Chapter 5 Results of Analysis

My involvement in this project began in 2002 through the field school excavation activities. At the end of the field season, my wife, Victoria Kline, and I brought the excavated materials into the lab for processing, and repeated the process in 2003. The results of our laboratory efforts were combined with activities and subsequent analyses performed during and after the curation processes. Additionally, activities and discoveries resulting after the presentation of Flemming’s 1999 thesis are also included. These results report on all categories of archaeological evidence with particular foci on the lithic tools, materials, and the debitage and what this tells us of the prehistoric occupational sequences of Lost Valley.



A diverse variety of flaked stone tools were excavated from the Lost Valley field school project over seven years of activity. The smaller tools were made from white milky quartz, obsidian, or other high quality cryptocrystalline silicates, while the larger, heavier tools identified as choppers or scrapers were more likely made from quartzite, the diverse metavolcanic materials, or the various Santiago Peak volcanics. The vast majority of the smaller tools were found in a fragmented state.

There were surprisingly few utilized flakes recorded in the database in relation to the size of the collection and the number of units excavated. As demonstrated in Table 5, there were only 122 total flake stone tools, variously identified as scrapers, choppers, edge modified flakes, or utilized flakes. The numbers of these that were made from the various raw materials seem consistent with the material types found in other artifact categories of chipped stone.

Further study of flaked stone tools was not pursued in this work.  Before an in-depth research project can be applied, the artifacts should be separated and re-categorized on stricter criteria. The artifacts were all collected by undergraduates, or graduate students,

Table 5. Flaked Stone Tool Quantities by Raw Material

MATERIALS Edge Modified Flakes Choppers Scrapers Utilized Flakes Total
Quartzite 32 6 2 4 44
Quartz 30 5 0 22 57
Obsidian 4 0 0 4 8
Meta-Volcanic 2 0 0 2 4
Chert 3 0 0 2 5
Granite 1 1 0 0 2
Schist 1 0 0 1 2
Total 73 12 2 35 122

many who were involved in their first field experience. Some of the sub-categories in the database seem to overlap. This was likely caused by the difference in individuals’ observations and the vagueness of some sub-categorical criteria.  Additionally, re-evaluation of the raw material identifications should be conducted, as some students most likely have classified materials differently.

Kayleen Fleming (1999) reported on the flaked stone tools that were specifically within the site CA-SDI-2508, and her findings compare consistently with the additional materials from the sites around Shingle Spring that were the source of the Lost Valley collection of excavated materials.

Diagnostic Projectile Points

Generally, Paleoindian types of projectile points are hypothesized to have been used with a thrusting spear. Prior to the 2002 excavations, no Paleoindian component had been identified from sites in Lost Valley. Archaic occupations typically consist of medium sized projectile points that were used with the atlatl, a throwing apparatus, designed to lengthen the thrusting motion of the arm. Two previous examples of possible archaic forms were presented in the Wiatava report from nearby site VS-766B, located approximately 3,600 feet (1097 m.) south southwest from Shingle Spring. Both specimens were excavated from shovel test pits (Pigniolo 1998: 99-101). The first of these examples is described as a large obsidian biface tip that was leaf shaped and from a source other than Obsidian Butte. The source at Obsidian Butte is known to have been available to prehistoric procurement at intermittent times due to the fluctuating lake levels (Dowd 1960; Downs and Woodward 1961; Wilke 1978; Waters 1981, 1983).

The second example is a large quartz biface suggestive of a leaf-shaped form. The size of these two specimens suggests that they were more likely used as dart points (Pigniolo 1998: 99-101). Sources other than Obsidian Butte “were typically used during earlier periods” because the increased depth of Lake Cahuilla would have resulted in the Obsidian Butte quarry being submerged and therefore inaccessible (Pigniolo 1998:101). Late prehistoric projectile points, or arrow points, commonly were significantly smaller and thinner, and were normally hafted to a lightweight, fletched arrowshaft, for use with a bow. It is possible that the technologies discussed above, briefly overlapped in time and across space.


The earliest known human occupations in North America are known as the Paleoindian period, spanning from the earliest known human presence, which may go back some 13,000 years ago or more, to around 8,000 years ago. Sites from these periods yield fluted points of which the classic Clovis and Folsom styles are representative.

These points are thought to have been hafted to wooden spears with cordage or sinew, and used in a thrusting motion. It has also been proposed that some of these fluted points show evidence of use as a knife, or as a multifunction tool (Rondeau 2006) (Appendix D). These artifacts seem to span different times over a wide territory, for example the Great Basin province yields slightly different examples over different times than does the Great Plains. The California situation differs significantly as well. Judith A. Willig (1991) introduced the concept of a “Western Variant” Clovis. Willig presents this notion as not necessarily morphologically distinct, but variant in a geographical and temporal paradigm, separating these from the “classic” Great Plains Paleoindian sites. Robson Bonnichesen (1991) points out that no absolute dates exist on this western variant.  Although Paleoindian artifacts are rarely found in buried sediments in California, those that have appear to exhibit slightly altered attributes. The record of Paleoindian artifacts is also blurred from incomplete and dated literature and overstated counts (Rondeau 2006).

Fluted points are a rare component in any site, but even more so in southern California. There have only been two reports of Paleoindian fluted points surfacing in San Diego County ostensibly underscoring this anomalous phenomenon. Additionally, Hyland and Gutiérrez (1995) document that only two fluted points have thus far been identified in the whole length of the Baja California peninsula. They announced the discovery of one specimen that was found in an existing collection that had been curated since the mid 1950s. They further point out that of these two rarities, sadly only one is available for analysis, as the other has been lost. It was not immediately collected due to legal restrictions, and was neither photographed nor sketched for documentation, and was not relocated upon returning to the site. This particular example is described as a basal section that appears to have been broken during use. It had additional flakes removed along the base and along the proximal portion of both blades after its successful fluting, and exhibits basal grinding.

The excavations of Lost Valley in 2002 unearthed what appeared to be a fluted point (Figure 7). The unit from which the fluted point was excavated reached bedrock at between 1 m. in the northwest corner and 130 cm. in the southeast corner where the only other evidence found at that depth was a feature consisting of a grouped collection of gneiss-schist cobbles, two of which may be cores (see Figure 15, on page 92), and from which, large flakes appear to have been removed. This depth reflects one line of evidence relative to the artifacts antiquity. Although bioturbation seems to be a problem with recording a visible stratification, the soil changed perceptibly with depth. The soil gradually became increasingly sandy and sterile with depth.

The fluted point was discovered not in situ, but nearly so. The unit under excavation had yielded several 10 cm levels with no cultural materials under the trowel and dustpan so the methods were altered to a flat shovel tossing the matrix soil up and into a ¼ inch mesh screen, where before we were screening through a 1/8 inch screen in all units. The sandy matrix at these lower levels was being shoveled into the screen at such a rate that an entire shovel-full of sand would pass unabated through the screen with little more than a single shake. The instant the fluted point was heaved into the screen, all activities stopped as a telltale scream-like shout heralded the find.

Instantly Dr. Leach inquired as to exactly where the last shovel-full of soil originated to effectively pinpoint the provenience as close as possible. Before another iota of work was accomplished, all focus on the site was directed at this spot. Soil samples, column samples, and several small nodules of charcoal were collected from within or near the 100 cm level.

Figure 7
Figure 7. The Lost Valley fluted point, obsidian.

A feature consisting of a group of four gneiss/schist cobbles within this level and unit was photographed in situ, and collected. These specimens appear to be cores with flakes removed from multiple sides. No flakes of this material were present nearby or within the entire collection.

We were unable to discern the point’s raw material until somewhat later when an unfortunate collection incident caused a small chip to become separated from the basal edge. The accidental incident revealed a newly exposed surface making it clear that the artifact was crafted from black glassy obsidian. The small fragment was saved for further analysis.

When an object of this rarity and presumed antiquity is discovered, it is requisite that a plan of research is immediately imposed. This action did not occur with immediate vigor, because of certain unavoidable circumstances, the implementation of a plan was delayed until this thesis research geared up in early in 2006.

Follow up research was planned to perform any non-destructive tests first to learn as much as possible about its age and origins. We immediately knew from the fluted side of the specimen that its affiliation with the Paleoindian period was likely but we assumed the obsidian raw material originated from either Obsidian Butte or the Coso complex of obsidians to the north.

