraction, distribution and consumption) tends to take precedence over politics (access to, and control over, resources).
There is always some variability in natural plant resource abundance and predictability. Seasons have an impact as well as local rainfall. Plants may produce an abundant "crop" one year and be sparse in another. But the relative abundance also tend to vary from location to location. When plants matured and were relatively abundant in one location, social groups nearby would hear about this and fission off members who would fuse with the people exploiting the relatively abundant resource. This usually occurs through kinship or friendship lines. The size and make-up of groups was changing as this process occurred. The lack of competition for resources allowed for non-threatening change. People did not belong to a polity as opposed to other competing polities, rather they adjusted their size and membership in relationship to the availability of resources.
Middle Archaic Period
This is the period when groups began to settle into the patterns that were found when Euro-Americans entered the Oregon territory.
The time period covered is roughly from 6000 years ago to about 2000 years ago. During the early part of this period, the climate reached its maximum warmth and dryness of the Altithermal which tapered off into the Medithermal (like modern conditions) about 4000 years ago. The dry lake beds of southeastern Oregon developed extensive dune fields from blowing sands.
The first 2000 years of the Medithermal were cooler and wetter than modern conditions and is sometimes called the Little Ice Age. Some think the glaciers of the Rockies, Cascades and Sierras that had melted during the Altithermal, reformed during the Little Ice Age, and are melting again under current conditions. There was a period of increased rainfall and snow that refilled many of the lakes in the basins of southeastern Oregon, and then started an evaporation cycle towards modern conditions again. Sea levels stabilized at about current levels around 5000 years ago with modern coastal habitats associated with cliffs and headlands.
Plants and animals moved up and down the mountains as temperatures fluctuated, but forms are all modern. The relative percentages of plants and animals fluctuated as well. For example, boreal trees expanded and shrank as temperatures opened up and then closed habitats. Timber lines moved up and down, allowing more alpine species at times. There was probably more grass in the sagebrush zones than today. Sage moved through cycles of expansion and fire related contraction. Coastal habitats increased in productivity when the sea level rise impinged on the rocky habitats, cliffs and headlands of previous interglacial high stands. The deep river valleys filled with sediments, producing richer estuary ecologies as well.
Around 3500 to 3000 years ago, systematic burning of climax vegetation by native peoples to reduce the biomass tied up in cellulose began in the lower elevations. This increased plant diversity in valued floral resources.
In general, projectile points become smaller and more regionally diverse. the atlatl and dart was still the primary hunting weapon. In the Great Basin area, two early sites in Lake County, dating around 8000-9000 years ago, contained large pits full of rabbit bone and charcoal. This is evidence for communal rabbit-net drives, a form of hunting in the basins found among the Paiute at contact!
Fishing technology erupts into the archaeological record during this period, but it is unknown how much of this is a sampling problem. Most fishing technology is fragile, and easily destroyed by the ravages of time. Plant processing tools also increase in relative abundance, but again, sampling is a potential problem. Strong evidence for exploitation of camas, acorns, wapato, bitterroot, and other plants known from contact period peoples appears throughout the archaeological record.
The most interesting change in this period is the first evidence for houses and evidence for increasing sedentism. Given limited evidence from around the world, houses were probably used at much earlier periods as well, but will be hard to find. Increasing use of houses with expanding population densities and technologies relating to storage mark this period. Shell middens appear all along the coast as groups took advantage of the expansion of shellfish into rocky habitat created by stabilized and modern sea levels. This period also marks increasing regionalism in styles associated with the manufacture of artifacts and in art or religious objects. There are the clear beginnings of a NW style, a Plateau style, a Basin style and styles associated with the California region. Trade networks become more complex. Many more localized tool-rock sources are exploited, and trade increases in its distances and in scope. The beginnings of the Columbia River trade network are developed as a natural outgrowth of the river system and its intrusion through many different ecological regions deep into the continental interior. The first burial complexes (cemeteries) also appear in the archaeological record.
Kathryn Toepel excavated the Flanagan Site as part of a University of Oregon field school in the 1970's. The site contained both Middle and Late Archaic occupations. Kathy named a Middle Archaic phase after the site: the Flanagan Phase, dated between 4000 and 2000 years ago. The projectile points for this period are dart points with wide necks. The site contained choppers and earth ovens. Kathy suggests that the choppers were used for getting the wood burned in the ovens for cooking camas. The ground stone tools suggests acorn gathering and processing was also important during this period when oak reached its maximum expansion in the Willamette Valley. A range of artifacts types came out of the excavations: projectile points, both dart and arrow, knives, drills, bifaces, scrapers, gravers, spokeshaves, utilized flakes, cores, abraders, choppers, hammers, mortars, pestles, milling stones, edge-ground cobbles and celts. She felt the site was a summer and fall base camp from which hunting, plant gathering, food processing, wood working, hide processing and tool manufacturing went on.
Middle Archaic (6000-2000 BP)in The Willamette Valley
The following basins are organized from north to south.
