My 2009 report recommended that "lead isotope ratios should be obtained for the four lead balls" discovered at Kuykendall Ruins, and that these ratios should be compared to ratios from lead balls found at other known or suspected Coronado sites.(1) We followed this course by measuring the lead isotope ratios (LIR) of shot from Kuykendall Ruins (Chichilticale), Doubtful Canyon (Camp 23 June 1540), Hawikku-Kyakima, and the Jimmy Owens sites. In addition we used the LIR presented by archaeologist Charles M. Haecker for two lead shot and one copper crossbow bolthead found at the Piedras Marcadas site and for two copper crossbow boltheads found at the Jimmy Owens site. We analyzed this data and generated an internal report. Subsequently, we submitted our findings to a professional consultant to obtain an independent appraisal.(2) The isotope ratios we measured, as well as a discussion of our analytical methods and conclusions, will be presented in a future issue of the New Mexico Historical Review. The following report pertains only to ten lead shot determined to be from a Spanish source location.
As a prelude to formal publication of our findings, we published a Research Note in the Winter 2010 issue of the New Mexico Historical Review.(3)
In September 2004 I attended a presentation concerning the Coronado Trail by historians Richard Flint and Shirley Cushing Flint. I learned that the trace of most of the trail followed by Spanish explorer Francisco Vázquez de Coronado during 1540-1542 was both unknown and the subject of vigorous historical debate. This puzzle inspired me to begin exploration in October 2004 for evidence of Coronado's route between Señora (Río Sonora valley, Sonora, Mexico) and Cíbola (Hawikku, New Mexico), and for the legendary Red House called Chichilticale. My study of the historical documents convinced me that identifying Chichilticale was integral to solving the mystery of the expedition's route across present eastern Arizona and western New Mexico. My ongoing effort has resulted in the likely discovery of both Chichilticale at the Kuykendall Ruins in southeastern Arizona and the camp that the expedition's Advance Party made on 23 June 1540. Two reports of my search for Coronado's route have been published by the New Mexico Historical Review. (1)
My 2009 report recommended that "lead isotope ratios should be obtained for the four lead balls" discovered at Kuykendall Ruins, and that these data should be compared to those from lead balls found at other known or suspected Coronado sites. (2) My team followed this analytical course by using Thermal Ionization Mass Spectrometry (TIMS) to measure isotope ratios of lead shot found at Kuykendall Ruins (Chichilticale), Doubtful Canyon (Advance Party's Camp 23 June 1540), Hawikku-Kyakima in New Mexico, and the Jimmy Owens site in the Texas Panhandle. The team added to its data the isotope ratios presented by Charles M. Haecker for two lead shot found at Piedras Marcadas Pueblo (LA 290, Mann-Zuris site). For comparative purposes, the team assembled a database using published data from Spain, Portugal, Mexico, New Mexico, the Rocky Mountains, the mid-continent United States, the Caribbean region, Central America, and the Mediterranean region. In this analysis, each region was treated as a conceivable source location for lead carried by members of the Coronado Expedition of 1540 – 1542 or by travelers who visited Coronado sites after 1542. The team compared its ratios to those in the database to conduct a robust analysis and to produce an internal report that assigned lead source provenience to the shot under consideration.
Subsequently, with the intention of obtaining corroboration or objection to the team's conclusions, lead isotope abundances (Pb204, Pb206, Pb207, Pb208) from its database and from its samples were analyzed independently by an unaffiliated party to determine numerically their similarity for the purpose of designating origin of the lead. (3) The team's isotope ratios of the lead shot, a discussion of its analytical methods and conclusions, and a report on the discovery of likely sixteenth-century iron artifacts in Doubtful Canyon, will be presented in an article published in a future issue of the New Mexico Historical Review.
The team's isotope analysis strongly supports the interpretation that shot composed of Spanish lead sources were present at five proposed Coronado Expedition sites: Kuykendall Ruins, Doubtful Canyon, Hawikku, Piedras Marcadas Pueblo and Jimmy Owens. The positive correlation between Spanish lead and these five individual Coronado sites suggests a nexus between the otherwise disparate locations. The common thread connecting all five sites of Spanish lead, the team believes, is the Coronado Trail. The results of the lead analysis has caused the team to modify its ever-evolving exploration model to include the likelihood of Spanish lead at Coronado sites and the probability that Spanish lead found within specific geographical corridors is diagnostic of the Coronado Expedition.
1. Nugent Brasher, “The Chichilticale Camp of Francisco Vázquez de Coronado: The Search for the Red House,” New Mexico Historical Review 82 (fall 2007): 433–68; and, “The Red House Camp and the Captain General: The 2009 Report on the Coronado Expedition Campsite at Chichilticale,” New Mexico Historical Review 84 (winter 2009): 1-64
2. Brasher, "The Red House Camp and the Captain General," 54.
3. TIMS measurements were conducted by geochemist Dr. Franco Marcantonio at the Radiogenic Isotope Geochemistry Laboratory in the Department of Geology and Geophysics at Texas A&M University. Isotope ratios for the Piedras Marcadas Pueblo are from Charles M. Haecker, "Tracing Coronado’s Route through Trace Element Analysis." Paper presented in the symposium, Between Entrada and Salida: New Mexico Perspectives on the Coronado Expedition. Charles Haecker and Clay Mathers, organizers. Society for Historical Archaeology Annual Conference, Albuquerque, New Mexico, January 12, 2008. Dr. Michael J. Rothman, Michael J. Rothman & Associates, LLC, a data analysis and visualization consulting company, in Hopewell, New York, conducted a comparative data analysis.
Because our findings are best displayed as colored, three-dimensional graphics, the New Mexico Historical Review has kindly agreed that we should also present them on this website.(4)
Artifacts discovered during the 2005 – 2008 Coronado Exploration Program included four lead balls from Kuykendall Ruins and one from Doubtful Canyon.(5) Our team desired to determine the sources of the lead from these five shot and to compare them to lead balls from other recognized Coronado sites.
We successfully obtained three samples from Hawikku, one from Kyakima and four from the Jimmy Owens site. All samples were solid lead. In February 2009 the LIR of these thirteen samples were measured using Thermal Ionization Mass Spectrometry (TIMS) by geochemist Dr. Franco Marcantonio at the Radiogenic Isotope Geochemistry Laboratory at the Department of Geology and Geophysics at Texas A&M University.
Exploration subsequent to 2008 resulted in additional finds suitable for LIR analysis. In January 2009, at and around the place known to the Apaches as Tsisl-Inoni-bi-yi-tu (Rock-white-in-water) in Doubtful Canyon, we discovered twelve lead shot.(6)
Dr. Marcantonio measured the 206, 207 and 208 LIR of these twelve pieces by Inductively-Coupled Plasma Mass Spectrometry (ICP-MS). Our comparative analysis indicated that four of the twelve lead balls found at Doubtful Canyon in January 2009 were likely composed of Spanish lead. Consequently, we measured solid samples of these four artifacts by TIMS. In summary, we measured by TIMS the lead isotopes of seventeen shot, with an additional eight shot measured by ICP-MS.
To complete our collection of lead isotope ratios from suspected Coronado sites, we included the LIR measured by Charles M. Haecker from two lead shot found at the Piedras Marcadas site.(7) This addition brought our total lead shot for analysis to twenty-seven.
"It is certainly true that archaeological scientists as a whole do not have a good understanding [of lead isotope ratio analysis.]"
Noel H. Gale and Zofia Anna Stos-Gale,directors
Isotrace Laboratory of the Research for Archaeology at the University of Oxford.