Obsidian Butte would possibly have been underwater at the time, depending on the  early cyclical course actions of the Colorado River delta, and I anticipated that we would learn something new here. Before implementing a plan of research into the specimen, I delivered a presentation to the Society for California Archaeology (SCA) in Ventura, California with the intent to acquire as much input from the many working professionals present, before trudging on with my own plan. Luckily, I was received with many gracious offers to perform analyses. The Lost Valley fluted point has been subjected to an intense research agenda thus far and has contributed significant new information. Several analyses were applied to this specimen as explained previously in the methods chapter, the results of which are reported below:

Richard Hughes conducted a geochemical analysis on the fluted point, which indicated that the artifact’s raw material was from a quarry near Mammoth Lakes, California, over 350 air-miles distant. It was geochemically sourced through wavelength dispersive x-ray fluorescence analysis as a positive match with Lookout Mountain of the Casa Diablo Complex of obsidians (Hughes 2006) (see Appendix B). The Casa Diablo source complex is actually three local sources with a slightly different trace element composition. The Lookout Mountain quarries are the northernmost distinct source, while Sawmill Ridge is located just 1 to 2 miles (1.6 to 3.2 km) south, and the Prospect Ridge source is located about 3 miles (4.8 km) southeast of the Lookout Mountain quarries (see Hughes 1994).

The next phase of research was a blood protein analysis. I sent the point to Dr. Linda Scott Cummings at the Paleoresearch Institute in Golden, Colorado for this test. Katherine Puseman and Jaime Dexter performed the procedure, using the cross over immunoelectrophoresis method (CIEP). The specimen tested positive for Cervidae, one of the families of browsing herbivores that includes deer, elk, and moose. Results identified in CIEP are limited to the animal’s family level. Species that are related relatively close will have common serum proteins. Some cross-reactions will occur between closely related species and sometimes between distantly related animals (Puseman and Dexter 2006) (see Appendix A).

This information suggests a theoretical application of the Far Western Prearchaic model of subsistence rather than the Paleoindian big game hunter model applicable to the Great Plains (Elston and Zeanah 2002). This model is based on a behavioral ecology theoretical approach for an early Holocene or Pleistocene Holocene transitional period. According to Robert G. Elston and David W. Zeanah, this situation would fit into men’s “high variance foraging behavior”, and would be expected to be found in valley piedmonts and passes. Lost Valley is a mountain valley located at an intersection of mountain passes leading between the low desert floor near Borrego Springs and the inland valley of Warner Springs and Lake Henshaw.

The next phase of this project was a technological analysis of the point. Mike Rondeau conducted this analysis, and determined that this fluted point is indeed of Paleoindian origin, and has identified the chronology of events in the production and use of this specimen.  He determined that it is of Paleoindian origin by identifying five specific traits that are consistent with Paleoindian biface point technology.

First, there are several flakes along the blade edge that cross the midline that are attributable to a Paleoindian time frame. Second, the remnants of two individual step scars were identified as a collapsed platform that represents a failed attempt at fluting the reverse side near the basal margin. The third attribute is a series of elements that suggest that this artifact may have once been a part of a larger biface. These include general thickness and the lateral margins that diverge at the base. The point also has biface thinning scars, three of which cross the midline and two others that just reach the midline. These scars are consistent with known Clovis era technology of overshot and nearly overshot biface thinning flake scars. And finally, the channel flute scar, ending in a feather termination, and a parallel flute guide scar, clearly denote the Paleoamerican origin of this specimen (Rondeau 2006) (see Appendix C).

Large Projectile Points

Archaic temporal occupations are commonly identified by the presence of Elko series projectile points, or large leaf-shaped biface points. Elko points commonly show remarkable variation in shape and are at times difficult to define into a single typology. Projectile point typologies often exhibit a diverse array of variations on a single theme or typological designation. Several researchers have suggested that these variations in morphology resulted from either procedural artistic modes or conceptual artistic modes. Flenniken and Raymond (1986) tested a theory contrasting the Elko series nomenclature as it has been typologized into roughly two variations, the corner notched, and the eared varieties. Additionally these two subdivisions are marked by a distinct variation in length and concavity of blade edge. The authors, believing a “last mode or activity” morphological determination, offered an explanation to address the variation in styles within this typological designation. Lithic reduction technical analysis demonstrated that projectile points were continuously used, they were lost, found any number of years later and salvaged, then re-sharpened and reshaped throughout their effective use life (Rondeau 1997).

Pigniolo et al. suggested that an Early Archaic component is present in Lost Valley. He based this on artifacts and extensive site depth (Pigniolo 1998) and associated artifactual evidence from the prior Bleitz and Porcasi (1991) survey. Two large biface artifacts were identified in The Wiatava Report. Pigniolo et al. reported on one obsidian biface that is likely an archaic preform from its large size, it is sufficiently larger than one would expect for an arrow point and it was probably intended for use as a leaf shaped dart point. It was recovered from the 0-10 cm level of a shovel test pit from the site VS-766B-18. He described it as an incomplete bifacial tip that probably was not from Obsidian Butte. The other biface represents an earlier occupation. It is described as a large quartz biface preform. This preform is also of sufficient size to surmise that it was intended to become a large leaf shaped point (Pigniolo et al. 1998).

There have been two examples that point to an archaic presence from the recent excavations in Lost Valley, demonstrated by the presence of Elko series projectile points. Two Elko Series points were excavated from CA-SDI-2506, and both are apparently made from the local white, milky quartz. These points are notched in the “Elko Series” method but as we have seen in the late prehistoric points of this material, the notches are significantly shallower than normal Elko specimens made from other materials (Figure 8). The tips are missing, as are large portions of the bases of both examples. Both artifacts are very thick relative to their length and width dimensions. The example on the left measures 0.8 cm. thick and weighs 2.3 grams, while the specimen on the right measures 0.4 cm. at its thickest point, and weighs 1.29 grams. The weights and sizes are substantially greater than any of the specimens included in the late prehistoric assemblage, suggesting that they were used as heavier dart points in conjunction with an atlatl during the archaic (Pigniolo et al. 1998; Bleitz and Porcasi 1991).

Late Prehistoric

Projectile points that are typically found in late prehistoric sites in San Diego County and throughout southern California and are classified by Noel T. Justice (2002) as Western Triangular Cluster, include the Cottonwood Triangular type that is recognized by its small size and un-notched blade edges. These points were used as tips for arrows. Within this designation, there is immense variation. Justice splits this variation into many sub types, many of which are contemporaneous and found in widespread and overlapping regions. Base shape is slightly convex, to straight, to deeply concave. Their overall shape varies considerably from equilateral, to isosceles triangles, and may exhibit excurvate, straight, or slightly incurvate blade edges (Justice 2002).

Figure 8
Figure 8. Large projectile points. Possible Elko eared or corner notched. Tips and bases appear to be missing.

Desert Side Notched Cluster types are another style recognizable as late prehistoric time markers. These are small, triangular points, but with notches on their proximal blade edges. The variety of sizes and details among these types are almost as great as the Cottonwood varieties. They occasionally exhibit serrated blade edges (Justice 2002).

The projectile point assemblage from Lost Valley, specific to the late prehistoric era, was substantive and varied enough to apply statistic analysis comparing the raw material types and stylistic attributes. While the excavations were underway, I noticed a strong tendency for white, milky quartz points to be of the un-notched Cottonwood style, while other cryptocrystalline silicates that flake more predictably were more likely to be represented in the Desert Side Notched style. The late prehistoric period of occupation left a large number of complete diagnostic projectile points manufactured from a variety of lithic raw materials. The majority of the lithic artifacts are represented here by the white, milky quartz material as would be expected since it is the only material found locally in quantities sufficient to support a lithic technology. A small number of uncommonly used lithic materials were present. These include a metamorphosed mudstone and slate shaped into a triangular un-notched projectile point. The fact that these materials were used to the large extent that they were suggests that higher quality materials were more difficult to acquire and that superior materials only appeared through trade from outside sources.

The projectile points that were made from white, milky quartz were categorized along with other lesser quality lithic raw materials, such as basalt, meta-sedimentary and meta-volcanic materials. These lithic materials are represented in primarily Cottonwood triangular series points (Figure 9), while the better quality materials such as obsidian and chert, occur more often in the formal style of Desert Side Notched Series points (Figure 10).

Most projectile points of the minimal quality materials that were notched, appeared to have been notched in such a way as to be just a slight indention of the blade edge, and not really a formal notch per se. This form may have been added only as an attempt to facilitate hafting after the projectile point was completed. Figure 11 exhibits the notching spectrum of quartz projectile points from finely notched to rudimentary notches. This abbreviated form of notching would have been sufficient to eliminate a sharp edge so as to not sever the cordage hafting material, and to hold the point in place, but would appear as what Andrefsky (2001) terms an “informal” type. This was also true for some of the other less predictable lithic materials. Those of lesser quality, such as the aforementioned slate/mudstone material were also not notched.

When examining the diagnostic points, including all materials shown on a chi square table and graph, it is evident that projectile points of a lesser quality lithic material were not as likely to be notched, while other more predictable, higher quality materials were more likely to be notched (Figures 12, p 80) and Table 6, p.81). The chi square test statistic for significance of the un-notched late prehistoric projectile points made of minimal quality lithic materials, as compared to those that are notched and made from optimum quality lithic raw materials show that this occurrence is not a random phenomenon. The test shows a direct correlation between raw material and style as is exemplified in Figure 12.