Pettigrew (1981) tested a series of sites in the Sauvie Island area for his doctoral dissertation. His primary purpose was to establish a chronological sequence for the Portland Basin and to develop a descriptive system for the diagnostic artifacts. He was convinced that this was "one of the major centers of prehistoric development on the Northwest Coast" (1981: 1). He selected ten sites to test, but was able to excavate at seven. The Merrybell Site (35MU9) is the type-site for the Merrybell Phase. The site produced C14 dates of 850 ±90, 870 ±90, 2070 ±70, 2850 ±85, 2850 ±95, and 2880 ±155. The projectile points were generally small dart points with medium to wide neck widths. The Phase is characterized by its broad-necked dart points, stemmed drills, flaked cylindrical bipoints, flaked crescents and perforated ground stone pendants. The Phase was dated to 2600 BP - 1800 BP. Other artifacts included burins, cores, celts, abrading stones, atlatl weights, netsinkers, hammerstones, pestles, anvils, antler wedges, antler flakers, ground and stone pendants.
On the Washington side of the river, 45CL31 on Vancouver Lake produced a C14 date of about 3400 BP in association with Cascade-like points and cobble tools.
Jenkins & Churchill (1984) tested the Fox Bug Site (35MA48) on the Detroit Ranger District just south of the Clackamas County line. The site is at an elevation of 3800 feet. The 18 test units (mostly 50 x 50 cm) revealed a disturbed site. The three broad necked points and lanceolate biface suggested the Middle Archaic date. Also in 1984, Winthrop & Gray tested the Hugh Creek Site (35CL61) at an elevation of 2400 feet. The site contained broad necked darts. It was tested with a block of 32 1 x 1 meter units and three 1 x 2 meter units. Only 1095 pieces of debitage and 57 worked tools were recovered. The points consisted of three ovate and one side-notched dart.
Woodward (1987) excavated a mixed Middle and Late Archaic site at the proposed location for the Clackamas Historical Society Museum near Willamette Falls in shallow soils scattered in basalt bedrock. The projectile points were a mix of narrow-necked arrowheads and medium necked dart points similar to the Flanagan Phase proposed by Toepel. The upper strata suggested many hot fires and 60 carbonized huckleberries. A shard of mid 19th century glass and a bugle type glass bead were recovered from sampling the burned matrix. A crevice produced a basalt spall cluster (40 spalls and a hammerstone). A bedrock milling stone was identified from a 130 cm circular smoothed area. A possible petroglyph was also noted where a series of mechanical looking deep parallel scratches occurred on basalt bedrock. A possible second glyph consisting of a circle and wavy line were also reported. Four lanceolate bifaces were listed as knives. Three gravers and eight end-scrapers rounded out the small tools. Three large basalt flakes were interpreted to be hide scrapers and there was one hammerstone. Twelve fragments of "hemtite (SIC) red-orange ochre and 2 fragments of a cream-white clay" were reported as well. The upper layer contained 12 fragments of black bottle glass from high-kick mid-19th century bottles. The illustrations included pumice abraders, a pestle, and a girdled net weight.
Wilt and Fagan (1996) with Archaeological Investigations Northwest tested four sites for the Federal Highway Administration along the Clackamas River near Fish Creek Mountain. While only one dart fragment was recovered from one site, obsidian hydration suggested the sites were occupied from 4350-3650 BP and around 2000 BP. A scraper fragment was sourced to the Clackamas River source and flakes were sourced to Obsidian Cliffs, Devil Point and Newberry Crater. Some material was provisionally sourced to Whitewater Ridge. The sourcing to eastern and central Oregon quarries suggests a trade network down the Deschutes to the Columbia then down the Columbia to the Portland Basin and then up into the Clackamas River drainage.
The Middle Archaic sites illustrate the changes in adaptations over time. As populations grew, the human groups learned about the relative abundance, predictability and activity of the drainage ecology. The larger numbers meant adaptive strategies that included more ways of dealing with non-cooperating other human beings who were in direct or indirect competition for the necessary and sufficient resources needed for survival. Mapping out access to, and control over, resources would have been more complex and demanding. Groups would be required to use resources that were no longer the cream of the crop. Lesser sources would have taken on greater significance and diversification in resources occurred. Technological changes involving production, distribution, storage and consumption efficiency took place. This is the time of full adaptation to the natural limits of the non-managed environment. At some time in this period, groups reached that limit and had to change their fundamental adaptation.
The change was the introduction of fire to decrease organics tied up in non-food (cellulose) and increase the plants, and animals dependent on those plants, that were of greater value to human groups. Fire became a tool to manage the environment.
Ellis and Fagan (1990: 10) suggest that this was a time of major population growth linked to the start of environmental change through burning. Populations groups were increasing in size and developing the winter village pattern associated with storage of foods harvested throughout the year. During the spring and summer, groups would fragment into harvesting parties at resource specific locations. There was a trend from broad spectrum forays into narrow spectrum gathering of select staples that could be processed for winter storage.
Ellis & Forgeng (1998) tested sites 35WN22 and 35WN45 with three 1 x 1's for a sewer line project. Five artifacts were sourced from 35WN22 to Glass Buttes, Massacre Lake/Guano Valley and Dog Hill from eastern Oregon! A dart fragment was recovered from 35WN45. The obsidian from this site was traced to Obsidian Cliffs, Inman Creek and Whitewater Ridge (far NE are in the Blue Mountains!). Two charred acorn nut fragments were also recovered.
No archaeological data at this time.