N. H. Gale and Z. A. Stos-Gale, "Comments… II," Archaeometry 35, 2, (1993) 253.
In order to compare our measurements to others, we assembled a database of LIR using published data from Spain, Portugal, Mexico, New Mexico, the Rocky Mountains, the mid-continent United States, the Caribbean region, Central America, the Mediterranean region, and the United Kingdom, these considered possible source locations for lead carried by members of the 1540 – 1542 Coronado Expedition or by post-1542 travelers to sites visited by Coronado. The database contained 2,018 identified samples plus our unidentified samples.(8)
My reading of publications concerning the application of LIR to identification of lead sources for artifacts suggested that the recommendation to use LIR for archaeological provenience studies might have originated with a 1967 publication by geochemists Robert H. Brill and J. Marion Wampler. Seven years later Brill published an application of the technique.(9) In 1992 and 1993 occurred a lively verbal joust amongst more than a dozen of the most respected researchers in the field of LIR analysis for archaeological purposes.(10) The discussion centered on methods of interpretation of isotope data, including the problems of overlapping populations, treatment of outliers, and selection of appropriate graphics. My analysis carefully and respectfully considered the comments and techniques of these researchers.
Of particular interest to me were the 1992 and 1993 comments of nuclear physicist E. V. Sayre and his co-workers. In 1992 Sayre et al wrote: "Lead isotope ratios should be handled as a three-dimensional problem. No single two-dimensional plot of the data fully or definitively defines the relationships between the measured specimens." Later that year researchers T. J. and C. L. Reedy wrote, "Sayre et al made multiple fixed plots to find a viewpoint that visually separates source groups. Modern statistical graphics programs now allow continuous rotation of point clouds to find an advantageous viewpoint. For two groups it should be possible to automate the process of finding a viewpoint that gives maximal apparent separation."(11)
The following year Sayre et al reiterated and added to his position: "The suggestion [offered by Reedy and Reedy] that one should use a computer program to rotate continuously the three-dimensional data through all arbitrary angles until one can visually observe, in a projection on a computer screen, the best separation between the specimens in question is an excellent one. The only reason we have not done so for our recent publications was that we did not have the required computer software readily at hand."(12) This caught my attention – at the time of their 1993 publication, Sayre and his co-workers were associated with the Conservation Analytical Laboratory at the Smithsonian Institution, and Sayre himself had previously been associated with the Brookhaven National Laboratory. How could he not have the software on his or a co-worker's or colleague's desktop? My own experience included rotating data points on a Macintosh computer beginning in 1988.
In 1981 in Glenwood, New Mexico my research partner and I began development of a stand-alone workstation, that is, a computer system independent of a mainframe computer.(13) Our design was for petroleum exploration. We called our system an ExplorationStation. Ours was an interdisciplinary effort between an electrical engineer and a petroleum geologist. Our initial equipment was a Radio Shack Model II, a 64K computer with a single 8" floppy-disk drive. In 1982 Radio Shack released its Model 12 with double disk drives. We "stepped-up" the Model 12 by installing a Motorola 68000 chip and adding extra RAM so as to produce 1 MB of random access memory. Although there was no commercial Fortran compiler for such a "step-up," we secured one through channels in Massachusetts. In 1984 Radio Shack released its Model 16. This computer followed our lead by also using a Motorola 68000 chip. Using the Radio Shack system, we wrote our own software, including a database manager, a statistical package utilizing multivariate analysis, regression analysis and discriminant function analysis, and a contouring package.
In mid-1987, Apple Computer, Inc. released the Macintosh II. This equipment utilized a Motorola 68020 chip and was the first full 32-bit microcomputer. We opted to shift from the Radio Shack system to the Macintosh. To expedite this we joined the Apple developer group and began to convert our CP/M interface to the Mac interface. This required conversion of our software, which we accomplished by March 1988.
We enhanced our ExplorationStation by adding auxiliary commercial software. Two such programs were MacSpin and DataDesk. MacSpin displayed data in three-dimensional form and allowed the user to "spin" data, that is, to rotate it in three dimensions and to animate data points. MacSpin was first developed on a VAX computer at the Stanford Linear Accelerator, but was re-coded in the mid-1980s to operate in the Macintosh environment. DataDesk likewise offered the capability to "spin" the data. We served as a development and test site for both MacSpin and DataDesk. Rotating data points in three-dimensional space was common practice for us by early 1988. Below is an example of a rotated three-dimensional scatterplot created on DataDesk software. All graphics presented in this section of the website, including two-dimensional scatterplots and three-dimensional rotation plots, were produced on the commercially available 1988 version of DataDesk.
Two-dimensional scatterplots are most commonly presented in the literature pertaining to archaeological lead source identification. The figure below is a two-dimensional 207/204 vs 208/206 scatterplot generated from our database. For my purposes, the first ratio given is the x-axis and the second the y-axis
I selected the specific, non-traditional 207/204 v 208/206 isotope ratios to plot because they generally demonstrate separation of the data points into four regional populations. In 1992 Sayre et al recommended such an approach to plot selection:
"It is not practical to include large numbers of plots in a publication, therefore a brief consideration of the reasons for the selection of the ones that were included might be in order. Two-dimensional scatter plots are, of course, projections on to a plane of the three-dimensional distributions of the isotopic data. If two groups of specimens truly overlap in the full three-dimensional space there can be no such two-dimensional projection of them that shows them to be separated. Hence, if any two-dimensional scatter plot shows the groups to be separate they must truly be separate in the three-dimensional volume that fully defines them. We, therefore, use such plots as providing sufficient proof that such separation exists between groups or between individual specimens and a group, selecting for publication whichever plot most clearly shows the separation. Conversely, no matter how widely separated two groups may be in the three-dimensional space one can always find two-dimensional projections in which the two groups falsely appear to overlap."(14)
Sayre expanded his recommendation the following year:
"No individual two-dimensional scatter plot of three-dimensional stable isotope ratio data can correctly display these data… [With respect] to the choice of two-dimensional scatter diagrams to display the three-dimensional isotope ratio data, [here are] some general comments… Overlap between groups of specimens or the occurrence of specimens in close proximity to each other on any one such two-dimensional scatter plot does not provide adequate evidence that the groups or specimens are actually even close to one another in the full multi-dimensional space. In fact, no matter how widely separated the groups or specimens are in the full dimensional space one can find a two-dimensional projection in which they appear to coincide. Conversely, if specimens truly do lie close to each other or if the groups do overlap in the multi-dimensional space it is not possible to find a two-dimensional projection of the data in which they will appear to be separate… Therefore, if any two-dimensional scatter plot of the data shows specimens or groups of specimens to be well separated from each other, they must be at least as well separated in multi-dimensional space, but these same well-separated groups or specimens may well falsely appear to be overlapping in some such plots… Because no two-dimensional scatter plot can provide a truly correct relationship for these three-dimensional data, it is useless even to consider establishing a single conventional two-dimensional plot for such representations. If one is attempting through such a plot to illustrate separations between specimens, it is only good common sense to use the plot that best shows the separation.
"We hope that our remarks have succeeded in convincing most readers that stable lead isotope ratios are three-dimensional data sets that require multivariate data handling to be properly interpreted… No treatment exclusively involving only two of the three ratios can produce a reliable interpretation of the three-dimensional data. Therefore, any attempt to establish a conventional two-dimensional scatter plot as a standard basis for interpreting these data is intrinsically misdirected."(15)
Clearly, Sayre et al were convinced that a three-dimensional analysis would prove productive for identification of source locations. The three dimensional 206/204, 207/204, 208/204 plot below illustrates better separation between Spanish and Mesoamerican populations than the two-dimensional 207/204 v 208/206 plot presented above.
For data analysis purposes, Mesoamerica is an analytical term to describe the Middle New World according to our lead isotope ratio database. This includes all of modern Mexico, Central America, and the Caribbean Islands. Our Mesoamerica includes, but is not limited to, the region traditionally called Mesoamerica by the anthropological community.