When considering the quality of raw materials and the likelihood of breakage when attempting to apply a notch to the finished point, the points made from the minimal quality materials were almost always un-notched, and never deeply notched. Conversely, most notched projectile points of optimal quality lithic raw materials were in fact distinctly well notched.

The Lost Valley late period projectile point assemblage strongly parallels the pattern suggested by William Andrefsky (1994) that the locally available, poor quality material has been found to be shaped into “informal tools” or tools that have undergone “little to no [extra] effort” in their manufacture. Most informal tools are made from the more commonly available and poor quality milky quartz and other minimal quality materials such as slate. While the “formal tools” are composed of optimum quality materials such as chert and obsidian, which were imported from outside the ethnographic territory of the Cupeño people as identified by Kroeber (1925), and Strong (1929).

Obsidian is one of the optimum materials for the production of a finely worked, precision projectile point. This volcanic glass is much like modern window glass in the way it behaves when put to stress. There are numerous ways to apply stress pressure to predictably shape this material into the desired form. Chert, flint, or chalcedonous silicates are also prime mineral materials that are capable of highly predictable concoidal fracture, but are also significantly harder. Chert, flint, and jasper are classified by geologists as an alternate form of quartz and likewise reflect its Mohs hardness of 7.0 as shown in Table 4 (AGI 2003:55-6; Thompson 1998:46).

The application of notches along the proximal blade edge and on the central base, as in the Desert Side Notched series points, require less effort and pressure when obsidian is used. As for chert or other similar chalcedonous materials, the pressure needed to apply notching is increased, but the reliability of success without accidental breakage is sufficiently less than with the white quartz material. Chalcedonous materials may also have been heat treated to lower the Mohs hardness measurement to a hardness factor measurably closer to that of obsidian. Even so, the chalcedonous materials are more homogenous with less inclusions or flaws than in the quartz.

Figure 9
Figure 9. Typical un-notched quartz projectile points.

Figure 10
Figure 10. Typical notched projectile points.

All late period prehistoric projectile points discussed in this portion of my thesis are associated with the Cupeño People who were believed to have come into this area around A.D. 900 (Rogers 1945:168-170). Of these classificatory projectile points, Noel Justice (2002:367-401) has identified two distinct style-type categories that pertain to this study. The Western Triangular Cluster, with the sub-categories termed Cottonwood Triangular, Bull Creek, and Canaliño Triangular, and the Desert Side Notched Cluster, a highly variable designation with many variable forms that are not all relevant to this particular study. The main theme on which I wish to focus is the act of applying a side notch to the point’s proximal blade edge. All of the late prehistoric notched points in this collection are small triangular points similar to the Western Triangular Cluster examples, but with opposing notches on either side of the proximal blade edge, as identified by Noel Justice (2002:379). The sub-category - Desert Side Notched features a variable basal form. Some have a centered basal notch and are designated Sierra Side Notched, and some of the specimens exhibit a deeply concave base along with side notches categorized as Delta Side Notched (Justice 2002:380-83). Figure 9 shows a representative sample of the side notched projectile points in the Lost Valley collection. Justice also suggests that many triangular projectile points may have been mistaken for desert side notched preforms and that the processes for manufacturing them are very similar. Hafting for both cottonwood and desert side notched points was evidenced by a few examples that appeared to be similar, with possibly more emphasis on fore-shaft-to-basal wrapping and additional wraps covering a significantly larger portion of the blade edge in a crisscross fashion (Justice 2002:367-68).

To separate the nominal variable of material into two distinct categories, the quartz points were separated into two categories, crystal quartz being an optimal quality material and milky quartz a minimal quality material. Jasper, chert, and obsidian were placed into the optimal category and schist/mudstone material along with the basalt, quartzite, and metavolcanic materials were combined in the minimal quality category. This made the data comparable in a two-by-two, cross tabulation table, (nominal, bivariate analysis) or chi-square statistic analysis as shown in Tables 6 and 7.

The chi square analysis shows that the projectile points are not randomly distributed, and supports the conclusion that I have introduced, that formal (notched) projectile points were more likely to be made from optimum quality raw materials that were imported from outside the Cupeño ethnographic territory than the local poorer quality materials. Conversely, the analysis also shows that non-formal (un-notched) projectile points were more likely to be made from minimal quality and locally obtained materials.

Figure 11Figure 11. Notched quartz projectile points. Bottom row are made from clear crystal quartz.

Figure 12Figure 12. Comparison of lithic raw materials to styles of arrow points in the Lost Valley collection.

Table 6. Frequency Tables Showing the Raw Material of Artifacts Compared to Projectile Point Style Types

Quality of Lithic Raw Material







Minimal Quality Material



Optimal Quality Material












Style Type: Notched or Un-notched
















Experimental Results

My first exposure to archaeology spawned an interest in flintknapping and I immediately took up the art as both a hobby, but more so for the insider’s knowledge of how the various lithic materials behave under differing applications of force applied to them. I learned mostly by doing, but also by the observation of other flintknappers. My particular interest was focused on the usages of locally available materials, both lithic raw materials, and knapping tool materials.

Table 7. Results of Chi Square Tests for Points by Material Type. Computed for a 2x2 Table

Chi-square Tests


Degrees of Freedom

Asymptotic Significance

Exact Significance (2 sided)

Exact Significance (1 sided)

Pearson Chi-square

25.53 *





Likelihood Ratio






Fisher's Exact Test






N of Valid Cases






* 0 cells (.0%) expf , 5. Min exp  7.44…
Note: Null Hypothesis: The distribution of projectile point styles and raw materials are random and significantly equal. With degrees of freedom at 1 and a p-value of less than .05, a Chi-square critical value is 3.84. Since the Chi-square value of 25.53 is greater than the critical value of 3.84, we must reject the null hypothesis and determine the distribution is not equal or random.

With more practice, the experiments with the white milky quartz began to yield adequate triangular projectile points mirroring many of those found in the excavations. The material produced a substantially large amount of waste and the success rate of producing an intact usable point was found to be noticeably lower than better quality materials. This may be due to my having more practice with obsidian and other cryptocrystalline silicates, but it did not take long to learn the behaviors of this new material. Another aspect concerning the white quartz is that many of the flaws and inclusions are more difficult to see due to its white, low contrast visual nature. It appears that the manufacture of tools made from this material can be quickly learned and it would be logical to assume that prehistoric people would benefit by using locally available materials in spite of the lesser quality, given the alternative of trading for far distant materials or traveling a long distance carrying a heavy load of stone.

Experiments with milky quartz have shown that flakes of sufficient size can regularly be removed by the process of hard hammer percussion from a core. These can then be pressure flaked into an adequately sized, simple triangular, “informal type” (Andrefsky 2001) projectile point. I have produced numerous unbroken examples of similar Cottonwood and Desert Side Notched projectile points through experimentation from an adequately sized flake at an estimated success rate of approximately 70%.  Upon scrutinizing the experimental debitage, I found similar properties to that of the quartz debitage from Lost Valley and of the experimental debitage described by Boudreau (1981).

Additionally, experiments with milky quartz showed problems when attempting to apply a notch for hafting. The numerous flaws and inclusions in the white quartz tended to cause breakage at a high rate making the practice of producing a “formal style,” such as a Desert Side Notched point, probably not worth the risk as long as the point can be hafted without unnecessary waste of effort as well as the lithic raw material. If the quartz material is somewhat homogenous and the inclusions or natural facets are not present in the triangular preform, notched points can be produced with a reduced level of risk.

Both the experimental and artifactual debitage collections yielded few large flakes. Instead they appeared mainly as angular debris, shatter, and pieces so minute that the 1/8” screen that was used in the excavations would not contain them. This debitage ranged in size from very small to rarely larger than ¾” in any dimension. A debitage analysis of the data collected in the experiments was not conducted here since the point of the experiments was to determine the process and problems in producing a usable (and comparable) projectile point of the milky quartz material.

Experimentation in the reduction of two large football-sized cores yielded interesting results. Hard hammer percussion was the only reliable method for producing usable flakes. Soft hammer percussion produced small angular shatter of a size too small for any possible artifactual use. Indirect percussion yielded similar results. Both direct soft hammer and indirect soft hammer percussion methods caused rapid significant wear on either the bone punch tool or the deer antler billet tool.

The direct use of hard hammer percussion on these two cores produced many thin usable flakes that were either quickly broken and rendered useless, or were quickly formed into a usable triangular point. A few random strikes produced larger thicker flakes, which were easily refined into a strong bifacial cutting or scraping tool. These large flakes tended to accept soft hammer percussion to a better extent than did the cores. Pressure flaking necessitated a significantly increased force on the working edge than did the more homogenous materials, such as obsidian and chert, but was made somewhat easier when the angle of detachment was slightly less than 90o.
When in the process of reducing the experimental cores, I found that certain directions of impact worked better than others. There appeared to be at least one direction that produced only crumbly, useless, coarse gravel or sand, while most other angles of force produced usable pieces. There appeared to be a “grain” to the quartz that either facilitated or hindered its predictability. This “grain,” however, tended to change at the odd angled intersections with the random nature of the cleavage planes.