Pettigrew (1980) tested a number of site in the Hager's Grove area within the Mill Creek drainage and the southern edge of the Pudding River basin. Site 35MA7 was tested and 21 cubic meters was sifted through screens. A series of pits, hearths, rock clusters (camas oven remnants) and living floors were revealed. Nearby site 35MA9 also had a living floor. On this floor were 4 points, 11 flaked cobbles, 4 percussors, 1 core, 2 flaked unifaces, 1 graver, 1 inverted mortar (to keep rain water out and winter freeze), 1 notched cobble, 1 pestle and a boulder anvil. At site MA9 the broad necked dart points were mostly made from CCS while the arrow points were obsidian. Charred organics included camas bulbs, hazelnuts, onion, acorn, and cherry. There were more arrow points than dart points, and Pettigrew (1980: 57) speculated that either improved hunting success or the ease and relative quickness in making arrow over darts may explain this difference. It may simply be a sampling issue or relative amounts of time or changing population densities with population growth. C14 dates for 35MA9: 400 ±90, 1140 ±90, 1220 ±110, 3740 ±260; and 35MA7: 1190 ±120, 2680 ±150, 2850 ±140, 2870 ±130 and 3800 ±140. This indicates there was an early (Middle Archaic) and later occupation at MA9 (Late Archaic) with a period on non-occupation while the occupation at MA7 was at the end of the Middle Archaic.
"Hager's Grove appears to have been used as a warm season camp, and probably no more than several families used the site areas at any given time. Camas was clearly a very important resource for the people who used the Site MA 7 after 1000 B.C., since many charred camas bulbs were found associated with several features interpreted to be earth ovens. No camas bulbs and no certain earth ovens were found in the earlier component of Site MA 7 (2000-1500 B.C.), nor at Site MA 9, whose major occupation took place during Period A (2000-1500 B.C.). It appears probable on this basis that camas was not a major resource collected from these two sites during Period A. The most likely explanation for this seems to be that camas was not as abundant or for some other reason not as available during Period A as it was later. It is possible that the environment surrounding the grove at that time was too dry to support a large camas prairie, since camas typically grows in moist ground. It must be considered a possibility that period A at Hager's Grove was during the later part of the Oak Maximum of mid-postglacial time, proposed for the Willamette Valley by Heusser (1965) and Hansen (1947). During the Oak Maximum the climate appears to have been drier than at present, and camas may not have been as widespread as it was after 1000 B.C. It is interesting in this regard that the Luckiamute Hearth, in which was found a number of charred acorns, is dated to 3300 B.C., in the middle of the proposed Oak Maximum" (Pettigrew 1980: 75).
Since systematic burning in the Valley may have begun about 1500 BC, the expansion of Prairie, and therefore camas habitat, may also have played a role as well. "Some time prior to 900 B.C. the site area of Site MA 7 became exposed, perhaps through the annual burning of the valley floor by the local people, and the area was used as a warm-season campsite by people who utilized the area in large part to collect camas, which was by that time more abundant, owing to the increased moisture content of the soil" (Pettigrew 1980: 76). The fluctuation in occupation may have related to over exploitation of local wood supplies along the Mill Creek gallery forests, and when the trees returned in sufficient quantities, the sites were re-occupied for camas processing.
Reckendorf and Parsons (1966) reported on the C14 date of 5250 ±270 on the Luckiamute Hearth, exposed in a cutbank of the Luckiamute River near its junction with the Little Luckiamute River. The hearth was 6 meters above the river and 1.5 meters below the present surface. The hearth was lined with sandstone, siltstone, basalt and quartz stones. Five charred acorns were found in the hearth. At the time it was recorded, it was the earliest hard date for occupation in the Willamette Valley.
Smith and Baxter (1996) tested two sites on the lower Luckiamute River with their field school in 1990. A single hearth feature was encountered which gave C14 dates of 4800±80 and 4810±90 BP. A single lanceolate dart point was recovered near the hearth. Dorothy Freidel looked at the geomorphology and concluded the sites were in the Champoeg surface formed at least 34,000 years ago. The obsidian was sourced to Inman Creek.
Jenkins & Churchill (1987) tested two sites at an elevation of 4000 feet near the dividing line between the High Cascades and the Western cascades physiographic zones. While both sites today are in a snow zone that would limit activities to late Spring through fall, the time of occupation corresponded to the maximum warmth and dryness period of the Medithermal. The location of the sites precludes their use from a local base camp at lower elevations. Their date suggests that lower elevation activities were carried out at higher elevations during the period of maximum warmth and dryness. Both sites are located next to marshy areas today, suggesting water was an important element in site location.
Spencer (1988) tested the Squaw Mountain North II site (35LIN353) at an elevation of 4150 feet and the Canyon Owl Confluence site (35LIN336) at an elevation of 1760 feet. He placed 24 50x50 cm shovel probes and 4 1x1 meter test pits into the former and dated it as Middle Archaic on obsidian hydration, with the material coming from Obsidian Cliffs. The latter site was tested the same way and contained five foliate dart points with hydration rinds that also were in the Middle Archaic range, with the material coming from Devil Point and Obsidian Cliffs. He notes that Lindberg-Muir had proposed that the further sites are from specific quarry sources, the more likely it is to contain obsidian from other sources.