My objective in using LIR was to assign an unknown lead artifact to a known population of lead source material. To accomplish this I explored the data to see if the lead shot samples we had measured could be shown to belong to a recognized source location population of lead. For an overview of how Southwestern and Mesoamerican archaeologists employed LIR to approach provenience puzzles, I studied four papers published in 1996 or later, presuming that these would offer the current state of research.(16)
Archaeologist Dorothy Hosler and geologist Andrew Macfarlane, in 1996, compared LIR measured in Mexican copper artifacts to LIR associated with copper ore deposits in Mexico. They presented their findings on 206/204 vs 208/204 scatterplots, writing that such ratios are "geological convention." The plots showed considerable overlapping of ore fields, so the authors used historical and archaeological evidence to assign sources to the samples measured from the artifacts.(17)
In 2007 archaeologist Deborah L. Huntley and her co-workers compared LIR from ore sources in the Río Grande valley of New Mexico to LIR obtained from glaze paints on Río Grande Glaze Ware made at two pueblos in the region in order to identify ore sources. Huntley et al published their findings on 206/204 vs 207/204 and 206/204 vs 208/204 scatterplots. They wrote: "The first step in the data analysis was to plot ore and glaze sample lead isotopic ratios in two-dimensional space using various combinations of stacked plots of pairs of ratios… Since 206 Pb, 207 Pb and 208 Pb isotopes are normalized to 204 Pb in geochronology and in characterizing lead ores, we have used these ratios… In addition, we examined plots of other isotopic ratios (e.g., 206 Pb/207 Pb: 208 Pb/207 Pb) to evaluate group membership, as recommended by Sayre et al [in 1992]… we relied on visual examination of multiple bivariate plots to identify general trends for ore source… In this paper we present those plots that we feel best illustrate group separation." Huntley et al, like Hosler and Macfarlane, observed overlap of ore source fields, and likewise employed "other lines of evidence" to assign sources to samples.(18)
Also in 2007, archaeologist A. M. Thibodeau and her co-workers compared LIR from galena excavated at the La Isabela, Hispaniola site to LIR in a database including measurements from the Caribbean and Spain. As had Huntley et al, they presented their findings on 206/204 vs 207/204 and 206/204 vs 208/204 scatterplots. Thibodeau et al used "historical circumstances, geographical proximity and mineralogy" to assign a source location to the La Isabela galena.(19)
In January 2008, archaeologist Charles M. Haecker presented a paper at the Society for Historical Archaeology conference comparing LIR from artifacts found at the Piedras Marcadas and Jimmy Owens archaeological sites to LIR from Mexico and Spain. Following suit, his findings were presented on 206/204 vs 207/204 and 206/204 vs 208/204 scatterplots.(20)
None of these researchers demonstrated or even intimated that rotation of three-dimensional data had been employed. Only Huntley et al cited the work of Sayre et al, and they purposely acknowledged that bivariate, not trivariate, plots were used.(21)
All four of the researchers employed the technique of enclosing data points representing individual ore bodies inside a perimeter. I have declined to use this technique because the population and distribution of individual ore bodies are almost certainly unknown due to incomplete sampling. Sayre et al fittingly pointed out that the size and distribution of an ore body population must be defined by the data, not by perimeters drawn by the interpreter. In the absence of complete data, I aver that any perimeters enclosing data clouds are irrelevant.
Sayre et al addressed the issue of incompleteness of source group populations: "The major evidence that many source groups have not been fully characterized is that within the specimens from every source region from which a reasonable large number of specimens have been analyzed there have been at least a few highly deviant specimens… The specimens in question may be samples of significantly different ore deposits in the area, which are not yet explored, or highly improbable members of the ore groups that have been characterized… There is, of course, the additional possibility that further sampling of the region might produce yet other specimens that show more, significantly different, deviations." These possibilities illustrate the caution encouraged by Sayre and his co-workers that "one can argue whether a correct choice of specimens has been made to represent the population."(22) In the absence of conclusive evidence of population size and distribution, I have elected to avoid enclosing data clouds in perimeters on two-dimensional plots, regardless of the statistical or interpretative technique used to generate such perimeters. On any account, my reliance on three-dimensional and four-dimensional analysis overrides any necessity to apply such perimeters.
Contrary to the observed disregarding of three-dimensional analysis for provenience studies by the four researchers cited above, I employed three-dimensional rotation of LIR data points as the very foundation of my lead isotope ratio analysis, and the technique significantly influenced my sample provenience identification conclusions. I employed three-dimensional rotation as a non-parametric method for exploring the dataset in search of likely populations to which our twenty-seven lead samples of unknown origin belonged. My initial objective was to observe motional and proximal relationships rather than to measure distances in a parametric manner.
My approach employed point rotation to visually determine which known samples irrespective of specific source location or whether more than one source location was involved, appeared to move through three-dimensional space with a specific unknown source location sample. Once I had identified the known points that traveled with an unknown point, I reduced the dataset to include only members of the source location populations represented in the "equi-movers." I considered these known points plus the unknown point to represent a single population based on similarity of motion and proximity to one another. Because I was rotating ratios, I did not attempt to introduce a parametric component by measuring distances between points.
I was not concerned if multiple source locations were represented in the population of equi-movers. The verbal joust of 1992 – 1993 included a comment by Budd et al warning that researchers might be expecting too much from the data: "There has been a marked reluctance on the part of some workers to accept that ore source fields may truly lack the resolution that archaeometallurgists would like."(23) My experience with geological and petroleum data caused me to agree with Budd et al. Working with oil and gas exploration data has convinced me that oftentimes data does not contain sufficient information from which to draw precise conclusions, although the same data does indeed allow accurate, general inferences.
The historical setting of sixteenth-century Spain included mining activities dating from the Phoenicians, and included mining operations during Carthaginian, Roman, Vandal, Visigoth, and Arabian times. Active mines in the 1500s provided lead for lining ship's hulls, conduits, covers and armaments. Lead from Spain was available to travelers reaching the New World in time to have accompanied Coronado on his expedition. For reasons based on Spanish and Mexican history, and on geography, I reasoned that identifying Spanish lead would produce the resolution I needed to differentiate some samples of Coronado Expedition lead shot from other lead. Consequently, I was concerned with being able to discriminate between larger populations, such as between Spanish samples and Mexican samples.
I also reasoned, based on history and geography, that if lead from Tarascan sources could be differentiated from lead sources along the Camino Real, that I could support a Coronado source for Tarascan lead found at suspected Coronado sites. Conversely, lead from Camino Real sources at suspected Coronado sites would indicate a most likely post-Coronado source. This objective required discriminating between smaller populations possessing similar isotopes, such as between individual Mesoamerican mines or mining regions.
At the conclusion of my exploratory, non-parametric, visual, three-dimensional rotation analysis, I assigned as precisely as my confidence permitted a source location identification to each of the twenty-seven lead shot samples of unknown source origin.
Researcher Brill and his co-worker's early attempt at interpretation produced a fundamental observation. "One disadvantage of this method is that although individual mines or deposits of galena ore are characterized by definite isotopic ratios, these ratios are not necessarily unique and mines from different areas… may have similar isotope ratios."(24)
This caveat seems to have been lost during the subsequent years as evidenced by the coinage and unfortunate usage by many researchers of the term "fingerprints" when referring to lead isotope ratios. For example, Huntley described her 2007 study of prehistoric glaze paints as "… based on the principle that lead ores… can be distinguished by 'fingerprints' of the ratios of the four stable isotopes." Geochemist Malin E. Kylander wrote that by "using isotopic fingerprints it is possible to make detailed source assessments;" archaeologist Judith A. Habicht-Mauche wrote that "Lead ores from distinct geologic deposits… can be 'fingerprinted' by their stable lead isotopic composition;" geochemist Sanghamitra Ghosh refers to isotopes as "fingerprinting tools;" geochemist E. Pernicka refers to the "isotopic fingerprint."(25)
Professor Noel Gale and his wife Dr. Zofia Stos-Gale, as well as Dorothy Hosler and Andrew Macfarlane presented a different characterization of LIR application. Gale and Stos-Gale wrote, "A fundamental point… is that lead isotope analyses allows one unequivocally to exclude a metal object from having come from a given ore deposit…" Hosler and Macfarlane wrote, "Lead isotope data are most effective in excluding certain deposits as likely source areas. Positive source identification is more difficult because not all deposits can be sampled."(26) My interpretation employed conclusions of source locations based on exclusion as well as inclusion.