Throughout the reduction process, flat planar crystal facets were occasionally revealed. If a platform could be prepared that was nearly perpendicular to this facet, a large usable flake could be removed.  The experimental core reduction ended with the remains of a 4 to 6 cm. diameter nodule that was almost spherical in shape and left no facets of sufficient angle to produce flakes of sufficient size to warrant its further use. This small size left the fingers vulnerable to hard hammer impact, testing the first aid skills, and/or prompting muted vocal expletives. These experimental spent cores mirrored those suspected spent core nodules identified from the excavations.


As explained in the methods section above, the analyses typically performed on most cryptocrystalline silicates are not readily applicable to this collection due to the nature of the locally available milky white quartz. Comparatively, very few obsidian flakes appeared in the excavations. The quantity of flakes present, the proportion of projectile points and biface fragments collected in relation to the debitage, together with the absence of any cores, suggest that the finished artifacts were either traded as finished products and only repaired, re-sharpened, and disposed, or that the obsidian artifacts were produced elsewhere.

What can be analyzed successfully are the various raw materials present. Below I compare them statistically and quantitatively (Figure 13). High quality cryptocrystalline lithic debitage materials were found to be significantly lower in number in comparison with the white milky quartz. In comparison to the quartz debitage, quartzite, obsidian, metamorphosed volcanic materials, and chert were the most numerous of the remaining debitage materials and when combined, made up only 4.21% of the lithic detritus (Figure 13). Clear crystal quartz was found to be quite rare, possibly less than .02%, but is included in the collection together with the milky quartz, therefore, this is only a rough estimate, as exact quantities are unknown at this time.

What is important in Figure 13 is the imported material debitage counts. The total chert and obsidian debitage numbers and corresponding weights are substantially under-represented in comparison to their relative artifactual components. This suggests that most of the imported materials were either traded as complete artifacts or were manufactured to a complete form at another location and then brought in. The very low counts of these imported materials in debitage form may represent tool re-sharpening, or another form of tool use and/or maintenance.

As Figure 13 exemplifies, quartz vastly outnumbers all other lithic debitage materials in percentage. Quartz is the most commonly available local material that can be made into a projectile point despite its inferior predictable flaking attributes. Quartz is capable of maintaining a sharp and rugged edge for use as a projectile point, vastly outweighing the predictable flaking properties of other, less than adequate local materials. Quartzite is best suited for larger tools such as scrapers and choppers, and finished tools made from quartzite are common throughout southern California. Next to quartz, and quartzite, obsidian was more common, followed by the various metavolcanics and then chert.

Figure 14 reveals the amounts of lithic debitage raw materials in relation to the levels from which they were excavated. Note that some level categories overlap. This is because different units were excavated at varying increments depending on the amount of cultural materials present. When convenient, many of these separately excavated levels were combined to produce the data exhibited in Figure 14. Although chalcedony and basalt occurred rarely, they were encountered more often than materials such as schist, slate, and jasper, which would be represented only as a slight trace when compared to the more common materials. The levels from the surface to around 50 cm. produced the bulk of the lithic debitage, suggesting a Late Prehistoric period occupation with some significant bioturbation.

Figure 13
Figure 13. The percentage and the numerical quantitative differential relationships of the major lithic debitage materials.

Figure 14
Figure 14. Quantities of raw materials of debitage by depth.

Deeper levels represent a possible mix of late period materials that have been churned deeper through bioturbation processes, with some older occupation sequences. The excavated material from the surface down to about 50 cm. contained the bulk of the late prehistoric period evidence. Quartz is the most commonly used lithic material in both the artifact and debitage categories, probably for two reasons – its local availability and its tendency to create comparatively more waste than the more predictable and more manageable lithic raw materials. Quartzite flakes were only slightly more numerous than obsidian, probably because they were also available locally.

Chalcedonous Materials

The chert specimens found in the excavations appeared to be from numerous diverse quarries. One known source near Lost Valley was probably of silicified tuff of the Santiago Peak Volcanics. D. L. True, Rosemary Pankey and C. N. Warren (1991) identify Piedra del Lumbre (PDL) chert as the only widely known local chert source in San Diego County, from the “Piedra del Lumbre Canyon, which drains the Coast Range north of the present city of Oceanside” (True, Pankey and Warren 1991). They also point out that smaller outcroppings may exist relatively nearby, but have not as yet been identified.


Quartzite flake tools commonly exhibit hard hammer percussion flaking to shape a single edge. Quartzite colors range from gray-white to rosy pink. The homogeneity is similar throughout the collection. It is likely that much of the quartzite was obtained locally from metamorphic sources within the valley, or from Santiago Peak Volcanic residual cobbles found in streambeds and canyon arroyos within the surrounding mountains and passes.

Other Lithic Materials

The metavolcanic materials, a term that likely causes geologists to cringe, not unlike a reaction from the sound of a badly tuned violin, is a term widely used by archaeologists to represent a variety of lithic materials that are formed from volcanic processes. During the Mesozoic Era, the natural application of heat and pressure of a nearby magma chamber altered the preexisting sedimentary, or extrusive or intrusive igneous rocks throughout the Peninsular Range. This phenomenon often created a stone that is homogenous, nearly cryptocrystalline, and either fractures predictably for flaked stone artifacts or maintains a shape and is durable, making its use as either a ground stone or a flaked stone lithic material popular.


White, milky quartz is a hard crystalline mineral with numerous inclusions, facies, cracks, and flaws, making the practice of shaping a tool design extremely difficult. Concoidal fracture predictability becomes more risky with this material, and breakage would be more of a common occurrence than one would expect, even with a mediocre quality obsidian or chert material. This property of a less than adequate material may be responsible for the additional quantities of quartz waste found in the excavation. In addition to white, milky quartz, crystalline quartz was present as both waste flakes and tools however this material was not represented in the numbers comparable to the milky quartz. Some of the quartz contained inclusions of black crystal tourmaline (shorl) with a crystalline structure in the form of “needles” and groupings of individual needles combined into “bars” within the flakes, angular debris and spent cores, as well as in at least two projectile point artifacts. This is probably from a locally obtained resource, as there are still today several operating gem mines in the area that extract garnets, tourmaline, large quartz crystals and other semi-precious gems.


try to ascertain whether this lithic debitage represents a single event or a combination of manufacturing events may seem somewhat irrelevant when we consider that this group of sites most likely represent a living area, and that most tool manufacture would have taken place where debitage would have been disposed of well away from the living area. The dispersion of the debitage did not seem to be concentrated in any one locality. Bioturbation and erosion could be contributing factors. Among the many other possibilities that can explain this evidence, the spatially dispersed lithic evidence may likely represent tool maintenance in a workplace where tools were used. It appears that the accumulation was formed by actions such as taking off a flake or two to refresh an edge or fashioning an “expedient” tool for a job at hand. Another possibility is that they may simply be unfinished or finished tools that became lost. If this is a series of residential sites, it is likely that tool use, breakage, and repair was commonplace. The amount of detritus, while comparatively large, does not seem to be representative of a lithic tool manufacture locus. The poor quality of the white quartz material should be represented by a much larger relative quantity of waste. The lithic debitage reflected a homogenous spatial distribution with no obvious, visible dense concentrations. This equitable distribution may also be due to bioturbation (Herz and Garrison 1998).


Comparatively, very few obsidian flakes appeared in the excavations. The majority of the obsidian debitage is probably from Obsidian Butte, located approximately 50 kilometers to the southeast. This is the closest known source of obsidian, which was likely transported to Lost Valley by trade networks. The Obsidian Butte source is known to have been unavailable during high water stands of the ancient Lake Cahuilla, but the resource was exposed as long as the water level remained below sea level, as the currently named Salton Sea does today. The body of water filling the Salton Trough only recently became renewed by accident early in the twentieth century when Colorado River levees breeched, thus filling the vast trough over a period of a few years. This trough was once a northern continuation of the Sea of Cortez until Colorado River sediments formed a delta, which met with the peninsular range, land-locking the basin to the north from the gulf. This delta blocked seawater from entering the basin, and the dry hot environment evaporated the remaining water causing a salty dry basin well below sea level. Intermittent changes in the course of the Colorado River over this delta alternately fed into the basin and back into the Sea of Cortez. The resulting Lake Cahuilla fluctuated from high levels, as evidenced by high water stains and erosion features on the hillsides on the western shores, to complete desiccation, as it was when Europeans first entered the area.

It remains possible that the rarity of the obsidian found at Lost Valley can be attributed to the more recent availability of the resource in the late prehistoric to the times of abandonment of the site. Only 0.8% of the chipped stone debitage was obsidian. This, combined with the distance from the resource, also logically explains its near absence. The few flakes present, the proportion of projectile points and biface fragments collected in relation to the debitage, together with the absence of any cores, suggests that the finished artifacts were either traded as finished products and only repaired, re-sharpened, and disposed, or that the obsidian artifacts were produced elsewhere. The unique properties of obsidian make it convenient to use as a relative dating tool, and additionally, it can be reliably sourced through several types of geochemical analyses. There were sufficient obsidian flakes present to cast a glimpse of the span of time that Lost Valley was occupied.