Catherine Lindberg-Muir. in her 1988 Masters thesis at OSU, sourced six sites in this drainage, and she found six sources: 90% from Obsidian Cliffs, followed by Devil Point and four unknown sources (at that time). Two of the unknowns turned out to be Inman and Clackamas River. Yukwah's (35LIN118) sample of 30 consisted of 28 Obsidian Cliffs, 1 Devil Point, and 1 Inman. Soda Fork Way Trail II's (35LIN230) sample of 10 were all from Obsidian Cliffs. Lost Prairie West's (35LIN322) sample of 18 were all Obsidian Cliffs. Monument Peak Trail One's (35LIN342) sample of 10 were 5 from Obsidian Cliffs, 3 from Devil Point, 1 from Clackamas River and one from Unknown B. Dane Saddle's (35LIN32) sample of 12 were all Obsidian Cliffs. Tombstone Summit's (35LIN341) sample of 10 were 8 from Obsidian Cliffs and Unknown A.
Nilsson (1989) tested the Bear Saddle site (35LIN301) in a saddle on a ridge at an elevation of 4320 feet. She noted that Carl Davis (1987) had characterized such "lithic scatters" as:
- being located above 3000 feet in mountainous terrain;
- located in the Pacific Silver fir zone on ridge lines and terraces;
- ranging from 2 to : an acre in size;
- the bulk of them are shallow (less than 30 cm);
- they are found in clay loams developed from tuff and breccia;
- they are mostly obsidian flakes and flaked tools;
- their tool density is low;
- a limited range of chipped tools are present;
- all phases of lithic reduction are present;
- there are no features;
- they range from Early to Late Archaic in age;
- all have been altered to some degree by natural and human actions.
The sampling data upon which these conclusions were based was minimal. Most sites have never been tested to the sample levels necessary for determinations of data redundancy. The samples upon which these conclusions were based were not elucidated. Many of the conclusions can only be tested with large numbers of sites tested to the 40-60% levels of data recovery, and perhaps more. For example, the work at Bear Saddle (testing this concept) recovered 12,385 flakes from 20 contiguous 1x1 meter units. This material was taken from a site covering 1500 square meters, a sample size of only 1.3% of the surface area. No features or organic remains were recovered within this sample. Tools included bifaces, projectile points, edge-modified flakes and drills. The projectile points were all Middle Archaic dart points. Cores were recovered as well. Eighty-six percent of the material was obsidian. Samples of the obsidian indicated Obsidian Cliffs, Devil Point and Newberry Crater as the sources. Given the sampling, there was no basis for testing the Davis "characterization" of such sites.
In my opinion. there is a difference between cumulative experience and statistical sampling. Sampling each site at very low rates that never hit a feature does not mean one can add up that experience (add up the samples) to conclude that the sites do not contain features. Sampling is not additive in this way.
Clayton Lebow (Applied Earthworks) tested four sites along Oak Creek of the Calapooia River in 1996 for a private development project. Three of the four sites were on the Calapooia/Senecal (Pleistocene) land surface and one was on the Ingram surface, dated to less than 5000 BP. Sourcing the sites on the older land form indicated the obsidian came from Inman Creek and hydration gave two peaks: one around 5100 BP and another at 2250 BP. The latter peak matches a C14 date of 2450±90 from charcoal found in an earth oven but two other dates from the same feature were 4490±80 and 4390±60 (Lebow et al. 1996).
The presence of two hydration peaks was thought to be evidence for a two component site on a land surface theorized to be a non-depositional setting. Local processes related to flood events is thought to be the explanation. The older component suggested a temporary campsite. The FCR oven with charred camas bulbs related to the upper component, also a temporary camp, but where camas was processed. Lebow felt the change to camas processing in the upper component was evidence for plant use intensification when the area had been burned over and changed from closed pine and fir forest to open grass/oak savanna.
The site on the Ingram land surface was occupied about 2450-2250 BP and contained a roasting oven and a pestle fragment, also indicative of plant processing.
Draper (et al. 1993) with 4D CRM, tested the Lundsford Saddle site at an elevation of 3550 feet. The site contained foliate or willow leaf and side-notched dart points he ascribed to the Baby Rock Phase of the Middle Archaic between 5000-3000 BP. Obsidian from the site was traced to Obsidian Cliffs, Devil Peak, Newberry Crater and an unknown quarry. He felt the location in the saddle allowed hunter-gatherers to access a wide range of animals and plants, but only in the late spring, summer and early fall before the snows. The Middle Archaic is a time of maximum warmth and dryness. Draper feels the site:
"was used intermittently by transient groups who periodically stopped to repair and replace broken equipment (i.e., 'gearing up') enroute to hunting tracts located elsewhere on the landscape. As such, the site can be perceived as stop-over locus along a travel corridor between areas containing more abundant and/or predictable game resources. owing to the abundance of Obsidian Cliffs and nearby Devil Peak obsidian in the sample, the site may also have been part of a larger trade or exchange network" (Draper et al. 1993: 5.41-42).