The results of my exploration-oriented, non-parametric, equi-mover analysis resulted in confident source ore identification of the twenty-seven unknown lead shot samples based on visual and relational examination. However, my technique lacked parametric measurements. To obtain a parametric analysis, chemist and data analyst Dr. Michael J. Rothman designed a procedure we defined as the "Rothman Method" (RM) to accurately measure how near or far unidentified samples are located relative to identified population centers.
"As we try to determine the origin of a sample, we look in detail at what the lead itself can tell us. Specifically we are trying to establish similarity between lead from unknowns and lead from various reference samples. Previous authors have approached this problem by looking at points in ratio-space. See for example graphs where the ratio of measured abundance of Pb207/Pb204 is plotted versus the ratio of measured abundance of Pb206/Pb204.
"We have taken a different approach. Similarity in this circumstance is a matter of distance, and distance is a function of the definition of the axes and the choice of a specific metric. Why should we choose these various ratios as the axes? It seems that they have been selected because these are the numbers that are typically reported by the laboratories who analyze lead isotope abundances. We have taken a more deliberate approach. Given the complex geophysical mechanisms that produce the lead isotope ratios of each sample, we decided that each isotope abundance should be treated as equally important. A simple way to do this is to transform the data into a 4-dimentional z-space, where the axes are the number of standard deviations from the mean for each of the four isotope abundances: Pb204, Pb206, Pb207, and Pb208.
"The means and standard deviations were computed over the entire database, with one exception. When we looked at the distributions of each isotope we found bimodal distributions. Some samples from central areas of the United States contain what has been termed “anomalous lead.” Their isotope abundances are far removed from those of the samples from the remainder of our database. We therefore chose to exclude these samples when we computed means and standard deviations and z-values."
The "Rothman Method" (RM) is outlined below:
1. I computed the mean and standard deviation for abundance of Pb
, Pb206, Pb207 and Pb208 for the selected dataset.
2. For each entry in the dataset I computed the number of standard deviations (or "z-units") that sample was from the mean for each of the 4 isotopes. Each point is thus located in a 4-dimensional Z-space.
3. I computed the Euclidian distance from each entry in the database to each of the 27 unknowns. This is just the square root of the sum of the squares of the differences between the z-value of each isotope of an unknown and each reference database entry.
4. I computed the mean distance by SET (geographic populations, such as Aegean, Southeast Spain, etc) to each of the unknowns.
The database as collected contained 2,018 individual samples worldwide plus the twenty-seven unidentified samples. The individual samples of known provenience were assigned to one of seventy-four geographical regions included in a definition named Set 1.
During my visual examination I experimented by lumping the seventy-four regions into different sets, such as by combining six southeastern Spanish regions (Cartagena, Alhamilla, etc) into a single region called Southeast Spain.
Using each of my experimental sets, Dr. Rothman produced a matrix displaying distances separating the mean isotope centers of each of the regions. We interpreted the matrix as illustrating which regions are most isotopically similar or dissimilar to one another. In the end we settled on Set 4 – this study definition satisfied the objective of discriminating between Spanish and New World lead isotopes.
We interpreted the matrix as illustrating wide separation of isotopes by region. The single notable exception to wide isotopic separation of regions was the similarity of Aegean Sea isotopes to Southeastern Spain isotopes, but I did not view this as problematic because of historical considerations.
Application of the RM to Set 4 produced confidence that we could distinguish Spanish lead sources from those of other regions of interest. The parametric RM analysis identified ten samples as being Spanish lead, which matched the conclusions of my non-parametric examination of the twenty-seven samples.
The Set 4 matrix displays three instances (RM# 3, 6, 10) where an Aegean source is a possibility for the lead shot under consideration. Anthropologist Carroll L. Riley kindly provided me a brief description the sixteenth-century historical setting pertaining to Spain and the Aegean Sea.
"The war between the Ottoman Turks and the Spanish-German Empire of Charles V and Phillip II went on from the early sixteenth century. The Turks controlled most of the eastern Mediterranean and for much of the century Spain was on the defensive. They did manage to occupy Tunis in 1535 but suffered a devastating defeat at Algiers in 1541. Spain did control Malta through the Knights of St. John and the Turks under Soliman I decided to seize this crucial group of islands in 1565. In an horrific battle, the Turkish army and navy were beaten off.
The final great battle was at Lepanto in 1571 when an allied Christian fleet under overall Spanish command gathered at Messina and proceeded eastward to the Gulf of Lepanto where after a seesaw battle the Turkish forces were defeated. From then on Turkish overall naval power declined.
I don't say there was not trade going across the Mediterranean during the Coronado period but it made conditions rather more difficult than otherwise."(27)
Because of these aforementioned historical, plus geographical factors, I elected a Spanish source as most likely for lead shot RM#s 3, 6, and 10. My interpretation technique is consistent with that advocated by other researchers. Hosler recommends that where overlapping populations exist the researcher "can eliminate some sources on the basis of historical and other evidence." Huntley insists that researchers must "rely on other lines of evidence to bolster an interpretation." Thibodeau argued that "the combination of the specific historical circumstances and the geographical proximity of lead ore deposits to the port of departure" influenced her conclusion that lead found at La Isabela in the Dominican Republic was from Spain.(28)
Application of the RM to Set 1 produced a more detailed look at specific geographical locations.
The RM provided a parametric analysis that resulted in corroboration of my non-parametric visual observation technique and instilled confidence in my conclusions.
We wanted to rotate in three-dimensions the Euclidian distances calculated by Dr. Rothman. Problematic to such rotation was the four-dimensional nature of these distances. Our solution to this problem was to compute the principal components of the samples in our dataset.
"Although we have defined a 4-dimensional z-space for our analysis, there are only 3 independent dimensions; the sum of the four abundances is constrained to be 100%. In order to most clearly show differences in the samples, we used software from the SAS Institute, Cary, NC, version 9.1, to compute the three principal components. The eigenvectors and eigenvalues are shown below. The first principal component accounts for 90% of the variance in the dataset and is useful in and of itself in distinguishing the differences between samples.
Principal Components (eigenvectors)
Note that the elements of each principal component are the z-values of that isotope’s abundance.
Prin1 = 0.50*Pb204 – 0.52*Pb206 + 0.51*Pb207 + 0.46*Pb208
Prin2 = 0.41*Pb204 + 0.09*Pb206 + 0.42*Pb207 - 0.80*Pb208
Prin3 = 0.76*Pb204 + 0.26*Pb206 - 0.59*Pb207 + 0.11*Pb208
E1= 3.64 (91.0% of the variance)
E2= 0.32 (8.5% of the variance)
E3= 0.02 (0.5% of the variance)
Using the principal components calculated by Dr. Rothman, I rotated in three dimensions Prin1, Prin2, and Prin3. The results corroborated the non-parametric visual observation technique and the Rothman Method technique, elevating confidence in our conclusions.
Of twenty-seven lead balls examined, we concluded that ten were composed of lead shot from a Spanish source.
Readers should keep in mind that my conclusions were controlled by the database I assembled. It is possible that one or more of the ten samples I have concluded to be from a Spanish source location are actually from a source not represented in the database, while at the same time possessing isotope ratios similar to those of Spanish samples. Thibodeau recognized this dilemma by writing that "isotope data alone cannot rule out the possibility that the [samples she concluded to be from Spain] came from some more distant part of Europe, the Mediterranean or North Africa."(29) With the hope that my database contains the information needed to truly determine the source of the lead shot found at five presumed Coronado Expedition sites, I present below the results of my analysis by non-parametric visual examination.
Reviewers are encouraged to keep in mind that the principal objective of this presentation is to show that the samples under consideration are Spanish lead, with a secondary objective of determining a more precise source location within Spain.
