Example: Buttons from Fur-Trade-Era Minnesota
These buttons from the fur-trade period in Minnesota were analyzed for an interesting project that unfortunately was cancelled due to disagreement among the collaborators. Historical archaeologists know local historical societies and clubs can sometimes be excellent resources and offer a wealth of useful information. Other times, however, they can be sources of inaccurate information and unfounded lore. Local "experts" can be individuals with experience and insight, people with only "hand me down" information and collector lore, and anything in-between. The project was an attempt to determine the accuracy of local experts' identifications of fur-trade-era artifacts. In particular, these experts were to identify the composition (i.e., copper, brass, bronze, iron, etc.) of buttons. These compositions would also be determined using the electron microprobe as well as tests available to museum curators.
This project started with the two buttons shown above. The button on the right was excavated from a Northwest Fur Company's trading post on Big Sandy Lake in central-eastern Minnesota. The button on the left was found just to the east, during an excavation of the Savanna Portage trail that connected the Mississippi River and the Great Lakes. The project would eventually have included a wider variety of buttons from this period.
To keep this story professional, I will refer to my three collaborators only by title: The Professor (an archaeology professor at a Minnesota university), The Curator (a curator at a local museum), and The Student/Intern (a student of The Professor and an intern of The Curator). I was to be a fourth collaborator on the project.
So what happened? This project was initiated by The Professor, and these buttons came from the collections at his institution. Local experts would also be recruited by The Professor, and undergraduates in his lab assigned initial identifications to the buttons visually. The Curator would contribute by offering assessments based on his experience and by providing The Student/Intern access to tests commonly available to museum curators. The tests involve using indicator paper or applying nitric acid to corrosion scrapings and observing any color change. My role was to analyze the buttons using the electron microprobe to provide accurate compositions of the buttons.
Although the project was organized and funded by The Professor, The Curator started dictating how it should be conducted, including how the microprobe analyses should be done. The Professor sought permission from the State Archaeologist to polish the corrosion off small areas on the buttons. This would allow analysis of the unaltered metal without application of a conductive carbon coat. The Curator, though, insisted that the microprobe analyses must be done without any polishing. He insisted that the buttons' compositions would be sufficiently determined by analyzing through their corroded surfaces. I insisted the corrosion and altered surface layer would certainly be several microns deep, meaning that the unaltered metal would remain hidden. For a good example, see the previous post -- "A Lost (and Found) Brass Spoon" -- where corrosion completely hid zinc and tin beneath. Analyzing corrosion on the spoon, though, showed that there likely was originally a silver coating on the spoon. Just analyzing the corrosion would have provided incomplete information regarding the composition of the spoon, perhaps leading to mistaken archaeological interpretations. I suspected the same would be true of these buttons. I also offered citations of articles about surface alteration of coins as evidence to The Curator. Instead he insisted that he was an expert on analyzing metal artifacts and had "successfully" analyzed them without polishing in scanning electron microscopes.
After discussing the issue with The Professor, I agreed to try to do the microprobe analyses as The Curator had insisted, but then I would also do the analyses my way, on small polished areas. This would add another layer to the evaluation: what effect does the corroded layer have on electron microprobe analysis? The Curator and The Student/Intern would write up their report as a portion of the final paper, and I would write up both sets of microprobe analyses as separate report for the paper. The Professor would write the other portions of the final paper. The Curator, though, objected to this and continued to try exerting control. Soon The Professor killed the project.
I, though, had already started to collect the data, so I'll present my initial findings here.
The button on the right above, excavated from a Northwest Fur Company's post on Big Sandy Lake, came to me with no initial identification of its composition. I had already polished a small area on the back of this button when The Professor ended the project. Analysis of the polished area revealed that it was pure iron. Analyzing the corrosion, on the other hand, revealed aluminum (2%), silicon (6%), phosphorus (6%), and tin (8%). The incorporation of aluminum, silicon, and phosphorus in the corrosion might be explained by the burial conditions as these are geologically common elements. That is very unlikely for tin, though, which likely would have been deliberately added.