Flenniken, Ozbun and Markos (1990b) tested the Swamp Peak Way Trail Site (35LIN373) on the Sweet Home Ranger District. They dug 20 50x50 cm shovel probes and two blocks consisting of three 1x1 meter units. They only recovered 877 fragments of debitage and 13 formed artifacts. The projectile points were dart forms. Fifteen obsidian fragments were submitted for x-ray fluorescence and traced to Devil Point (12), and the rest (3) from Obsidian Cliffs.
Marion Creek Site (35MA114) was tested as part of the NW Pipeline Project down the Willamette Valley. The site is on the Winkle surface and the original survey recorded a maul, seven pestles, three pestle fragments, two stone mortar bowls and six fragments, a stone wedge, a digging stick handle and 28 projectile points (lanceolate, broad-necked and side-notched dart points). Three trenches and four blocks were excavated in 1992. A concentration of charcoal and oxidized soil in a pit shape was associated with Douglas fir, Western redcedar and hemlock. A second feature was a dense concentration of basalt cobbles in a pit and its last use dated to 1160 ±60 (790 AD). Feature 3 was concentrated FCR with no definable shape.
Thirty-six projectile points were recovered from the excavations. Fourteen were classed as arrowheads and the rest were fragments or dart points:
Seven of the arrow points and two of the dart points retained patches of remnant ventral surfaces, evidence of their manufacture from flake blanks. Three of the dart points were manufactured from bifacial blanks. Two of the arrow points and four of the dart points exhibited differential luster suggesting intentional heat treatment (Fagan et al 1996: 2-42).
Most of the obsidian came from Inman Creek followed by Obsidian Cliffs, Devil Point and Newberry Crater. Based on hydration rims, the site was thought to have been occupied about 3000 BC followed by sporadic intensive occupations between 1000 and 250 BC.
They also tested the nearby Two Gun Shot Site (35MA105). Three dart points had been noted during the survey, four dart points were recovered from the surface and another 21 points, preforms, or fragments from the excavations. Four were arrowheads, nine were darts, four indeterminate, two dart preforms, one arrow preform and one indeterminate preform. A drilled siltstone pendant was also recovered.
The obsidian was source to Inman Creek and Devil Point. Hydration gave a Middle to Late Archaic date of occupation. The total assemblage suggested the site was occupied from 1050 BC to AD 775.
Long Tom Basin
The type site for the Flanagan Phase is the Flanagan site reported by Kathryn Toepel (1985). The work was done as part of the University of Oregon field school from 1976-78. Seven features were recorded: one post hole (1720±100), six rock hearths (460 ±80, 980 ±100, 3300 ±220), and one camas oven (3230 ±150). Based on the range of artifacts (hunting, plant processing, wood working, hide scraping, and mfg of tools) the site was interpreted as a base camp occupied in late summer and fall (Toepel 1985:73). Both charred acorns and camas were recovered.
The projectile points were broken into classes by neck width. Narrow necked points had neck widths less than 7 mm, moderate were from 7-9 mm and broad necked were >9 mm. Narrow necked points are considered to be arrow points used with the bow. Medium and broad necked points are probably dart points used with the atlatl. The small stemless and narrow necked points were concentrated in the upper levels of the site while the moderate broad necked one were in the middle strata and the heavy broad necked were most often in the lowest layers. The fourteen C14 dates ranged from 460 ±80 to 5750 ±200.
The lowest levels ranged from 6000-4000 BP and contained just over half of the ground stone found in the site, but no ovens or hearths. The middle component from 4000-2000 BP contained the remainder of the ground stone along with hearths and the camas oven. The top layers dated from 2000-200 BP and is marked by a change to bow and arrow technology as well as stone hearths.
The obsidian was traced to the following sources: Inman, North Sister, Tucker Hill, Newberry, and Coglan Buttes.
She notes that "The distribution of aboriginal groups in the Willamette Valley conformed to a definite pattern in which each "tribe" or "band" occupied its own valley or basin formed by one of the larger tributaries on the Willamette River" (Zenk 1976:18; Toepel 1985: 27).
"The correspondence of the hydrologic basins with the Kalapuyan languages and the sub-basins with Kalapuyan dialects strongly suggests that it may be profitable to use the basin/sub-basin distinctions as natural units of study in Willamette Valley archaeology. Archaeological patterns - functional, temporal, and cultural (or ethnic) - mat be defined in each sub-basin, and these patterns can be compared with those observed in other sub-basins in developing a picture of the archaeological record within the region as a whole" (Toepel 1985: 29).
The type site for the Lingo Phase (35LA29) is the Lingo Mound dug by the Kalapooya Chapter of the Oregon Archaeological Society from 1962-1966 and then in 1965-1966 by a University of Oregon field class. Three shallow basin fire pits were uncovered with associated charred camas bulbs. Ten burials were discovered in trenches dug through the site. The stemmed broad-necked points represent the end of the Middle Archaic period. Additional artifacts included drills, scrapers, gravers, used flakes, pestles, and beads. The site was interpreted as a seasonal camas camp used during summer months.
Beginning about 7700 years ago and ending about 4400 years ago, there was a period of erosion (Freidel el al. 1989: 97) that destroyed many of the archaeological sites. From 4400 to 3900 years ago, there was a period of rapid deposition and erosion events. From 3900-2100 years ago, deposition predominated.