The first isotope measurement reported to me by Dr. Franco Marcantonio was from a 0.31 caliber lead shot found west of Kuykendall Village on 10 March 2008 on the north bank of Ruins Arroyo on the Gill property.(30) This lead piece represents the westernmost sixteenth-century artifact found during our exploration of Kuykendall Ruins from January 2006 through 11 April 2008 and thereby establishes the western limit of demonstrated Coronado presence at Chichilticale. Explorers seeking to offset this artifact discovery should proceed to the west along both sides of Ruins Arroyo because we have not yet searched there.
On 16 February 2009, Dr. Marcantonio identified the piece without need of our yet-to-be-constructed database – he simply compared it to a list of data in a book from his library. He casually, but quite confidently, declared that the lead source was obviously the Río Tinto region of southwestern Spain.(31) Subsequent detailed analysis proved him to be correct.
I translate the phrase Rio Tinto as "Red Tinted River." The Gill ball is off-spherically-shaped and is a mottled red-gray color. When I found it by "swinging sticks" (handheld metal detector), so employed at that spot because of difficulty to survey by the Blennert Sled method, I held the shot in my hand but could not see it. The color blended perfectly with the sand, silt and gravel surrounding it. My magnet did not attract a target, so I knew that the "hit" was not iron. Only by isolating with my pinpointer the "that's a pebble isn't it?" object did I find the shot. Only by poking it with a sharp point did I see the silver sheen that tells the explorer that the object is lead.
A few meters from the Rio Tinto lead ball I found an Apache iron "arrowhead" made from a barrel strap. John Blennert identified the object as one commonly found and generally termed a "trade point." For motive of a complete record and full disclosure, I report the find and Blennert's identification. As to whether an Apache dropped the Spanish Rio Tinto lead shot at the same spot, I will leave such speculation to the reader, begging only that the probability of such an occurrence be considered.
For a measured relationship of the Rio Tinto ball to other artifacts, interested readers should consult my Map 2 published in the Winter 2009 issue of the New Mexico Historical Review or refer to the New Mexico Historical Review 2009 section of this website.(32)
Following is a gallery of graphics supporting my conclusion that the Kuykendall Gill shot is composed of lead from the Rio Tinto – Huelva regions of southwest Spain.
On 1 April 2007 on the Cimarron Ranch, east of prehistoric Kuykendall Village, on the north bank of Ruins Arroyo, on the surface of a wind-swept bald spot (blowout) littered with potsherds and lithic chips, by the "swinging stick" method, on a windy, unseasonably cold afternoon, I found a 0.33 caliber lead ball (Cimarron 57-01, from latitude 31.52.57). Subsequent isotope ratio analysis showed this shot to be from Southeastern Spain. Ninety-four feet from the Spanish lead ball, I had found on the surface that same day, less than an hour before, part of an iron crossbow bolthead ferrule. Surrounding the blowout remained piles of broken, burned rock. The spot was almost certainly a campsite before and after Coronado. Stumps of ancient trees nearby attest that Ruins Arroyo flowed beside the spot before being captured miles to the east by Turkey Creek.
Nine days later, campmaster Gordon Fraser, along with my wife Karen and I, returned to the spot. The day was hot and clear, and while the sandhill cranes had departed, the buzzards had returned to soar in the thermals over Sulphur Spring Valley. One hundred fourteen feet west of the Southeastern Spain lead shot I had found, Gordon Fraser, by "swinging stick," discovered 0.49 caliber lead ball 57-02. Much later, isotope ratio analysis showed this second shot to be from the Mid-continent USA, likely Missouri. Although we found nothing else of interest that day, the event represented the final hunting day of the 2008 season – on the 11th of April we steered the basecamp coach out of Cochise and took her home to Glenwood, New Mexico. It marked the end of an era.
I have previously reported that Turkey Creek captured Ruins Arroyo "after 1800 but before the earthquake of 1887," and that until that stream capture "Ruins Arroyo [was] the principal watercourse, enjoying pools and running water that supported trees for fuel wood, and exposed stream cobbles for use with fires."(33) My geological interpretation is supported by shot 57-02 and by history. The United States lead source and the spherical shot point to an American ball-shot gunner visiting the Cimarron site prior to or shortly after widespread use of conical bullets began during the American Civil War. Even if shot 57-02 reached the Cimarron site via a Native American, the Missouri lead was likely brought into the Southwest by an Anglo visiting the region after the Jackson party left Santa Fe, New Mexico and traveled through Apache Pass in 1831. Respecting the campsite itself, if an American visitor left the ball there, he had come to an old and favored camp, one that was previously enjoyed by Native Americans and Captain General Francisco Vázquez de Coronado, and his visitation attests to the duration of the Cimarron spot as a sojourn.
The following gallery of graphics offers support to my claim that the Cimarron lead shot is from Southeastern Spain.
The 0.54 caliber lead ball I found on 9 March 2005 on the north side of Doubtful Arroyo directly on the Butterfield Trail in Cozzens Valley at Doubtful Canyon was the first "real artifact" I found with my newly acquired White's MXT metal detector. This find initiated my procedure for "bagging and tagging" potential Coronado residuals. I captured the GPS of the ball location, then buried in the same hole from which I had extracted the artifact a plastic vial containing a modern nail, and recorded the date and description of the find, as well as named any witnesses present. My dog Zippho was the only witness present that clear day. I reported this find in two issues of the New Mexico Historical Review.(34)
My first attempt to use geochemistry to help identify sixteenth-century lead balls, including this shot, was in 2005 and has been previously reported:
"[Charles M.] Haecker recalled that researchers had studied lead-ball composition to discriminate between American and Mexican militaries at a site in Texas. He obtained three lead balls from [Jonathan E.] Damp [at the Zuni Cultural Resource Enterprise] that had been found at Hawikku (Zuni) and one lead ball found at Kyakima (Zuni) for comparison purposes. I arranged for a metallurgic analysis of the lead balls from Zuni and the one from Camp 23 June [Doubtful Canyon Cozzens lead shot]. Ellery E. Frahm conducted this examination at the University of Minnesota Electron Microprobe Laboratory in Minneapolis. Frahm provided data that I used to generate graphics useful for non-parametric comparisons of the five lead-ball compositions. The lead ball [Doubtful Canyon Cozzens] from Camp 23 June (0.535 caliber) and Hawikku FS138 (0.48 caliber) correlate on individually measured metallurgic composition."(35)
Regrettably, the measurements taken at the Minnesota laboratory suffered contamination from silicon-carbide polishing discs. On 6 December 2007 Frahm reported this contamination on the internet.(36) The contamination renders the correlation mentioned above invalid. Moreover, because of this contamination and the resulting uncertainty, we cannot offer here the chemical composition of the Doubtful Canyon Cozzens, the Hawikku, or the Kyakima lead shot.
The following gallery of graphics offers support to my claim that the Doubtful Canyon Cozzens lead shot is from Alhamilla, Southeast Spain.
Doubtful Canyon lead shot numbers 3, 4, 9, and 10 were all found during a three-day search of 12-14 January 2009. Kuykendall Ruins veterans John Blennert, Gordon Fraser, Loro Lorentzen and I conducted a Blennert Sled and swing-stick exploration on private land in the canyon.
We found 0.39 caliber lead shot number 3 on 12 January 2009 with a Blennert Sled. This discovery lay about 1.5 miles east of where Doubtful Canyon Cozzens lead ball was found in 2005. Like the Cozzens ball, lead shot number 3 was found on the north side of Doubtful Arroyo directly on the Butterfield Trail. That same day, within a few steps of the number 3 ball, we found sixteenth-century iron artifacts.
Our isotope analysis shows that Doubtful Canyon lead shot number 3 is from the Catalonia Coastal Range of northeastern Spain, along the Mediterranean coast near Barcelona.
The following gallery of graphics offers support to my claim of this source location.
