The most notable exception to the homogeneous availability of raw materials is in the northwest corner of the state, where an exceptionally flat clay and silt plain covers all or parts of nine counties (Figure 1). This is the dry bed of Glacial Lake Agassiz, an area also known as the Red River Valley. Rock of any kind is rare or absent in the lakebottom plain. Cobbles are, however, abundant in beaches of the extinct lake. The lake stabilized at different levels during the several thousand years of its existence, and separate beaches were formed at each level. They presently appear as sand and gravel ridges rising above the surrounding clay plain. Some of the beaches are quite massive and may be traced for a hundred or more miles (Fenton et al. 1983). Locally they contain significant quantities of silicate cobbles, and would have been potentially good lithic raw material sources (e.g., 21RO11, discussed below).

A second exception to the homogeneous availability of raw materials in the Western Resource Region is the apparent lack of Tongue River Silica in the northern part of the region (see Bakken 1985). This distribution is presently hard to explain in terms of geological and glacial history, and needs further examination.

Swan River Chert (SRC) is the most naturally abundant and most commonly used raw material in the Western Resource Region (Bakken 1985, 1993; Campling 1980). SRC is found over a very wide area, from eastern Saskatchewan through southwestern Manitoba, the eastern Dakotas, much of Minnesota and presumably into northern Iowa. This material is extremely variable, especially in color. The "orange peel" texture of fracture surfaces, cloudy yellow color of transmitted light, and microscopic, agate like banding are normally the most useful diagnostic characteristics. The spongy or ropy texture of the cortex is also distinctive, when cortex is present. SRC is a difficult material to work. Cobbles are extremely tough, and even breaking open a raw cobble can be difficult. At some sites, analysis of lithic debitage suggests that whole SRC cobbles -- rather than partly reduced blanks -- were heat treated (e.g., Bakken 1995a). Presumably this was a response to the tenacity of the material, which impedes even the earliest stages of reduction. Once SRC was reduced to the form of a tool, however, it the tool should also have be tough and durable.

Red River Chert (RRC) is the second most commonly utilized local raw material in the region (Bakken 1985, 1993). Distribution ranges from parts of southern Manitoba through the eastern Dakotas, much of Minnesota and presumably into northern Iowa. In Manitoba, RRC would be identified as either Cathead Chert or Selkirk Chert, both of which are known from bedrock outcrops. In Minnesota, where these materials are known only from glacial drift, they are commonly grouped under the broader category of Red River Chert. RRC is usually light in color, varying from white to light brown or light grey. The color is usually homogenous to broadly mottled. Texture and translucently are somewhat variable. Transmitted light normally shows little color. Some pieces of RRC contain fossils, usually in the form of molds. Segments of crinoid stems are the most common type of fossil. There is little to separate RRC from other fine grained, light colored cherts. Happily, no similar materials are known to occur naturally in the Western Resource Region. It appears that RRC is relatively more common in drift than at sites, although relative abundance is difficult to evaluate objectively (see Bakken 1985). This difference may be because many RRC cobbles fracture uncontrollably into small, blocky chunks which are useless for further reduction. Workable RRC cobbles are of variable quality, although there was probably selection for better pieces. These better pieces would be fine grained, homogenous, and generally comparable to the better quality raw materials available in the state.

Quartz is available in glacial drift throughout the state, including the Western Resource Region (Bakken 1985, 1993). It is normally easily identified by the presence of flat fracture plains following the material's crystalline structure. Although quartz is difficult to work, it was utilized extensively at sites in some parts of the region and is therefore included as a primary raw material (Table 1). Based on initial studies, it appears that specialized bipolar reduction techniques were often used with quartz, possibly with the goal of producing expedient flake tools.

Drift derived quartzite found in the Western Resource Region occurs in two general forms: a white to golden yellow form which tends to have larger crystals, and a dull red to dull purple form which tends to be finer grained. Both forms are very brittle and largely useless for standard lithic reduction. The flakes occasionally found at sites are probably detached from quartzite cobbles used as hammerstones.

Rhyolite is less common than SRC or RRC, both naturally and at archaeological sites. The distribution of this material is not well documented. Bedrock sources and associated quarries are known in parts of northwest Ontario (see Lynch and Lovis 1988), in east central Saskatchewan (McDougall 1980), and possibly in intermediate parts of Manitoba. I would expect that this material occurs sporadically in glacial drift throughout the Western Resource Region. Rhyolite may be identified by its greenish grey to (less common) grey color; the presence of patches and streaks of brown to orange brown; and the presence of clear, colorless phenocrysts in a finer grained matrix. It sometimes resembles siltstone. The two may be distinguished by relative opacity; siltstone is always opaque, while rhyolite displays moderate to strong translucency. Rhyolite varies in texture from very fine grained and almost cherty, to more coarse and granular. Its working characteristics vary from marginal to moderate.