Tin within the corrosion indicates that the button originally had a bronze outer coating over an iron center. Bronze is a copper alloy with tin as the principal additive, but other elements are frequently added, namely phosphorus, aluminum, silicon, and manganese. Therefore, the detected aluminum, silicon, and phosphorus (see above) can also have come from bronze. It appears the button has a pure iron exterior and had a bronze exterior.
The button on the left, excavated on the Savanna Portage trail, has an image of a locomotive on the front (which should certainly aid in its identification by button experts). This button came to me with two labels. The first label is a standard artifact inventory tag and describes the item as "brass button with steam engine design on face." In different handwriting on the back of the tag is another assessment: "cut on front copper cover, Fe back." The second label is a Sticky Note inserted into the artifact bag. It reads: "iron back, copper face." There are absolutely no indications who made any of these identifications or wrote these notes. The mention of a "cut on front" could indicate that there was a simple scratch or wet-chemical test conducted by a student, curator, or someone else.
Analyses of the second button revealed that the front and back parts did have different compositions:
Front (Corrosion):
Cu: 94%
O: 2.2%
Zn: 1.1%
Na: 0.9%
Fe: 0.5%
Si: 0.4%
Al: 0.2%
Ca: 0.2%
K: 0.2%
S: 0.1%
Back (Corrosion):
Fe: 88%
O: 9%
Na: 0.7%
Cu: 0.4%
Si: 0.4%
Cl: 0.4%
K: 0.3%
Mn: 0.2%
S: 0.2%
Al: 0.1%
Ca: 0.1%
The front was indeed mostly copper and the back mostly iron. Many of the other elements are geologically common could have come from the burial environment of the button, and the presence of iron in the copper corrosion and vice versa is explained by cross "contamination" of the two corrosion surfaces. The only probable exception is zinc on the button's front. Brass is a copper alloy with zinc as the main additive. One percent zinc is a very low concentration in comparison to most brasses. Zinc, though, is susceptible to selective leeching out of brass as it corrodes, especially when in wet environments. The result is a copper-enriched surface. I would have to polished an area on the front to determine if it is entirely a low-zinc brass or a typical brass with a copper-rich surface. It appears that, at present, the different assessments ("brass button" vs. "copper front") are both partially accurate.
So what was learned from these exercise? First, it was a reminder than metal artifacts can be layered materials, either by design (e.g., silver plating, gold gilding) or due to surface corrosion. Second, it demonstrates, as analysis of the spoon did, that analyses of both the corroded surface and unaltered subsurface are necessary -- both can provide important clues about the original composition of the metal artifact. Third, simple scratch or chemical tests commonly available to curators can be misleading under some circumstances. The iron back of the second button was properly identified. Its front, however, was not. When there is surface alteration, such as a copper-enriched surface, the metal may be misidentified, such as brass mistaken for pure copper. Fourth, and perhaps even most important, it should be made common practice to record who made an identification of a material and its basis or reasoning (e.g., experience, visual identification based on color, scratch test, chemical reaction with nitric acid, etc.). Later archaeologists will have no idea whether "brass button" is a guess or based on sophisticated chemical analysis.
6/4/07
Added:
Electron Microprobe Analysis in Archaeology
Electron microprobe analysis (EMPA), also known as electron probe microanalysis (EPMA), is an analytical technique that combines scanning electron microscopy (SEM) and compositional analysis using x-ray spectrometry. The ability to determine structure and chemistry of samples makes EMPA very versatile. This is a dominant analytical technique in geology, but it is not as commonly used in archaeology despite similar materials in studied both fields. Here I will post about topics in EMPA, artifacts I have analyzed, archaeological studies that use EMPA, etc. If there is a topic you'd like to see posted here, please let me know.
Ellery Frahm
Doctoral Candidate, Archaeology
Research Fellow, Geology & Geophysics
University of Minnesota - Twin Cities