One individual flake was sourced through edxrf by Richard Hughes at Geochemical Research Laboratory, and it was determined that the trace element composition matches that of Obsidian Butte in Imperial County, California. This flake was subjected to obsidian hydration analysis and presented two rind measurements from separate internal cracks that had penetrated the surface of the flake. These measurements were calibrated to 7.21μm on one crack, and 8.15μm on the second.  An obsidian hydration rind was visible on the surfaces of the flake but were in such a weathered condition that it was impossible to accurately measure the depth of penetration. The flake also was covered with a relatively thick patina that when viewed under the microscope, this appeared as a layer of grit obscuring the surface “edge” from where a hydration measurement would terminate. The depth of the surface hydration appeared consistent with the cracks that were measured with substantially enhanced accuracy. For complete and detailed results of the obsidian hydration analysis, refer to the obsidian hydration analysis section below. Additionally, five obsidian waste flakes were previously subjected to geochemical source analysis by Kaylene Fleming (1999) and were also attributed to the Obsidian Butte source.


There are 136 quartz cores in the Lost Valley collection. They were relatively small in size ranging from approximately 5 cm. or less in diameter. They were all of the poor quality white milky quartz material. Again, as in the debitage, there were negligible “classic flake features” present on the surfaces of these cores except for traces of cortex contrasting with clean angular facets. These white milky quartz cores were all small enough to assume that they were spent. No facets were large enough, or exhibited the correct angles to facilitate the removal of a sufficiently sized flake in which to manufacture a projectile point or a usable tool. No clear crystal quartz cores were identified and no cores of any of the higher quality cryptocrystalline silicates were present either.

A total of 60 larger cobble sized quartzite cores were present together with a few large quartzite flakes that may be identified as, or could have been used for scrapers or small choppers, but over all, quartzite flaked cores were not represented in the quantities that would be expected as they are elsewhere in the region.

A feature consisting of four cobble-sized gneiss-schist possible cores (Figure 15) were found grouped, and at approximately 1 meter depth, within the same unit and about 50 cm. horizontally from the Paleoindian component and may be associated. As these specimens show, flakes have been removed from many sides, and the flake margins are sharp and clearly visible. A total of 6 gneiss-schist cores are present in the collection. Other cores in the collection consisted of metavolcanic (N=8), Granitic (N=4) and Chert (N=1).

Figure 15
Figure 15. Four possible cores made of gneiss-schist, excavated from the same unit and level as the Paleoindian fluted point.

Obsidian Hydration

The limited quantity of obsidian flakes in the collection was sufficient to perform statistical analyses to obtain a reasonably reliable conclusion. Only a few large flakes 2.5 cm. diameter, which may be described as closer to angular debris than as flakes, were evidence that some limited reduction work with obsidian was accomplished in the vicinity. Small “classic” pressure flakes, many of which were too small in diameter and thickness to easily submit for a hydration analysis, represented the vast majority of obsidian debitage. Obsidian hydration rind measurements ranged from the smallest at 1.26 μm, representing the most recent use of obsidian, to the largest rind penetration at 8.15 μm, evidence of the oldest use of obsidian. All obsidian hydration samples were prepared and hydration rinds measured at the San Diego State University Obsidian Hydration Laboratory under the direction of Dr. Glenn S. Russell. All measurements were made a 400X magnification using images captured from a Meiji model ML9430 polarizing microscope with a INFINITY2-3C microscope mounted digital camera. The optical image was measured with IMT iSolution Lite, Version 7.0 software.

Although I did not apply a large portion of the obsidian to sourcing analysis, I am assuming that the majority of the obsidian comes from the Obsidian Butte source in Imperial County, California. I am making this assumption based on the grounds that this is by far the nearest source, and based on data from other studies (Fleming 1999), and reassurance from local professionals active in the business of archaeology that Obsidian Butte is the most likely source.

The one obsidian flake that was subjected to a geochemical analysis was determined by Richard Hughes (Appendix D) to match the trace element composition of Obsidian Butte. This particular flake was found to exhibit the largest rind measurements, and this finding, together with a noticeable patina not seen on any other obsidian flakes, influenced my decision to pursue the sourcing analysis. Hydration rinds preserved along two separate internal cracks were measured and calibrated to 7.21 μm, and 8.15 μm.

Geochemical analysis can be applied to finished artifacts in a different way than in the debitage. When a source is identified on a finished tool, it can be assumed that this tool could have been made elsewhere and was transported, or was exchanged to the place where it was eventually collected, or the tool was fashioned locally from an imported core. There also exists the possibility of material salvage and subsequent transport from another environment with an altogether different hydration rate. The desert floor, just a few miles to the east, would represent one of these environments where obsidian hydration rates would differ significantly. A debitage flake presents a different story altogether. Flakes would not likely travel far. A debitage flake was probably struck off of a core, biface, or tool in the place near where it was discovered. Small waste flakes in a prehistoric context would not likely have been intentionally transported.

It is well known that obsidian hydration rates are a contentious issue with a wide variation of rate equations published for the Obsidian Butte source (Laylander 1992). While some are linear equations, denoting a constant rate over a limited span of time, others are set on a curve showing a diffusion rate. The two equations are  (T = k d) and (T = k d2).  The letter “T” represents the time an obsidian surface has been exposed to the elements, “k” represents the constant, or the rate of hydration, and “d” denotes the hydration rim measurement.

The equations with an exponent added to the rim measurement variable are those that are set on a curve and may be truer to reality because the rate at which water penetrates the obsidian is not a constant. The hydration process begins at a relatively rapid rate and slows its advance into the obsidian material as it progresses. These two equations are demonstrated in chart form in Figures 16 and 17.
There are many variables that affect the rate of hydration.

Environmental factors concerning long term mean temperature and humidity, whether or not the specimen was exposed on the surface, or buried, and the trace elements and water content in the obsidian are only some of the variables. The primary factors however, are the temperature of the environment and the composition of the material (Michels 1973; Friedman, Trembour, and Hughes 1997).

This study will try to avoid the rate-based mathematic fray by reporting the measurements as data points, so that this information can be recalibrated to a reliable rate equation in the future. Depending on the reliability of the obsidian being from the source at Obsidian Butte along the southern shore of the Salton Sea, we can assume that these dates are relative to one another, making the histogram in Figure 18 a graphic representation of Lost Valley occupation. The obsidian flakes from this group of sites in Lost Valley that were subjected to hydration testing were all exposed to the same environment and likely remained from the original deposition. All of the measurements were taken from waste flakes or flakes with minimal use-wear, and were less likely to have been transported after their removal from the core or biface. We can be reasonably assured the histogram in Figure 18 is a representation of occupation throughout the late prehistoric and into the archaic based on the reliability of the link from the smallest measurement to the time just before the Cupeño were forcibly moved to the reservation at Pala in 1903. The spike of measurements at around 3.4 μm should represent the peak of late prehistoric occupation. The peak of measurements at 5μm and the several measurements reaching out to 8.15 μm probably point to the archaic period, another line of evidence reinforced by the presence of large projectile points attributable to atlatl dart use, by both Pigniolo et al. (1998), and the Elko projectile points from the Lost Valley collection. It is useful to note that there is no correlation of hydration rind thickness with depth below surface, a clear indication that the vertical position of small artifacts has been heavily affected by bioturbation.

Figure 16
Figure 16. Obsidian Hydration Time Chart. Note: This is based on a diffusion model with the hydration measurement value squared and a constant multiplier at 66.77 years. T=kd2 ( T = time in years, k = constant based on assumed association between known hydration measurement and point in time, and d = individual hydration rind measurements squared).

Figure 17
Figure 17. Obsidian Hydration Linear Rate Chart. Note: This is based on a linear equation where a hydration rate of 66.77 years per micron is constant through time.

Figure 18Figure 18. Histogram showing the obsidian hydration analysis results of 66 rind measurements.

The most recent hydration measurements may be from no later than the fall acorn harvest of the year 1902 before the Cupeño were removed from their village, Kupa, at Warner Springs in February of 1903 (Castillo 1978). Phil Brigandi told me of a visit to Lost Valley by Rosinda Velasquez in 1987. She related to him that she was at a site near the rifle range as a child of about 10 years old and witnessed the processing of acorns in the nearby bedrock mortars. This information, together with the calculation of her age at the time, would have likely represented the last Cupeño occupation at Lost Valley (P. Brigandi personal communication 2008). 

Many past researchers have tied certain assumed or hypothesized variables into the hydration rate equations to calibrate the curve, such as low level lake stands of Lake Cahuilla that were reported by Waters (1983), but these dates are now under scrutiny as there are now many more levels reported from lakeshore sites, suggesting an increased variation in lake water levels.