On 13 January 2009 our four-member search team composed of Kuykendall Ruins veterans found Doubtful Canyon 0.39 caliber lead shot number 4 while swinging-stick about a tenth of a mile (534 ft) east of shot number 3. The ball was on private land on the north side of Doubtful Arroyo, on an escarpment above the Butterfield Trail, which at that point is in Doubtful Arroyo itself.
Our isotope analysis shows that Doubtful Canyon lead shot number 4 is from the Catalonia Coastal Range of northeastern Spain, along the Mediterranean coast near Barcelona. This is the same source region as Doubtful Canyon 3 lead shot.
The following gallery of graphics offers support to a Catalonia source location for shot number 4.
Doubtful Canyon 0.39 caliber lead shot number 9 was found on 14 January 2009. John Blennert discovered it while swinging-stick on private land on the cutbank downslope from the Butterfield Trail, which at that location was on the south side of Doubtful Arroyo. Near the number 9 shot Blennert found an iron sixteenth-century artifact. The number 9 ball and iron piece were found directly across the arroyo from where Doubtful Canyon lead shot 4 was discovered on the escarpment above the watercourse.
Doubtful Canyon lead shot number 9, as concluded from our isotope analysis, is from the Catalonia Coastal Range of northeastern Spain, along the Mediterranean coast near Barcelona. This is the same source region as Doubtful Canyon 3 and 4 lead shot.
The following gallery of graphics offers support to a Catalonia source location for shot number 9.
The expended Doubtful Canyon 0.43 caliber lead shot number 10, found by Blennert Sled on 12 January 2009, was presumably fired and subsequently struck an object that flattened it on one side. We found the number 10 shot on private land within steps of unfired lead shot number 3 at a location on the north side of Doubtful Arroyo directly on the Butterfield Trail, where we also discovered sixteenth-century iron artifacts. The proximity of the spent shot to other artifacts suggests that it was transported as such, that is, that it was not fired at the site where found.
We concluded from our isotope analysis that Doubtful Canyon lead shot number 10 is from the Catalonia Coastal Range of northeastern Spain, along the Mediterranean coast near Barcelona. This is the same source region as Doubtful Canyon 3, 4 and 9 lead shot.
In total, at Doubtful Canyon we found four shot having a Catalonian lead source location. Three of these lead shot were 0.39 caliber, the fourth being 0.43 caliber. Associated with three of these lead shot we found iron sixteenth-century artifacts.
The following gallery of graphics offers support for shot number 10 being Catalonian lead.
Archaeologist Jonathan E. Damp kindly provided for our analysis solid samples of three shot found at Hawikku and one found at Kyakima. Included in these was 0.56 caliber lead shot Hawikku 29. Damp reported that "approximately 1/3 of the ball" was broken away, and he speculated that this was due to a possible "pewter content." Damp also reported that Hawikku 29 "may be corroded from blood," this based on the opinion of a Zuni Police Department officer.(37) Damp's report did not include the exact location of the Hawikku 29 find, so I cannot offer information as to whether or not other artifacts were found nearby. It is doubtless, however, that sixteenth-century iron artifacts were found at Hawikku.
Our analysis determined that the Hawikku 29 lead shot falls within the lead isotope ratio population of the Cabo de Gata region along the Mediterranean coast of southeastern Spain.
The following gallery of graphics offers support to my claim that the Hawikku 29 lead shot is from Southeastern Spain, most likely from Cabo de Gata.
On 7 November 2008 my wife Karen and I visited the Floyd County Historical Museum in Floydada, Texas to obtain samples of lead from four shot found at the Jimmy Owens site. Following the instructions of Dr. Franco Marcantonio, and using the chemicals and containers provided by him, we swabbed the four balls with a dilute HNO3 ~10% solution, retaining for later measurement the swabs in separate sterile bags. In addition we extracted a solid sample by using four separate sterilized pins to prick away a tiny piece of solid lead from where we had swabbed, subsequently placing these solid pieces in four individual sterilized containers.
Dr. Marcantonio measured by Thermal Ionization Mass Spectrometry (TIMS) the isotope ratios of only the solid lead samples taken from the four shot, holding in abeyance any measurements of lead on the swabs.
Our analysis determined that 0.43 caliber lead shot AT-02, collected at the Jimmy Owens site on 26 March 2004, was from Southeastern Spain. The gallery of graphics below provides support for our interpretation of this source location for the lead in this shot.
Charles M. Haecker kindly and generously offered us the lead isotope ratios and graphics he presented in January 2008 at the Albuquerque, New Mexico conference of the Society for Historical Archaeology. The measured isotope ratios included those from two lead shot found at Piedras Marcadas, a Coronado site on the Río Grande at Albuquerque, New Mexico. We used these measurements to determine that the source location of the Piedras Marcadas "Small" lead shot is Ossa Morena in southcentral Spain. The following graphics support our interpretation.
Haecker presented his own graphics at the 2008 conference.
The scatterplot generated by Haecker shows the "Small" lead shot found at Piedras Marcadas as a black triangle located at approximately 17.7 on the 206/204 axis. The triangle is located near, but not within, a South Central Iberia population of isotope ratios represented by a blue color contained within an irregular perimeter apparently drawn in a non-arbitrary manner.
Our analysis revealed that Haecker used data published by Santos in 2004, but did not include data published that same year by Tornos.(38) We repeated the 206/204 vs 208/204 Haecker plot and added the Tornos data to show the impact of the additional data.
By adding the Tornos data and employing rotation of a three-dimensional plot, we concluded that the Piedras Marcadas lead shot is a member of the Ossa Morena isotope ratio population.
Our scatterplots showed eleven samples associated with the Piedras Marcadas Small lead shot sample. We applied the Rothman Method of Euclidian distances to these eleven samples to show their parametric proximity to the Piedras Marcadas ball.
The map presented by Tornos shows the topographic proximity of the eleven samples and illustrates that all are geologic members of the Cadomian volcanic sequence.
Notice that Haecker followed the convention of enclosing data clouds within an irregular, non-arbitrarily drawn perimeter – he girdled the Southeast Iberia, South Central Iberia and Mexico populations. Consequently he did not assign the Piedras Marcadas Small lead shot to southcentral Spain, even though it was proximally near. Had Haecker not drawn his perimeter, he might have interpreted the data differently. Sayre cautioned against interpretation by perimeter: ""One sometimes encounters the claim that if a specimen does not lie within a source area defined by enclosing the measured specimens from a mining region [that] it cannot possibly have been derived from ores of that source area."(39) The Piedras Marcadas example provides a sharp illustration of the pitfalls of researchers who enclose data clouds.
As discussed in the section "Recent Application of LIR by Archaeologists," I have declined to enclose data clouds within perimeters. Our own research showed that data cloud perimeters, even if drawn in a non-arbitrary manner, change shapes and areas dependent upon two-dimensional scatterplot selection, thereby obfuscating attempts to identify an unknown sample. Sayre observed the same phenomena:
"We routinely make 12 two-dimensional scatter plots of our data… It is impressive to see for oneself how different the relationships between data points can look in these different plots… It behooves us to use methods that take into account the uncertainties that are involved in working with less than optimumly large data sets…"(40)
The Piedras Marcadas example also illustrates the impact of missing or incomplete data. One of the reasons Sayre objected to perimeters was that he felt "that most source areas have not been characterized sufficiently well" enough to warrant perimeters, and that "further sampling might produce yet other samples."(41) Although Haecker had data from southcentral Spain, he did not have all the data, and when additional data was provided, the data point pattern was altered to an extent demanding revision of conclusions, resulting in the Piedras Marcadas shot being shown to belong to southcentral Spain. This eventuality haunts all of us.
Despite the above note of caution, I asked mathematics professor Dr. Oleg Makhnin to prepare a plot utilizing a perimeter for the Piedras Marcadas shot and the eleven Ossa Morena samples used in the Rothman Method Euclidian distance analysis. Below is his kindly offered contribution.