In the excavations at the Hurd site (35LA44) in 1969-70, a large shallow semi-elliptical pit feature was interpreted to be a house pit. Two large areas of disturbance interrupted the rim of the feature. The northern rim was dug about 35 cm deep and the southern rim was just detectible as a slight rise, producing a level floor on the sloping ground. The northern rim segment was 8.2 meters long and the southern one 2.2 meters long giving a maximum north-south feature diameter 7.5 meters by 5.4 meters from east to west. near the center was a cobble circle around a shallow saucer shaped pit about 22 cm deep and 69 cm across. Charcoal from this hearth gave a date of 2800 ±110 years (Gak-2660) and 2820 ±230. Possible post holes were recorded around the northern region of the house pit. Three form a line across the north end and three are found on the north rim of the pit as well as one on the south rim (White 1975:148-150).
Silvermoon (1985) reported the work at the J&K site (35LA254) at an elevation of 1475 feet on the floodplain of the river. A total of 153 tools were recovered, including 107 utilized flakes, 24 flake unifaces, 3 bifaces, 3 projectile points, several cores and a number of biface fragments. The points consisted of two leaf-shaped and one side-notched dart points.
Southard (1986) put some test pits into the Mill Creek site (35LA365) at an elevation of 800 feet near the junction of Mill Creek and the Mohawk River. He dug block units totaling 8.1 cubic meters and recovered 60 tools. His points were Early Archaic cascade-looking or Middle Archaic side-notched dart points.
Winthrop & Gray (1989) tested two sites on the Blue River Ranger District (35LA325 & 35LA857). Both sites were tested with twenty 50x50 cm shovel probes and four 1x1 test pits (yielding roughly 5.5 cubic meters in both cases). A total of 40 formed tools were recovered from 35LA325. The projectile points were broad necked dart points. Site 35LA857 yielded 55 tools and 941 fragments of debitage. The projectile points also were from darts. Both appear to have been small temporary camps at elevations between 1200-1300 feet.
Ozbun (et al. 1990) of Lithic Analysts excavated the J&K Enterprises site (35LA254) for the Willamette National Forest. Neither Ozbun nor Flenniken believe that projectile point style has any meaning as points can be reused, changing their morphological characteristics. For this reason, there reports do not contain illustrations of the material culture, making independent evaluation essentially impossible. Simple descriptive data is not present. They did note the absence of arrow points indicated a probably Middle Archaic occupation for the site.
Harvey Rice did a catchment analysis for the J&K site (the most likely resources exploited by the people using this site). A half mile radius was used based on Snyder's dissertation on land use patterns in the Cascade range. Most of the area was climax forest, with most of the biomass tied up in trees. Rice could find no concentrated plant resources that would have been of value. He thought the site was merely a stop-over place for groups traveling to higher elevations to look for bear grass, huckleberries and obsidian at Three Sisters. He listed some complex uses for plants that indicate how varied Indian technology was:
wood: fuel, halibut & dogfish hooks; bark: canoes; pitch: treatment for bows, dye, finishing agent; boughs: bedding, floor covering, berry baskets; branches: costume material; needles: baby powder, scent and deodorant, boiled for cough medicine.
Big Leaf Maple
wood: carving, sweat lodge frame, bows, fuel, canoe paddles, cradleboards, dishes, spoons, smoking salmon; bark: rope, baskets, fans, beaters, cambium eaten; leaves: cover food, cooking, drying berries, lining containers, rags, flavoring food; boughs: temporary shelters.
wood: dishes, spoons, platters, canoe paddles, canoe bailers, cradles, smoking salmon, fuel; bark: dye for camas, lining for storing elderberries, tea for colds; boughs: shelter frames; sap: food, cambium eaten; cones: eaten for dysentery.
wood: baskets; leaves: heated for colds. Patricia Whereat Phillips (e-mail personal communication) indicated that the roots were gathered during hunting trips.
wood: gambling disks, salmon harpoon foreshafts, pounding bracken ferns, charcoal for tattooing, bows, arrows, implement handles, combs; bark: tanning agent, dye, laxative, emetic.
wood: spoons; bark: dye; roots: dye for baskets, arrows, rope, edging for baskets, cradleboards; nuts: mixed with camas and tarweed, dried as winter food.
Western White Pine
wood: mat needles; bark: storage baskets, small canoes; pitch: waterproofing, medicine for stomach and cuts: pine nuts: food.
wood: palisade posts, fire drill hearth, dugout canoe, cradleboards, fuel, fishing weirs; bark: house covering, soap, food containers, lining storage pits, cambium eaten: roots: rope, lashings; gum: glue, paint; cotton: bedding, insulation, fire starting.
wood: fuel, torches, harpoon shafts, salmon spears, dip net handles, tepee poles, drying scaffolds, smoking racks, gaffhook poles, canoe cross braces, river poles, dugout canoes, fishtraps, snow shoes, coffins, bows; bark: fuel, dye; boughs: used in sweat lodge, temporary shelters, shading smoking fish and berries, bedding, padding; saplings: dipnet frames; pitch: caulking, mixed with sand for fishskin moccasin soles; knots: halibut and cod hooks; needles: yellow dye.
wood: chisel handles; bark: dye, laxative, stomach troubles, purgative; berries: food.
wood: bows, arrows, harpoons, canoe paddles, clubs, digging sticks, wedges, boxes, drum frames, spring pole traps, adze handles, prying sticks, mat needles, awls, dishes, knife handles, spoons, dowels, pegs, bark scrapers, canoe bailers, fire tongs, combs, gambling sticks, snowshoe frames; twigs: depilatory; branches: sea urchin spear, hide hammock support; leaves: bathe babies, internal injury, smoked in place of tobacco; fruit: eaten in small quantities.