Rather than use this approach to estimate a hydration rate for the Lost Valley obsidian, I assumed that the rate conformed to the basic diffusion model and equated the smallest hydration rind measurement with the last known aboriginal use of the area by the Cupeño in 1902. Although this is only a single data point, it results in the rate depicted in Figure 16.  Although very approximate at best, it does yield a rate that is generally consistent with the dates of the known occupation of the area. It is offered here only as a preliminary rate estimation, should not be applied to hydration data from other areas, and clearly needs to be refined using additional hydration data associated with other reliable absolute dates.

To calculate the hydration rate constant used in the Figure 16 graph, I used the smallest hydration measurement of 1.26 μm and equated that specimen to an assumed time of  106 years before the present (YBP).  With this assumption, I divided 106 by 1.262 (or 1.5876) to arrive at a 1.0 μm, which would represent an age of ≈ 66.77 years.


Ground stone artifacts in this analysis are classified into three categories representing their usage purposes.  Items used for the processing of food materials (milling stones) consisted of many fragmentary remnants in the form of manos, metates, portable mortars or bowls, and pestles. By far the most obvious feature found today throughout various sites circumnavigating the valley are the large bedrock milling stations. The nearest example to CA-SDI-2506 is a large, reasonably flat granite outcrop 20 meters due north of the site’s north datum stake. This feature consists of no less than 30 conical mortars of varying depths on a rock outcrop approximately 20 to 30 square meters in area. There are numerous similar bedrock milling stations located throughout the valley featuring numerous individual conical depressions, but few feature as many as these near the Bog Site.

Thirty two fragments of portable ground stone metates, one rim fragment of a stone bowl, twenty diverse forms of pestles, and sixty two variable sizes and shapes of manos were found during the Bog Site excavations. 

Only three items of physical body adornment were found. One is identified here as a zoomorph, and two others as pendants with a biconical drilled hole for cordage attachment. Finally, there were two examples of ground stone arrowshaft straighteners classified here as a shaped stone tool used in the production of arrow shafts. Twelve fragments of separate doughnut stones are included in the collection that were used as a weight in conjunction with a digging stick for collecting edible tubers. There were an additional 38 fragments of groundstone artifacts that have yet to be unidentified as to their use or purpose.


Most of the ground stone pestles recovered were of opportunistic shapes. Many were basically three-pointed rocks with one point at a less acute angle than the others, enabling its use without much additional effort. Pestles did exhibit great variability in raw material, size, and shape. Two small, finger-sized pestles were collected that may have been used for small precision work such as mixing pigments or medicinal ingredients. The larger specimens were likely used for processing acorns or other seeds in larger quantities. The presence of more than 12 bedrock mortars, and 6 grinding slicks were exposed with the likelihood of more hidden just below the surface beneath a shallow grus deposit, just north of the site CA-SDI-2506 boundary.

This site boundary is also the beginning of Site CA-SDI-2507. The presence of these mortars explains the occurrence of the many non-formal pestle types, and the variety of odd shapes. Raw materials were not as variable as the other ground stone artifacts. The vast majority of pestles were short and thick, and were slightly altered natural cobbles formed from quartzite or a granite material, but a few were intentionally shaped, elongate, and made from schist.

One extremely long and heavy schist pestle was found during the excavations of 2002 in the Bog Site (Figure19). This pestle was broken into three main segments and was recovered from the 50 to 70 cm. level in Unit 26S 10W. A total of seven pieces make up the complete artifact. The two ends of this pestle differ in that one end is blunt and flattened while the other is comparatively rounder and more pointed. The overall length is 65.5 cm. and all seven fragments combine to weigh 5.465 kg. When reconstructed, it is consistent with similar types found in neighboring territories, including areas near Borrego Springs, that were probably used for the purpose of processing Algaroba Mesquite beans. Historic Cahuilla people reportedly used long pestles of wood or stone in conjunction with wooden mortars made from Algaroba Mesquite wood or cottonwood stumps of up to about 2’ 6” (76 cm.) long. The production of these mortars reportedly commenced with first burning out the center to soften the wood, and then continued hollowing ensued, coring out the center of the log in this fashion until a bowl was formed (Barrows 1920:227; Bean and Saubel 1963:58; Kroeber 1908:40, 52). The length of these pestles facilitated usage in the standing position and also would remain ergonomically feasible as the mortar gradually wore a deeper depression in the wooden mortar (see another example in Figure 20).

Figure 19Figure 19. Schist Pestle, Catalog # 5276,  from CA-SDI-2506: Three major fragments (table width is 60 cm.).

Figure 20
Figure 20. Schist Pestle, Catalog # 5276,  from CA-SDI-2506: Three major fragments (table width is 60 cm.).

The material of the pestles in Figure 20 is not identified, but appears to be a granitic rock type or possibly gneiss. These examples are from the San Gorgonio Pass area and from the desert. They are not exhibited with a scale bar, but the longer pestle’s length to diameter ratio seems to be comparable to the example in Figure 19. Note the apparent “polish on the surface of the upper and narrower half of the longer specimen in Figure 18. Another similarity between these two pestles is in the two ends, one being somewhat blunt and the other rounded.

The schist from which the long pestle was manufactured could possibly come from a number of sources in the highlands bordering the valley to the south, or even possibly from within the valley. These southern mountains are geologically different than the relatively nearer northern and eastern mountains in that they are mostly metamorphic muscovite-biotite schists, sillimanite-biotite schists, gneisses, quartzites and banded quartzites (McCullough 1984:5).

Evidence supporting the Cupeño – Cahuilla pre-historical link can be illustrated in two historic photographs of Juan Chutnikat, Figure 21 shows a wooden mortar in a carrying net taken by J. P. Harrington in the early to mid 1920s (Bean and Smith 1978:590). Figure 22, also taken in the 1920s, was photographed near the transition of the Cupeño-Cahuilla boundary-interface at Palm Canyon. Ethnohistoric and linguistic evidence from Hill and Nolasquez (1973) suggests that the early Cupeño were once a lineage of mountain Cahuilla who splintered off and migrated south approximately 1,100 to 1,300 years ago (Pigniolo et al. 1998).


Two large mostly complete boulder metates were recovered from the sites, one of which is on display in the Lost Valley Boy Scout dining hall display case. Ninety-six other fragmented pieces were recovered throughout the excavations. The example in Figure 23 is representative of the majority of fragments, showing the fragile, decomposing condition of the granite. This metate fragment easily broke in two from an existing crack as it was removed from the ground and continues to crumble from minor handling. It is doubtful the rock would have been in this condition when in use since this sample broke under minor stress. A metate under normal use would have had to withstand significantly greater stresses. Additionally, this example, in the apparently weathered condition found, would have added a substantial quantity of grit into the food. The material easily crumbles with the slightest pressure between two fingers.

The raw stone material appears to be of the standard granitic material found throughout the valley and much of the surface of the local granite is in the decomposing state. Grinding slicks were noted on several of the flat granite bedrock outcrops throughout the valley in association with mortars.

Figure 21
Figure 21. Figure 21. Juan Chutnikat with a wooden Mortar. (Photo from Bean, Lowell John, and Charles R. Smith, 1978. Cupeño. In Handbook of the North American Indians, Vol.8, California, Robert F. Heizer, ed., pp 588-591. Washington D.C.: Smithsonian Institution.

Figure 22
Figure 22. Photo of Juan Chutnikat with two Cahuilla men and J. P. Harrington at Palm Canyon (east of Lost Valley) in the 1920s. Source: Heizer, Robert F. 1978   History of Research. In Handbook of the North American Indians, Vol.8, California, Robert F. Heizer, ed., pp 11. Washington D.C.: Smithsonian Institution.

Figure 23
Figure 23. Metate fragment of decomposing granite (Cat. #s 4134, and 4135).


More than 218 manos or mano fragments were recovered from the Lost Valley excavations. These varied in raw material from quartzite to gneiss/schist, and the ubiquitous granite. Few appeared to be of the formal type that were intentionally shaped with two opposing grinding surfaces and shouldered edges. The majority of the manos were found to be of the non-formal category - utilized river cobbles with both single-sided and two-sided forms. The manos were brushed clean in the lab to remove excess matrix soil and they remain in a state of preservation that would lend future analysis feasible.

Notable and Unique Ground Stone Artifacts

A pendant (Cat # 5310) was recovered from between 10 and 20 cm. in CA-SDI-2506 (Figure 24). It exhibits a biconical, drilled hole near the top edge, and polish along a portion of the biconical drilled hole that was probably used to suspend cordage. It has a highly polished edge all around, but the greater polish near the drilled hole still easily contrasts with its edge. One side of this pendant features four vertical parallel lines that appear to be deeply gouged, chiseled, or scratched into the piece. The raw material is probably a metamorphosed mudstone that may have been previously stream worn and flattened.