An initial survey of raw material resources in the northwestern part of the state (Bakken 1985) described this material but called it argillite (i.e., siltstone). As better information became available, it became apparent that this identification was incorrect. The availability of siltstone in the Western Resource Region remains unevaluated, although it seems unlikely that significant quantities would be available (Bakken 1993). This material is not similar to a brown rhyolite which occurs in northeastern Minnesota. The brown rhyolite has very poor flaking characteristics, and was only rarely utilized (Tony Romano, personal communication 1994).

Tongue River Silica (TRS) is found in most of the Western Resource Region although, as discussed above, it may be absent from the northern parts of the region (Bakken 1985, 1993). TRS has a relatively narrow range of variation and is easily identified. Its natural color is restricted to a fairly narrow range of ocher or yellowish brown. Weathering may produce an orange or reddish discoloration on the surface and along cracks or root molds. Heat treatment is easy to identify because the color changes to orange-red or red (Anderson 1978). Other diagnostic characteristics include a fine grained "sparkle" on fracture surfaces and the presence of hollow root molds. Most TRS is highly opaque; rare pieces are slightly translucent on thin edges. Although the material does not seem to be as tough as SRC, it is probably as difficult to reduce because the fracture is irregular. Heat treatment was regularly used with this material; as with SRC, it was probably applied early in the reduction process. Related but distinctive forms of grey TRS have been identified in North Dakota (Ahler 1977). These do not appear to occur naturally in Minnesota although they may occasionally be found at archaeological sites in the state.

Pebbles of chalcedony, jasper and silicified wood occur in the drift but are not common (Bakken 1985, 1993). They rarely occur in cobbles large enough to facilitate standard patterned reduction techniques. It should be noted that these terms are used here in a descriptive, generic sense. Chalcedony includes any fine grained, translucent to transparent, waxy textured materials which are usually pale brown, gold or colorless. Jasper includes fine grained, waxy textured material which is yellow, yellowish brown, olive, brown or red to reddish brown in color. The color may be solid or mottled. Silicified wood may be distinguished by its visible wood grain. Pieces of this material often fracture along lines between growth rings. The origin and distribution of these materials are not well understood. When materials matching these descriptions are found at sites in the region, some care should be taken in their identification. This is particularly true with chalcedonies and jaspers, especially if patterned tools made of these materials are found. Such pieces may represent the import of raw material from sources outside the region, especially from sources to the west where larger, better quality pieces are available. In some cases it may be possible to specifically identify such nonlocal raw materials.

Although Jasper Taconite (JT) is listed as a "local" raw material for this region, local availability is limited (Bakken 1985, 1993). Within the Western Resource Region, this material occurs only in small amounts and in small nodules which are often too friable to flake. Most or all JT found at sites here probably comes from the Eastern Resource Region where it is more common and of much better quality. This material is discussed in further detail in the section describing the Eastern Resource Region.

Locally available quantities of Knife River Flint (KRF) are also negligible; pieces also tend to be small and of poor quality (Bakken 1985, 1993). Most KRF found at sites here probably come from outside the region, specifically from the primary source area in west central North Dakota (Gregg 1987). An exception may be the reduction of locally procured KRF pebbles by bifacial techniques in order to produce expedient flake tools. KRF is still one of the most abundant raw materials at many archaeological sites in this region. It is not uncommon for up to 30% of the lithic debitage at a site to be KRF, especially at sites near the Red River. At Middle Woodland sites, the percentage may be even higher (Clark 1984). KRF is easily identified by its translucency, coffee brown color and pattern of diffuse inclusions (see Clayton et al. 1970). Occasionally it develops a white patination. Burning makes KRF nearly opaque and pale bluish grey in color. Even when patinated or burned, however, KRF displays the same distinctive brown color of transmitted light and may often be identified by this characteristic.

Other nonlocal and exotic materials will occur in small amounts at sites in the Western Resource Region. They include many raw materials available in other parts of the state, as well as porcellanite, rarely obsidian, grey Tongue River Silica and probably a few other materials which are usually not specifically recognized and therefore not reported in the literature. It is my impression that the Rainy River and Lake of the Wood may have provided a significant route for the arrival of lithic raw materials from the Eastern Resource Region (i.e., Jasper Taconite, Gunflint Silica and Hudson Bay Lowland Chert). These materials seem to occur in noticeably higher concentrations near this route, or at least at its western end near Lake of the Woods. This idea needs to be carefully examined, however, before it is given any interpretive significance.

At least one prehistoric "quarry" site has been studied in the Western Resource Region. Actually, it should probably be called a lithic procurement site since it is not a quarry in the narrower sense of the word. The Greenbush Borrow Pit site, 21RO11 (Peterson 1973), is located on a large Lake Agassiz beach. Cobbles were apparently being taken from the beach ridge and reduced on the spot. Swan River Chert was the primary material extracted and worked, with Red River Chert processed in smaller amounts. The presence of stemmed projectile points suggests a possible Late Paleoindian date for the site.

The views and opinions expressed in this page are strictly those of the page author.
The contents of this page have not been reviewed or approved by the University of Minnesota.