Most archaeologists and collections managers are aware of the “curation crisis” in the profession. It is a topic that scholars have been writing about for over 40 years. Many of the associated problems with archaeological collections—such as lost collections, lost contexts for materials, lack of space, and overall lack of care—have been thoroughly documented for decades (e.g., Bawaya Reference Bawaya2007; Childs Reference Childs1995; Kersel Reference Kersel2015; Marquardt et al. Reference Marquardt, Montet-White and Scholtz1982). Unfortunately, we will now add to this list, albeit in a relatively small way. Specifically, we highlight some of the results of a 15-year archaeochemistry project that has identified the infrequent, yet consistent, presence of unpleasant materials stored on the shelves of many different repositories. We emphasize that these are not items discovered in bags of unprocessed/orphaned collections, but rather items that originated from well-curated assemblages throughout the United States. The problem we raise is that being unaware of the specific contents on repository shelves can be a significant issue for collections management staff. The problem can range from generic (albeit extremely noxious) clean-up problems in the event of container breakage to moderate health-and-safety concerns for staff due to ignorance of contents, to extremely rare instances of potentially life-threatening issues.Footnote 1
Prior to reporting our findings, we wish to provide some context for our work and situate it as part of ongoing discussions on the relationship between archaeology and the collections that fieldwork generates. Our archaeochemistry collaboration began in response to materials recovered from a large field project in northern Idaho in 2008, where archaeologists recovered almost 600,000 artifacts (Weaver et al. Reference Weaver, Bard and Warner2014). A portion of the materials recovered had contents that required chemical analysis for identification—such as sealed bottles—leading to archaeologists reaching out to the chemistry department at the University of Idaho (Spinner et al. Reference Spinner, Lustig, Warner and von Wandruszka2011; Warner et al. Reference Warner, von Wandruszka, Spinner, Lustig, Freeman and Mitchell2014). What began as a project-specific partnership between an archaeologist (Warner) and a chemist (von Wandruszka) at the University of Idaho ultimately morphed into a much larger endeavor. Over the past 15 years, archaeologists and chemists at our institution have collaborated on an ever-expanding archaeochemistry project in which materials from historical archaeology sites throughout North America are chemically tested and identified (von Wandruszka and Warner Reference von Wandruszka and Warner2018). The project not only generates unique information for archaeologists that is regularly incorporated into reports (Campbell et al. Reference Campbell, Warner, Petrich-Guy, May, Bodley, Breen and Brooks2018; Prouty et al. Reference Prouty, Millward, Landry, Harrison and Horn2020; Swope and Grenda Reference Swope and Grenda2023; von Wandruszka et al. Reference von Wandruszka, Warner, Valentine, Kimball, Summer, Fink and Hunter2016; Voss et al. Reference Voss, von Wandruszka, Fink, Summer, Elizabeth Harman, Shapovalov, Kane, De Loney and Acebo2015) but also provides extensive training for analytical chemistry students. Since its inception, 36 students have participated in the project and analyzed more than 500 items.Footnote 2 The bulk of the items tested have been glass bottles and jars, but we have also tested slag from blacksmith shops, fabric, concrete, mortar, a tooth filling, gunpowder, and many other objects and materials.
Although our collaboration was initiated in response to a fieldwork-driven need, the project has continued based largely on testing materials that archaeologists had previously excavated and are now part of their collections. In this sense, our archaeochemistry project is now part of the growing body of scholarship on collections-based research. Given the volume of materials sitting on repository shelves—and the many reasons for emphasizing collections-based research—we strongly believe that this is an area on which archaeologists should focus their attention (e.g., Childs Reference Childs, Agnew and Bridgland2006; King Reference King2014, Reference King, Terry Childs and Warner2019; King and Samford Reference King and Samford2019; Schiappacasse Reference Schiappacasse2019; Voss Reference Voss2012).
However, as researchers expand work on archaeological collections, there can be unexpected findings unrelated to the original intent of the work. This has been the case with a number of items that we have investigated over the years. They include artifacts that could have health and safety implications for people working in repositories—or, on a more prosaic level, have been identified as disgusting materials that people may not want to put on their shelves if they knew what they were!
In addition to being part of a growing stream of collections-based research, our work also connects to present-day efforts of raising consciousness about health and safety issues in archaeology. Although the bulk of the scholarship in this area has focused on safety in the field (Klehm et al. Reference Klehm, Eifling and Peixotto2021; Meyers et al. Reference Meyers, Breetzke and Holt2021; Poirier and Feder Reference Poirier and Feder2001; Stapp and Longenecker Reference Stapp and Longenecker2009; White Reference White2012), there is also a small body of work on hazards associated with collections (Caldararo and Palmer Reference Caldararo and Palmer2008; Odegaard Reference Odegaard2019; Suits Reference Suits1998). Our work will add to this, but it should always be understood that the risks involved in these scenarios are relatively minor. In 15 years of work, we have identified one object (discussed below) that posed an immediate and serious health threat, and overall, about 6% of the artifacts that we have analyzed could be described as noxious, toxic, or disgusting. This includes materials containing toxins such as mercury, arsenic, lead, selenium, strychnine, and phosphine-producing compounds. It excludes substances that are only poisonous when used unwisely. In other instances, potential risks to personnel could arise through imprudent behavior, such as touching unknown materials or breathing their vapors. This is somewhat more of an issue with people working in archaeological labs given that lab technicians are generally not made aware of possible health risks while processing routine items such as bottle glass.