The biconical hole has concentric drill scratches near the hole’s edge. This specimen – much like another, larger tabular artifact – seems to have been used in a similar fashion due to the wear and polish evidence. The reverse side of this piece features a mineral inclusion in the form of a linear bulge, possibly a naturally occurring mineral crystal. It measures 2.35 cm. long, 1.98 cm. wide, and 0.4 cm. thick and weighs 6.93 g.

The zoomorphic shaped bead or pendant, also excavated from CA-SDI-2506, (Figure 25) features a biconical-drilled hole as well. This specimen (Cat # 5276) seems to be formed from the same raw material as the tabular pendant shown in Figure 24, but is more robust. It measures 3.61 cm. long, 1.50 cm. wide, and 0.93 cm. thick, and weighs 3.75 g. There are no apparent features representing eyes or other physical attributes other than what appear to be a nearly round carved head and a somewhat conical tail. Carving or shaping scratches are visible in the carved recesses of the piece. All outer surfaces of this piece exhibit a polish, but no evidence of suspension by cordage is discernable.

A large flat tabular pendant (Cat # 4978) was excavated from CA-SDI-2506 at a depth of 113 cm. It is made from slate-like stone and has a biconical, drilled hole near the top end (Figure 26). The area between the hole and the top end exhibits a polished surface suggesting suspension friction from attached cordage. One face features two rows of horizontally aligned micro-cupules drilled just deep enough to be easily visible (Figure 26a). The reverse (Figure 26b) side also exhibits the drilled micro-cupules but these are aligned180o out from the opposite side and are arranged in at least 5 unevenly spaced vertical linear alignments. The micro-cupules range in number from right to left at 6, 5, 5, 4, and 5 for a total of 25. The cupules on this side are shallower, less defined, and some additional features may be obscured by the breakage. It is also curious that scratches that appear on the reverse side are also similar in their alignment with the scratches on the opposite face, running nearly horizontal, possibly from a repeated pendulum-like motion. The pendant measures 6.99 cm. long, 3.34 cm. wide, and 0.29 cm. thick, and weighs 13.8 g. The multiple wear patterns suggest that this item was suspended by cordage.

Figure 24
Figure 24. Incised Pendant from CA-SDI-2506.

Figure 25
Figure 25. Zoomorph pendant/bead from CA-SDI-2506.

The pendant was recovered in two halves, both from a depth of 113 cm., and within unit 24S – 8W. This may seem a considerable depth in that most late period lithic materials were found at levels above 60 cm. If this item was a part of some funerary deposit, additional supporting evidence is no longer present. The example featured in Figures 26a and 26b were photographed with a blue-white LED light source at a nearly right angle to enhance the shading of the punctate pattern, but also added a false bluish hue that was digitallycorrected using Adobe Photoshop CS © by artificially returning the red, green and blue wavelengths. The photograph of the pendant microcupules shown in Figure 25 displays a truer color representation of the specimen.

Figure 26
Figure 26. Punctate patterned pendant, Cat # 4978, from CA-SDI-2506.

The arrangement of the punctate, microcupule designs on both sides is similar to a plethora of rock art designs found throughout the west and the Great Basin. Patterns of “dots,” whether in single or multiple lines, in circles, or uneven groups, appear alone or with other designs on vertical rock faces, albeit in larger scaled representations than on the pendant. The micro-cupules featured in Figure 26a are shown magnified in Figure 27. The cupules measure slightly over 1 mm. in diameter on average, and the linear arrangement spacing varies from 2 mm. to 3 mm. apart. The two rows of cupules are also spaced about 3 mm. apart.

This tabular stone pendant is similar to one defined as a constituent of the San Luis Rey Complex and is also described within the Cuyamaca complex (True 1970; True et al. 1991). The pendants from the Cuyamaca State Park cited by D. L. True are described as being made of steatite. D. L.True, Rosemary Pankey and C. N. Warren (1991) show a drawing of a similar pendant in the Tom Kav report attributed to the San Luis Rey Complex. Pendants from site CA-SDI-682 are described as made of “a slate, or schist-like material, shaped by edge grinding… Decoration, where present, consists of incised or drilled designs (usually geometric)”. Another slate pendant is identified within the Armogosa Tradition of the southeastern California playas, deserts and ranges (Chartkoff and Chartkoff 1984:179). Those pendants, associated with the Armagosa tradition, were also made of ground-slate and have a biconical drilled hole at one extreme end. The Armagosa tradition is regarded as lasting for 1,500 years, terminating around 1,000 years ago.

Figure 27Figure 27. Macro Photo of micro-cupules on #4978.

The tabular drilled stone pendant with the multi-linear punctate motif (#5310), and the carved stone zoomorph (#5276) are similar to artifacts described by D. L. True (1970) of the Cuyamaca Complex within the Kumeyaay ethnographic area to the south. These pendants (excepting the zoomorph example) also closely resemble specimens from the San Luis Rey II Complex. This mixture of influence from the surrounding culture groups can be explained by many themes such as stylistic diffusion, intermarriage, conflict, conflict aftermath, or trade activities.

Arrowhead Straightener Tools

Arrowshaft straighteners are manufactured from a smoothed fine-grained stone and were used to remove bumps and bends from arrow shafts to facilitate a smooth release from the bow and a reasonably straight flight. These are not to be confused with similar tool types fashioned from an abrasive stone such as sandstone or an extrusive igneous cinder or vesicular basalt. The abrasive types are likely a shaping tool used to remove irregularities, which would affect balance, and it may be argued that they were possibly used additionally as a sharpening tool for awls or other bone, wood, or antler implements (Toulouse 1939). Although little or no ethnological data support these assumptions, it seems logical that this use would be an effective additional use for the tool.
The arrowshaft straighteners however, are smooth (non-abrasive) and have a polished surface within the roughly semi-cylindrical groove with a noticeable residue build-up within the central two-thirds of the groove. The Lost Valley examples feature no decoration as do many examples found elsewhere nearby among the Cahuilla (Kroeber 1925:531), the Kumeyaay (True 1970), and the Luiseño (True et al. 1991). Paul Campbell (1999) describes an artifact with similar attributes, but measuring slightly smaller, that was found near Santa Catarina. The lack of a distinct or unique decoration or patterning implies that these were of strictly utilitarian usage. Decorations could serve single or multiple purposes not limited to connotations of supernatural powers, or even a simple method of identifying the piece with its particular owner and/or maker.

The two shown in Figure 28 appear to have been formed from quartzite. The example on the left is assigned Catalog # 4495, and the one on the right # 5003. The patina and residues present on the surface mask the subtle features that would positively identify the material. Artifact #4495 was excavated in 2002 from the southwestern quadrant of CA-SDI-2506 in unit 18S – 8W within the 20 – 30 cm. level. It measures 8.0 cm. long, 4.64 cm. wide, and 2.24 cm. thick. Artifact # 5003 was also excavated in 2002 from a nearby unit, 26S 10W at a comparable depth of 20 – 30 cm. This specimen measures somewhat larger at 8.58 cm. long, 4.86 cm. wide, and 2.66 cm. thick.

Figure 28
Figure 28. Undecorated quartzite arrowshaft straighteners.

They both feature blackened charcoal stains and residue in the notch from apparent use. The report, Wiatava, describes an arrowshaft straightener that is both decorated with incised lines and made of steatite. The decorated version was collected from CA-SDI-2508, a site only separated from CA-SDI-2506 by a single-lane, graded dirt road. This particular artifact from CA-SDI-2508 was described as decorated in the database, but the published photo (Flemming 1999) shows parallel, incised lines crossing the notch at a right angle and seems to be for the purpose of an abrasion structure. This would logically serve more as a form of function than as a decorative attribute.

Stone Bowls

Only two fragments of portions of ground stone bowls were excavated from CA-SDI-2506. One of these fragments is a rim portion and has a 3 to 4 cm wall thickness, consistent with stone bowls found elsewhere in California made from steatite and vesicular basalt. However the two fragments here were made from granitic rock. It is difficult to assess the overall size of the bowl from the two small fragments, or if they are from the same piece, however they seem consistent with the commonly found sizes and diameters seen elsewhere in the west. The presence of stone bowls may suggest its use as a portable mortar.

Perforated "Doughnut" Stones

Twelve fragments of perforated stones, shaped from steatite, quartzite and a granitic material, were found in the Lost Valley excavations at sites CA-SDI-2506, 2507, 2508 and VS-766C. The fragments of these specimens were excavated from all levels, down to a one-meter depth and also found on the surface. Steatite, also familiarly known as soapstone, merits this popular allonym from its smooth “soapy” feel when rubbed between the thumb and fingers. Steatite’s physical properties facilitate its rapid shaping into the desired form from its relative soft texture. The Mohs hardness of steatite, a form of talc, is measured at 1, which allows for the shaping with any material that exhibits a higher hardness number on the Mohs scale. Even ones’ fingernail, with a Mohs hardness of 2.5, can easily abrade steatite. The source of this steatite has not been identified. These examples of steatite appear micaceous, equigranular, and light green in color. The granite and quartzite fragments may possibly be made from locally available raw materials.