1. Nugent Brasher, "The Red House Camp and the Captain General: The 2009 Report on the Coronado Expedition Campsite of Chichilticale," New Mexico Historical Review 84 (winter 2009): 54.
2. Charles M. Haecker, "Tracing Coronado’s Route through Trace Element Analysis." Paper presented in the symposium, Between Entrada and Salida: New Mexico Perspectives on the Coronado Expedition. Charles Haecker and Clay Mathers, organizers. Society for Historical Archaeology Annual Conference, Albuquerque, New Mexico, January 12, 2008. Dr. Michael J. Rothman, Michael J. Rothman & Associates, LLC, in Hopewell, New York, generously provided respected counsel to our analysis.
3. Nugent Brasher, "Spanish Lead Shot of the Coronado Expedition: A Progress Report on Isotope Analysis of Lead from Five Sites," New Mexico Historical Review, (winter 2010): 79-81
5. Brasher, "The Chichilticale Camp," 461; "The Red House Camp," 10, 44.
6. The name Rock-white-in-water is from Edwin R. Sweeney, personal communication with author, 21 August 2009: "The papers of Gerenville Goodwin, written in the mid-1930s, contain the place name Tsisl-Inoni-bi-yi-tu, which is the Apache name for "Stein's Pass," or Doubtful Canyon." The white rock is rhyolite lava, or ignimbrite, which is exposed in the canyon in only a single spot. This outcrop forces water to the surface as a spring.
7. Haecker, SHA, Jan 2008.
8. Our lead isotope ratio database was constructed from the following published sources listed in chronological order oldest to youngest publication date: G. L. Cumming and S. E. Kesler, "Source of Lead in Central American and Caribbean Mineralization," Earth and Planetary Science Letters 31 (1976): 262-268; T. E. Ewing, "Lead Isotope Data from Mineral Deposits of Southern New Mexico: a Reinterpretation," Economic Geology 74 (1979): 678-684; George L. Cumming, Stephen E. Kesler and D. Krstic, "Isotopic Composition of Lead in Mexican Mineral Deposits," Economic Geology 74 (1979): 1395-1407; George L. Cumming, Stephen E. Kesler and Dragan Krstic, "Source of Lead in Central American and Caribbean Mineralization, II. Lead Isotope Provinces," Earth and Planetary Science Letters 56 (1981): 199-209; J. E. Dayton and A. Dayton, "Uses and Limitations of Lead Isotopes in Archaeology," in Proceedings of the 24th International Archaeometry Symposium, eds. J. S. Olin & M. J. Blackman (Washington: Smithsonian Institution Press, 1986) 13-41; G. L. Cumming and S. E. Kesler, "Lead Isotopic Composition of the Oldest Volcanic Rocks of the Eastern Greater Antilles Island Arc. Chemical Geology 65 (1987): 15-23; Antonio Arribas, Jr. and Richard M. Tosdal, "Isotopic Composition of Pb in Ore Deposits of the Betic Cordillera, Spain: Origin and Relationship to Other European Deposits," Economic Geology 89 (1994): 1074-1093; Z. Stos-Gale, N. H. Gale, J. Houghton and R. Speakman, "Lead Isotope Data from the Isotrace Laboratory, Oxford: Archaeometry Data Base I, Ores from the Western Mediterranean," Archaeometry 37, 2 (1995): 407-415; Dorothy Hosler and Andrew Macfarlane, "Copper Sources, Metal Production, and Metals Trade in Late Postclassic Mesoamerica," Science 272 (1996): 1819-1824; F. Velasco, A. Pesquera and J. M. Herrero, "Lead Isotope Study of Zn-Pb Ore Deposits Associated with the Basque-Cantabrian Basin and Paleozoic Basement, Northern Spain," Mineralium Deposita 31 (1996): 84-92; B. M. Rohl, "Lead Isotope Data from the Isotrace Laboratory, Oxford: Archaeometry Data Base 2, Galena from Britain and Ireland," Archaeometry 38, 1 (1996): 165-180; Z. A. Stos-Gale, N. H. Gale and N. Annetts, "Lead Isotope Data from the Isotrace Laboratory, Oxford: Archaeometry Data Base 3, Ores from the Aegean, Part I," Archaeometry 38, 2 (1996): 381-390; A. Canals and E. Cardellach, "Ore lead and Sulphur Isotope Pattern from the Low-Temperature Veins of the Catalonian Coastal Ranges (NE Spain)," Mineralium Deposita 32 (1997): 243-249; N. H. Gale, Z. A. Stos-Gale, G. Maliotis and N. Annetts, "Lead Isotope Data from the Isotrace Laboratory, Oxford: Archaeometry Data Base 4, Ores from Cyprus," Archaeometry 39, 1 (1997): 237-246; E. Marcoux, "Lead Isotope Systematics of the Giant Massive Sulphide Deposits in the Iberian Pyrite Belt," Mineralium Deposita 33 (1998): 45-58; Judith A. Habicht-Mauche, "Stable Lead Isotope Analysis of Río Grande Glaze Paints and Ores Using ICP-MS: A Comparison of Acid Dissolution and Laser Ablation Techniques," Journal of Archaeological Science 29 (2002): 1043-1053; Casilada Ruiz, Antonio Arribas and Antonio Arribas, Jr., "Mineralogy and Geochemistry of the Masa Valverde Blind Massive Sulphide Deposit, Iberian Pyrite Belt (Spain)," Ore Geology Reviews 19 (2002): 1-22; J. F. Santos Zalduegui, S. Garcia De Madinabeitia and J. I. Gil Ibarguchi, "A Lead Isotope Database: The Los Pedroches-Alcudia Area (Spain); Implications for Archaeometallurgical Connections Across Southwestern and Southeastern Iberia," Archaeometry 46, 4 (2004): 625-634; Fernando Tornos and Massimo Chiaradia, "Plumbotectonic Evolution of the Ossa Morena Zone, Iberian Peninsula: Tracing the Influence of Mantle-Crust Interaction in the Ore-Forming Processes," Economic Geology 99 (2004): 965-985; Michael B. Rabinowitz, "Lead Isotopes in Soils Near Five Historic American Lead Smelters and Refineries," Science of the Total Environment 346 (2005): 138-148; M. E. Kylander, D. J. Weiss, A. Martínez Cortízas, B. Spiro, R. Garcia-Sanchez and B. J. Coles, "Refining the pre-industrial atmospheric Pb isotope evolution curve in Europe using an 8000 year old peat core from NW Spain," Earth and Planetary Science Letters 240 (2005) 467-485; Deborah L. Huntley, Katherine A. Spielman, Judith A. Habicht-Mauche, Cynthia L. Herhahn and A. Russell Flegal, "Local recipes or distant commodities? Lead isotope and chemical compositional analysis of glaze paints from the Salinas pueblos, New Mexico," Journal of Archaeological Science 34 (2007): 1135-1147; A. M. Thibodeau, S. J. Killick, J. Ruiz, J. T. Chesley, K. Deagan, J. M. Cruxent and W. Lyman, "The Strange Case of the Earliest Silver Extraction by European Colonists in the New World," Proceedings of the National Academy of Sciences 104, 9 (2007): 3663-3666; Sanghamitra Ghosh, "Heavy Stable Isotope Investigations in Environmental Science and Archaeology," Florida State University, College of Arts and Sciences, Ph. D. Dissertation Summer Semester 2008; James K. Mortensen, Brian V. Hall, Thomas Bissig, Richard M. Friedman, Thomas Danielson, James Oliver, David A. Rhys, Kika V. Ross and Janet E. Gabites, "Age and Paleotectonic Setting of Volcanogenic Massive Sulfide Deposits in the Guerrero Terrane of Central Mexico: Constraints from U-Pb Age and Pb Isotope Studies," Economic Geology 103 (2008): 117-140; Haecker, Society for Historical Archaeology, January 2008. The Coronado Expedition included members from all parts of Spain, as well as Crete, Sicily and Italy. For example, expeditionary Diego de Candia from Crete carried an arquebus. (Richard Flint and Shirley Cushing Flint, Documents of the Coronado Expedition, 1539-1542: "They Were Not Familiar with His Majesty, nor Did They Wish to be His Subjects," (Dallas: Southern Methodist University Press, 2005): 161, Appendix 3, 605-615.)