What follows is a compendium of the odious materials identified in our work with a brief summary of the investigative processes undertaken by the student chemists.
MERCURY
The toxicity of mercury (Hg) is well known. Its toxicity was established in the second half of the nineteenth century, but it was still incorporated into a variety of products well into the early twentieth century. Interestingly, some of these had cosmetic and medicinal applications. An example of a cosmetically used mercury compound was provided by a sample of Gouraud's Oriental Cream recovered from a project in Sandpoint, Idaho (Figure 1). The material, an off-white creamy substance, turned out to be mercurous chloride (Hg2Cl2, aka calomel), which was advertised as a facial cream (Figure 2). It was touted for imparting a “pearly glow” to the user's skin, as well as removing unsightly blemishes. Curiously, it was noted in the advertising that the product was tasted to ensure that it was properly made. Around the middle of the nineteenth century, calomel was also taken internally as a cure against diseases ranging from cancer to ingrown toenails. Although calomel is not acutely toxic, it is recommended that archaeologists dealing with suspected samples handle it carefully. Exposure, especially ingestion, can lead to mercury poisoning.
FIGURE 1. Gouraud's Oriental Cream bottle from Sandpoint, Idaho (Warner et al. Reference Warner, von Wandruszka, Spinner, Lustig, Freeman and Mitchell2014:281).
FIGURE 2. Gouraud's advertisement (Salt Lake Tribune, August 26, 1908, p. 11).
A second example of mercury in archaeological samples is mercuric sulfide (HgS, aka cinnabar). The sample comes from a medicine vial excavated in the 1980s as part of the Market Street Chinatown project in San Jose, California (von Wandruszka et al. Reference von Wandruszka, Fink, Summer, Harman and Shapovalov2014; Figure 3). Such vials are quite commonly recovered archaeologically. The red mineral in the vial is the common ore of mercury, and it continues to find wide use in traditional Chinese medicine (TCM) as an antibacterial agent (von Wandruszka and Warner Reference von Wandruszka and Warner2021). Although potentially more poisonous than calomel, its toxicity is limited by its low solubility in stomach acid. Archaeologists are most likely to find the material in medicine vials like the one shown in Figure 3. However, the same material is also often contained in yinni, the stamp ink used to seal documents in China for thousands of years. Again, archaeologists should minimize exposure to the material, especially breathing its dust.
FIGURE 3. Medicine vial from Market Street Chinatown, San Jose, California (photo by Ray von Wandruszka).
PHOSPHORUS
Phosphorus (P) is an essential part of living systems, but many phosphorus compounds are deadly poisons. Historical archaeologists excavating sites associated with agricultural establishments or hospitals should be especially leery of materials that were used as rodenticides. Some of the most effective ones contain phosphorus, and many of those are acutely toxic to humans. One of the most pernicious of these is sodium dihydrogen phosphide, NaPH2. This is an off-white to light-yellow material that is kept in a sealed ampule (Figure 4) because it spontaneously reacts with moisture in the atmosphere according to this reaction:
PH3 is phosphine, a highly toxic gas that kills mice and rats, but it is also very dangerous to humans. One mode of action for an operator is to carry a quantity of solid NaPH2 in a sealed container, open it in a locale such as the hold of a ship or a grain-storage facility, and immediately leave the venue. Upon opening, the material continuously reacts with the moisture in the air, emitting foul-smelling PH3 (garlic/fish odor) until it is depleted.
FIGURE 4. Ampule recovered from morgue/hospital in New England (photo by Ray von Wandruszka).
The sealed ampule with NaPH2 shown in Figure 4 was recovered from the site of a hospital/morgue in New England, where it was presumably used to keep the vermin population down. Project archaeologists actually recovered two of these ampules. Upon identification of the contents, we notified the lab manager of the significant health risks and provided detailed guidance on how to dispose of the other ampule. Archaeologists who come across an artifact of this general appearance should be extremely careful not to accidentally break or open it. If it needs to be opened, this should be done in a good chemical hood.
Other phosphides were also used to generate PH3 and kill rodents. A prime example is aluminum phosphide (AlP), which reacts with water as follows:
AlP has been (and still is, in some parts of the world) distributed in tablet form, in which the phosphide is combined with inert excipients such as silicates. Although the mode of action is, in principle, similar to that of a pure phosphide, it proceeds much more slowly. Figure 5 shows a sample of such tablets that were sent to our laboratory from a school site in St. Augustine, Florida. The tablets were not found in a sealed container, and they had long since “outgassed”—that is, they had lost all phosphorus. Because of this, they presented no danger. It is, however, conceivable that tablets in a tightly closed container could still produce PH3 when exposed to the atmosphere.
FIGURE 5. Outgassed pills from St. Augustine, Florida (photo by Ray von Wandruszka).