Other Stone and Mineral Artifacts

Seven clusters of red-orange and yellow-orange ochre nodules, or iron oxide, were excavated in 2002 and 2003 within 6 units in the southwest quadrant of site CA-SDI-2506. Ochre is a common pigment that was used throughout the world for coloring textiles, ceramics, pictographs and other works of art, and as body paint. These nodules were very soft and would easily crumble with the slightest pressure between thumb and forefinger. It is logical to assume that these delicate lumps of mineral were brought into the site from surrounding points to the Shingle Spring camps for the use as a coloring agent, however it is possible that they could be naturally occurring. These ochre nodules were assigned 7 separate catalog numbers and curated according to the units and levels where they were found. The largest two nodule clusters were Catalog # 5426, which consisted of 12 nodules with a total weight of 11.5 grams, and #4979, 9 nodules weighing 8.5 g. A large, single nodule, Cat. # 5091, weighed 3.67 g. These three deposits were excavated from relatively deep levels in separate 2m2 units. Artifact # 5426 was found within the 70 to 90 cm. level from unit 24S 4W. Artifact # 4979 was recovered from a level just below 1 meter deep from unit 24S 8W. Artifact # 5091, the single nodule, was found at the 70 to 90 cm. level of unit 22S 2W.
Graphite nodules were also found in Lost Valley, which produce a black pigment. The 14 graphite nodules were found in three sites near Shingle Spring, four nodules were excavated from VS-766, eight from CA-SDI-2508, and two from CA-SDI-2506. These nodules were all relatively larger and firmer and would survive much more intense weathering and physical abuse without crumbling when compared to the ochre. These 14 nodules were widely scattered throughout the excavated units and depth levels ranged from just below the surface to 90 cm.

Crystals of shorl tourmaline were present that also formed within much of the quartz lithic debitage. These crystals ranged in size from individual black needle-shaped inclusions within the quartz, to groups or accumulations of needles forming short glossy black bars or shafts up to 1 cm. in diameter and 2 cm. in length. The presence of these crystals in such numbers is not surprising due to the geology of the immediate vicinity. Several mines are still working today yielding black shorl and other varieties of tourmaline, and also garnet, quartz, and other large crystals and semi precious gems for the commercial market.

One particular shorl tourmaline crystal came to light in the form of an aggregation of glossy, black, parallel, needles, shaped as a solid bar, with an apparently purposefully shaped notch circumcising one end possibly for the attachment of cordage and/or a cementing material (Figure 29). There is no visual evidence of any cementing material or residue present. A few fragments of clear, six-sided quartz crystals were found, but their use and the reasons for their presence at the site remains enigmatic. They may be debitage from the manufacture of clear quartz projectile points or other tools similar to some excavated from this group of sites. They may also have been used as sacred objects.

Fire Affected Rocks

Hundreds of small fire-affected rocks (FAR) were unearthed throughout the excavation activities, but were not collected or recorded other than in notes as to their presence. These rocks manifested angular fractured faces that normally occur when excessive heat is applied to a weakened rock causing thermal expansion unevenly across its surface and deep into the rock’s core. Many of the cobbles were composed of either quartzite, or granite, or intermediary metamorphosed stages of phylite, schist or gneiss, all of which are common to the immediate vicinity. It remains possible that any number of these FAR specimens could have been formed from periodic naturally occurring fires.


Several of the Tizon Brown Ware sherds featured biconically drilled holes, as the example in Figure 30 shows. These drilled perforations are commonly described as mends used to fix cracks. One local example from a Kumeyaay collection at SDSU still retains fragments of the cordage and traces of a mastic substance that was applied to repair a large olla. Several sherds in the collection have been identified as Salton Brown, Tumco Buff, and Lower Colorado Buff. These ceramic types all suggest trade or other associations with desert cultures to the east.


Most faunal bone fragments were in a very poor state of preservation. At least two specimens exhibited butcher marks. The few exceptions to the poor state of preservation were bone tools and bird bone beads. The two bird bone beads appeared to be highly polished on their exterior surfaces from use/wear (Figure 31). Their well-preserved condition is surprising compared to other less fortunate faunal specimen remains. These two examples may likely have been one single piece and were broken into two pieces, however they do not seem to fit together. They were both found near each other within the same unit and level and exhibit nearly identical size, color hue, and material properties. The one shown to the upper left was slightly longer measuring 6.375 mm. long, 3.44 mm. widest diameter, and the diameter from the flattened side is 3.225 mm. The bone bead shown in the lower portion of Figure 31 is 4.765 mm. long, 3.64 mm. widest diameter, and the diameter including the flattened side is 3.18 mm. These beads were excavated from site CA-SDI-2506, from the 2m2 unit 26S 10W, and from the 60 – 70 cm. level.

Figure 30
Figure 30. Drilled Tizon Brown Ware rim sherd.

Figure 31
Figure 31. Bird bone beads Cat. # 5040.

One particular bone fragment illustrated in Figure 32 may have once been used as a pressure flaking tool or as an awl. The tip of this specimen features wear marks remarkably similar to the flint knapping tools contained in my own personal flintknapping tool kit. One impact scar is also present about 3 cm. from the tip. The item has a freshly broken edge from excavation activities (Figure 32). Most of the collected bone was in a poor state of preservation. Several pieces crumbled under the slightest touch.

Figure 32
Figure 32. Bone tool fragment.

One fragile bone fragment, Catalog # 5252, had possible butcher marks along two edges (Figure 33). They do not appear to be from rodent gnawing, as one would expect from a rodent’s paired upper and lower incisors. The parallel lines on Side A are of slightly different lengths and spaced closely but not evenly. The marks on side B are randomly spaced along 2 cm. of one edge and appear to be possible chop marks. This bone fragment was extremely fragile and recent breaks can be discerned from the light colored surfaces along the broken edge. This specimen was excavated from CA-SDI-2506, unit 26S 6W, and from the 70 to 80 cm. level.         

Two examples of a purposely shaped bone awl were found that exhibit polish from use and scratches probably from their original manufacture. These both came from CA-SDI-2506, Unit 26S 6W, and from the 80 to 90 cm. level. One of these is illustrated in Figure 34. Curiously, these objects were well preserved. The preservation of highly used tools made from bone contrasts ostensibly with the non-utilized faunal remains that appear to be discarded. It is possible that the handling and use somehow provided a mechanism for preservation. This may be a taphonomy hypothesis for further study by a specialist in the faunal analysis of archaeology. At only 4 cm. in overall length, this tool could have been used for basket making or in the production of clothing, or any number of tasks where a hole needed to be made or enlarged in some relatively soft hide or textile material. Much of the faunal bone in the Lost Valley collection is in an extreme state of decomposition.

Figure 33
Figure 33. Possible butcher and other marks on bone fragment.

Figure 34Figure 34. Bone awl Cat. # 5257.

Bedrock mortars were found in large numbers at nearly any horizontally flat granitic outcrop near the spring or along the headwaters of Agua Caliente Creek, within the Lost Valley flatlands. Others were found all around the valley where the various species of oaks grow and a temporary water source is nearby (see Figure 35). These could be communal acorn processing localities, since there are often dozens of individual mortars within these bedrock features. The vast numbers of these features may suggest large scale processing for transport and/or trade. These features were so numerous that it would require a separate, intensive study to address them.

At VS-766C, rock groupings were recorded that may have been post-hole supports (Figure 36). This may be part of a shelter or possibly part of a granary structure. Another grouping of rocks was speculated to be a hearth due to the presence of charcoal. Note the blackened soil around a rock grouping in the lower left corner of Figure 36. Larry Leach obtained an AMS 14C date of 1,920 +  40 years BP from bone collagen from this site (See Appendix F). Within the excavation at CA-SDI-2506, numerous rock groupings were noted throughout the excavated units, which only looked suspicious from the lack of similar sized cobbles in an evenly scattered array. It seemed strange that nearly all of the fist-sized or larger rocks were found in groups, and at various levels. These were all sketched and photographed and will eventually appear in detail in the final site report. The composition of most of these cobbles were of metamorphosed igneous or sedimentary sources located elsewhere in the vicinity and likely were manually transported to this location from another nearby local source. The area encompassing the sites in this work are located at the relatively level valley edge, up to and against the base of the steeper granitic slopes, and slightly uphill and to the northeast. The sites around Shingle Spring are well within the muscovite-biotite leucoadamellite of Lost Valley. This geologic unit is characterized as garnetiferous, equigranular and homogenous, exhibiting a light gray color (McCulloch 1984). The immediate rock unit to the northeast and up the hill is the biotite-leucoadamellite of Lost Valley.

Figure 35Figure 35. Bedrock milling feature within CA-SDI-2507.

Figure 36
Figure 36. The Archery Site excavations of 2000.