Port Orford Cedar
According to Patricia Whereat Phillips (e-mail personal communication) the tree was used to make canoes. One Coos woman (Lottie Evanoff, daughter of Chief Doloos Jackson) mentioned that the canoes were heavier and more difficult to dock, but that some people preferred them. The wood was believed to be denser than red cedar.
wood: boxes, dugouts, house posts and planks, cradles, arrows, spindles, herring rakes, fire drill hearth, totem poles & mortuary posts, dishes, harpoon shafts, fish spreaders, dipnet hooks, fish clubs, masks, rattles, coffins, canoe bailers, drums, spirit whistles, paddles, fuel; bark: padding, sanitary pads, headbands, towels, clothing, plaited platters, mats, sails, line cooking pits; boughs: rope, lashings, basketry, fish weirs, bedding, green dye, hats, baby carriers, sew bark canoes; bark: roofing, sweathouse covers, canoes, food cache linings, separate foods during storage, baskets, bags, hats, mats, capes, blankets, cambium eaten; roots: baskets, box joints, lashings, nets, hats, mats.
Patricia Whereat Phillips also noted that some coastal groups made canoes out of redwood logs that washed up on the beaches. The Milluk Coos believed that a power had placed the logs on the beach and they should be left alone. "There is a story of, in 19th century, an Athabaskan man named Silas Tichenor who lived at Coos Bay who made a canoe from a beached red wood log. No Milluk wd get in that canoe. After Silas died - some time if 1800's - the Milluks sunk his canoe so no one wd use it again." (E-mail personal communication)
baskets, fish traps, hold roof planks, cradle swings, salmon tongs, fuel, spoons, bows, snowshoes, slat armor vests, arrows, implement handles, dipnet frames.
gambling disks, halibut line rigging, arrows, digging sticks, seed beaters, canoe frames, cooking basket frames, shields, euchalon rakes, shelter frames, drying mats, berries eaten.
bark soap, leaf wraps, leaf liners for steaming foods, berry stain, berries eaten.
stems hollowed as pipe stems, brooms, arrows, shellfish drying racks, digging sticks, gaming sticks, berries eaten dried.
Knowing uses for plants was technical knowledge required by every adult member of the society. This is just a small part of the many plants used by so called "primitive" peoples. There was nothing "primitive" about such knowledge. It was appropriate to its time and place and very complex and sophisticated.
Oetting (1994) tested seven sites on the Blue River Reservoir and six sites on the Cougar Reservoir. The sites with diagnostic points were all Middle Archaic in date containing a mix of Northern Side-notched, Elko and miscellaneous broad necked dart points. Also recovered were drills and scrapers. The reservoirs are at an elevation of about 1300 feet with the surrounding mountains at 3000-4000 feet.
Spencer (1989) tested the Dale Beam Site (35LA793) with a C14 date of 3110 ±90. The site is at an elevation of 3260 feet. Both foliate dart points and a single corner-notched dart/arrow head (7 mm neck width) were recovered. He dug 30 50x50 shovel probes and 5 1x1 meter test pits. Obsidian was sourced to Obsidian Cliffs (9) and Newberry (1). he notes the recovery of a "single microcore"... which in other reports would be called a thumbnail scraper. He does not report any microblades.
Coast Fork Basin
Bland (1989) reported on testing site 35LA265 for the Corps of Engineers. Twenty-none 50x50 cm shovel probes and three 1x1 meter pest pits were placed into the site. It had been heavily impacted by erosion, revealing the large number of tools.
The three excavated projectile points were each different. One appears to be a resharpened leaf-shaped point, the second a corner notched point, and the third a broad necked side or corner notched specimen. All showed evidence of fractures during use and had been removed from darts. The three points are illustrated with scrapers, bifaces, worked flakes, choppers and a core.
The obsidian came from Inman Creek (6), Obsidian Cliffs (1), Newberry (1), Silver Lake (1), and Spodue Mtn (1). This suggests trade routes primarily associated with the Eugene area to the north. Most of the stone was a locally available chert (80%). The obsidian hydration suggested a Late Archaic occupation while the points suggested Middle Archaic. Earlier surface collections did include a small narrow-necked arrow point.
Middle Fork Basin
The type-site for the Baby Rock Phase is Baby Rockshelter (35LA53) excavated in 1970 by the University of Oregon (Olsen 1975). The overhang sheltered about 85 square meters, and it had been badly vandalized at the time of data recovery. Red pictographs including a horse and rider were on the shelter wall, indicating contact period occupation.