9. Robert H. Brill and J. M. Wampler, "Isotope Studies of Ancient Lead," American Journal of Archaeology 71, 1 (January 1967): 63-77; R. H. Brill, I. L. Barnes and B. Adams, "Lead Isotopes in Some Ancient Egyptian Objects," Recent Advances in Science and Technology of Materials, Proceedings (1974): 9-27.
10. Archaeometry 34, 1, (1992) 73-105; Archaeometry 34, 2, (1992) 311-317; Archaeometry 34, 2 (1992) 327-329; Archaeometry 35, 2, (1993) 241-263.
11. E. V. Sayre, K. A. Yener, E. C. Joel and I. L. Barnes, "Statistical Evaluation of the Presently Accumulated Lead Isotope Data from Anatolia and Surrounding Regions," Archaeometry 34, 1, (1992) 97; T. J. Reedy and C. L. Reedy, "Comments… IV," Archaeometry 34, 2 (1992) 327-329.
12. E. V. Sayre, K. A. Yener and E. C. Joel, "Comments on P. Budd, D. Gale, A. M. Pollard, R. G. Thomas and P. A. Williams, 'Evaluating Lead Isotope Data: Further Observations', Archaeometry, 35 (2) (1993), and Reply: Comments…I," Archaeometry 35, 2, (1993) 250.
13. This history of development of our ExplorationStation was extracted from a confidential $6,600,000 private placement offering of 1990.
14. Sayre et al., "Statistical Evaluation," 81.
15. Sayre et al., "Comments on P. Budd," 247, 249-251.
16. Hosler et al, " Copper Sources," 1819-1824; Huntley et al, "Local Recipes," 1135-1147; Thibodeau et al, "The Strange Case," 3663-3666; Haecker, SHA, Jan 2008.
17. Hosler et al, " Copper Sources," 1819-1824, esp. 1819.
18. Huntley et al, "Local Recipes," 1135-1147, esp. 1139-1140, 1144.
19. Thibodeau et al, "The Strange Case," 3663-3666, esp. 3665.
20. Haecker, SHA, Jan 2008.
21. This observation caused me to recall the harsh appraisal of Noel H. Gale and Zofia Anna Stos-Gale, directors of the Isotrace Laboratory of the Research for Archaeology at the University of Oxford: "It is certainly true that archaeological scientists as a whole do not have a good understanding of [lead isotope ratio analysis]. Interested readers might care to refer to recent attempts to redress this and to explain the bases of the subject in, e. g. Farquhar and Vitali 1989; Gale and Stos-Gale 1989; Gale 1991; Gale and Stos-Gale 1992 and 1993." N. H. Gale and Z. A. Stos-Gale, "Comments… II," Archaeometry 35, 2, (1993) 253. Gale and Stos-Gale cite the following: R. M. Farquhar and V. Vitali, "Lead isotope measurements and their application to Roman lead and bronze artefacts from Carthage," MASCA Res. Pap.Sci & Archaeol., 6, (1989) 39-45; N. H. Gale and Z. A. Stos-Gale, "Bronze Age archaeometallurgy of the Mediterranean: the impact of lead isotope studies," in Archaeological chemistry IV (ed. R. O. Allen), Am. Chem. Soc. Advances in Chemistry ser., 220, Washington (1989): 159-98; N. H. Gale, "Copper oxhide ingots: their origin and their place in the Bronze Age metals trade in the Mediterranean," in Bronze Age trade in the Mediterranean (ed. N. H. Gale), Studies in Mediterranean Archaeology, 90, Paul Astroms Forlag, Jonsered (1991): 197, 239; N. H. Gale and Z. A. Stos-Gale, "Lead isotope studies in the Aegean (The British Academy Project), Proc. Brit. Academy 77 (1992): 63-108; N. H. Gale and Z. A. Stos-Gale, "Lead isotope studies applied to provenance studies," in Nuclear chemistry and its influence on modern science (eds. D. Adan-Bayewitz, M. Artzy and F. Asaro), Stanford University Press (1993): in press.
22. Sayre et al., "Statistical Evaluation," 82; Sayre et al., "Comments on P. Budd," 250.
23. P. Budd, D. Gale, A. M. Pollard, R. G. Thomas and P. A. Williams, "Evaluating Lead Isotope Data: Further Observations," Archaeometry 35, 2, (1993) 245.
24. Brill et al, "Lead Isotopes," 10.
25. Huntley et al, "Local Recipes," 1137; Kylander et al, "Refining the pre-industrial," 468; Judith A. Habicht-Mauche, Stephen T. Glenn, Homer Milford and A. Russell Flegal, "Isotopic Tracing of Prehistoric Rio Grande Glaze-Paint Production and Trade," Journal of Archaeological Science 27 (2000): 710; Ghosh, "Heavy Stable Isotope Investigations," 27; E. Pernicka, "Comments… III," Archaeometry 35, 2, (1993) 260.
26. N. H. Gale and Z. A. Stos-Gale, "Evaluating Lead Isotope Data: Comments on E. V. Sayre, K. A. Yener, E. C. Joel and I. L. Barnes, 'Statistical Evaluation of the Presently Accumulated Lead Isotope Data from Anatolia and Surrounding Regions', Archaeometry, 34 (1) (1992), 73-105, and Reply," Archaeometry 34, 2 (1992) 313; Hosler et al, " Copper Sources," 1819.
27. Carroll L. Riley, e-mail message to author, 15 December 2009.
28. Hosler et al, "Copper Sources," 1820; Huntley et al, "Local Recipes," 1144; Thibodeau et al, "The Strange Case," 3665.
29. Thibodeau et al, "The Strange Case," 3665.
30. Kuykendall Village is defined herein as the total number of buildings mapped by Jack and Vera Mills. See Jack P. Mills and Vera M. Mills, The Kuykendall Site: A Pre-historic Salado Village in Southeastern Arizona, El Paso Archaeological Society Special Report, no. 6, ed. Vernon Ralph Brook (El Paso, Texas: El Paso Archaeological Society, 1969)
31. Franco Marcantonio, e-mail message and telephone conversation with author, 16 February 2009.
32. Brasher, "The Red House Camp," Map 2, p. 12.
33. Ibid., 7.
34. Brasher, "The Chichilticale Camp," 461; "The Red House Camp," 44.
35. Brasher, "The Chichilticale Camp," 461-62.
36. Ellery E. Frahm, "Can EMPA Identify Coronado's Musket Balls?," (http://web.mac.com/elleryfrahm/iWeb/Microprobe/Electron%20Microprobe%20Analysis%20in%20Archaeology/CFB48F79-B208-4126-A32C-BCA7A181F1C7.html)
37. Brasher, "The Chichilticale Camp," 461; Jonathan E. Damp, The Battle of Hawikku, Research Series 13, Report no. 884 (Zuni, N. M.: Zuni Cultural Resource Enterprise, 2005) 52.
38. J. F. Santos Zalduegui, S. Garcia De Madinabeitia and J. I. Gil Ibarguchi, "A Lead Isotope Database: The Los Pedroches-Alcudia Area (Spain); Implications for Archaeometallurgical Connections Across Southwestern and Southeastern Iberia," Archaeometry 46, 4 (2004): 625-634; Fernando Tornos and Massimo Chiaradia, "Plumbotectonic Evolution of the Ossa Morena Zone, Iberian Peninsula: Tracing the Influence of Mantle-Crust Interaction in the Ore-Forming Processes," Economic Geology 99 (2004): 965-985.
39. Sayre et al., "Statistical Evaluation," 82.
40. Ibid., 97.
41. Ibid., 82.