Three 2 meter test pits were excavated into areas with the least looting disturbance. Four natural layers were distinguished. The number of dart points suggest a Middle Archaic date for the primary occupational layers which was thought to be contemporaneous with Cascadia Cave. Mazama ash was found in stratum C. Black-tailed deer hunting was the chief economic activity based on the presence of faunal evidence, but the presence of manos, pestles and stone bowls is evidence for plant gathering. This phase is also represented by the upper levels of Rigdon's Horse Pasture Cave (see Late Archaic below), and Cascadia Cave (see Santiam section). The stemmed broad-necked dart points, knives, end and side scrapers, drills, gravers, choppers, manos, pestles and stone bowls characterize this phase (Beckham et al. 1981: 168).
The Packard Creek site (35LA475) was tested by the Rigdon Ranger District (Heid 1987) with shovel probes and test pits. The site was radiocarbon dated at 2565 ±85 and 1780 ±80, at the end of the Middle Archaic and the beginning of the Late Archaic. The obsidian was traced to Obsidian Cliffs, Devil Point, Inman, and Quartz Mtn. There was no formal report, but the projectile points appear to have been a mix of darts and arrowheads.
Winkler (1988) tested and evaluated The Diamond Lil (35LA807) site, located in a saddle on a terminal ridge overlooking the river at an elevation of 1810 feet. While 40 50x50 cm shovel probes and two 1x1 meter units were dug, this was still only 0.3% of the surface area of the site. The work recovered 1489 flakes and 58 tools, a high proportion of tools to debitage. Twelve of the tools were projectile points... a mix of later Middle Archaic dart points and Late Archaic arrowheads.
Churchill and Jenkins (1989) reported on excavations of the Pepper Rockshelter dating to the Baby Rock Phase. While only excavating 1.2 cubic meters of cultural fill, they recovered 128 tools, 7,575 pieces of debitage, 61 floral specimens, and 5,587 fragments of bone! The tools included 14 projectile points, 66 bifaces, 6 scrapers, 2 unifaces, 14 worked flakes, 17 cores, 2 metates, 2 hammerstones, 1 anvil, and 2 probable pestle fragments. Dates included 205 ±60, 980 ±50, 1990 ±90, and 2440 ±110. Obsidian from the upper levels came from Inman Creek, Spodue Mtn, and Newberry. The lower samples came from Inman Creek, Mckay Butte and Silver Lake/Sycan Marsh. Plants included hazel nut, fir, bitter cherry, dogwood, yew, chokecherry, maple, pine and thimbleberry. Bipolar reduction (as illustrated) was used on small cobbles. Was this a Molala site?
Flenniken, Ozbun and Marcos reported on the testing of the Gate Creek Site (35LA295) on the Rigdon Ranger District of the Willamette National Forest. They placed 21 50x50 cm shovel pits and 2 6x6 meter blocks into the site and recovered only 1,623 pieces of debitage and 26 artifacts. Trace element studies was done on 30 fragments of obsidian: Inman Creek (7), Silver Lake (7), Spodue Mtn (5), Newberry Crater (3), Obsidian Cliffs (2), Unknown A (2), and 1 to Cougar Mtn, McKay Butte, and Unknowns B and C (now known to be Clackamas). Most of the sources are east of the Cascades. The projectile points were dart or lanceolate forms that could have been knives.
Flenniken, Ozbun and Fulkerson tested five sites on the Lowell and Rigdon Districts (1989a). The sites ranged in date from Early through Late Archaic, based on point styles. All of the sites are within the mapped Molala territory closest to Yoncalla and Winefelly Kalapuya lands. All five sites were disturbed lithic scatters. The Tire Creek site (35LA320) at an elevation of 1700 feet covered about two acres surrounding a marshy meadow. Two Early/Middle Archaic dart points were recovered. The obsidian was sourced to Obsidian Cliffs, Newberry Crater, Cougar Mtn and Silver Lake/Sycan Marsh. The Elk Saddle site (35LA444) contained a Late Archaic arrowhead and flakes sourced to Obsidian Cliffs. A pestle fragment was recovered. The Merrill-Exton site (35LA814) contained Early Archaic dart points and Late Archaic arrowheads. Sourcing was to Obsidian Cliffs, Silver Lake/Sycan Marsh and Inman Creek in Kalapuya territory. The Oakridge Spur site (35LA633) had a fragment of an Early Archaic dart point and was sourced to Obsidian Cliffs, Silver Lake/Sycan Marsh and Spodue Mtn. The Lupher's Ridge site (35LA632) dart point was Early Archaic in form and the obsidian came from Obsidian Cliffs and Silver Lake/Sycan Marsh. All five sites were associated with non-forested meadows where useful plants are more abundant. The sourcing suggested extensive trade networks or highly mobile groups accessing the quarries directly.
Baxter (1986b) tested the Colt site (35LA599) and the Saddle site (35LA529), both at an elevation of about 2000 feet on a ridge line adjacent to the Middle Fork. The Colt site artifacts included 69 projectile points: a mix of Middle Archaic dart points and Late Archaic arrowheads. The Desert Side-notched points were all found at the surface. The Gunther points were clustered near the surface as well. Other artifacts included bifaces, drills, gravers, end scrapers, utilized flakes, mortars, pestles, and anvils. The Saddle site yielded 26 classifiable points, unifaces, bifaces, drills, perforators, mortars, pestles, anvil stones, hammerstones and abraders. The obsidian came from Inman, Obsidian Cliffs, Three Sisters, Newberry Crater, Quartz Mtn and McKay Butte.