A third path by which potentially dangerous phosphorus can make its way into the world of archaeology is via hypophosphites. These are a suite of compounds (e.g., KH2PO2, potassium dihydrogen hypophosphite) that were sold in syrup form in the nineteenth and early twentieth centuries (Figure 6). They were part of a larger “patent medicine era” and were frequently accompanied by outrageous names and claims (Young Reference Young1961).
FIGURE 6. Pharmacy bottle from Washington, DC (photo by Ray von Wandruszka).
Hypophosphites can also form phosphine gas, but not in reaction with water. Instead, they disproportionate, especially at raised temperatures:
These compounds are somewhat less hazardous than the phosphides in an archaeological setting, but they should still be handled with caution. This is especially true for the example shown in Figure 6, because this syrup also contained strychnine (presumably as a stimulant), which is a notorious poison.
ARSENIC
Until the 1920s, when it was discontinued, Kellogg produced an insect poison known as Ant Paste in Los Angeles, California (Figure 7). It was a sugary paste, laced with arsenic (As), that had a sordid history involving murder trials and that caused the death of two children who ingested it (Willis v. State, 37 Ala. App. 185, 66 So. 2d 753 [1953], https://cite.case.law/ala-app/37/185/1696781/). A distinguishing feature of the container was a “rattle cap,” in which small shot pellets were placed in a hollow space in the lid, producing a rattling sound when the container was moved. Although this was meant to alert the user that the contents were poisonous, it may, in fact, have attracted the children.
FIGURE 7. Kellogg's Ant Paste advertisement (San Antonio Express, April 29, 1917, p. B8).
A small, unlabeled jar of ant paste was submitted to our laboratory from an excavation originally conducted in 1969. The jar is part of the Yreka Chinatown collection housed at the California State Parks’ State Archaeological Collections Research Facility in McClellan, California (Figure 8), and it contained a remnant of its original contents, including arsenic. It must be viewed as a hazardous compound.
FIGURE 8. Kellogg's Ant Paste jar from Yreka, California, Chinatown. Collection housed at the California Parks' State Archaeological Collections Research Facility (photo by Ray von Wandruszka).
PERCHLOROALKANES
A colorless glass medicine bottle with a square base containing a small amount of dark-brown sticky residue was submitted to our laboratory (Figure 9). It was recovered during property renovations of the still-standing “Cranky Sam” Public House in Missoula, Montana. Cranky Sam's was formerly associated with the city's red light district and Chinatown. The material in the bottle was found to contain hexacholoroethane (aka percholoroethane), a compound that was used as an anthelmintic in veterinary medicine.
FIGURE 9. Patent medicine bottle from Missoula, Montana (photo by Ray von Wandruszka).
In suspension with bentonite clay and water, it was used as a treatment for parasites in cattle and sheep until the 1940s. It was effective against common mature liver flukes but not against the larvae. Because of this, farmers had to frequently re-treat their livestock. In the 1920s, hexachloroethane was found to be toxic to humans and animals, primarily when it is absorbed through the skin. It is no longer produced in the United States.
URINE, NOT WHISKEY!
A sealed Iler's Malt Whiskey bottle with a remnant of a dark yellow liquid was recovered from a privy site in the state of Washington. It had been stored on the shelves of the Burke Museum in Seattle for approximately 15 years. A sample of this liquid, presumed to be whiskey, was submitted to our laboratory for analysis (Figure 10). It was found to be aqueous, with a notable content of potassium (K), phosphorus (P), and—especially—urea.
FIGURE 10. Iler's whiskey bottle from Washington State. Catalog number: 45SN409/2008/1278.1, Courtesy of the Burke Museum of Natural History and Culture, Seattle, Washington (photo by M. Caves).
These findings left little doubt that the liquid in the whiskey bottle was in fact not a distillery product but urine. Although the bottle obviously had been repurposed as a urine receptacle, its deposition was, of course, consistent with the privy from which the artifact was recovered. One may speculate that the bottle was used as a vessel of convenience, in order to avoid a nighttime trip to the outhouse.
ODOROUS OIL
The final sample we discuss is a large mason jar with a wire bail, approximately three-quarters full with a thick, dark-brown liquid (Figure 11). It was recovered from a trapper's cabin in Southern Idaho, and it was introduced by the archaeologists as a scent, “brewed” to attract animals to traps. Although this may have been the purpose of the material, when the jar was unsealed, it gave off a most unpleasant, penetrating odor.
FIGURE 11. Mason jar with trapper's “scent” from southern Idaho (photo by Ray von Wandruszka).
It was established that the liquid was not aqueous (i.e., not brewed), but rather an organic (carbon based) material. It showed signs of extensive oxidation, which accounted for the foul smell and strongly suggested that this was a rancid natural oil. Oxidative decay of this kind primarily happens with (poly)unsaturated oils, which may be of plant nature, but could, in view of the trapper's setting, also be animal oils. Mink oil and neat's-foot oil are cases in point. Whether the oil was originally produced to attract animals to traps or to treat their hides after tanning remains an open question.
CONCLUSION
Collections managers are all too familiar with the problem of inadequate funding to curate collections. This is particularly the case with historical collections. The volume and variety of materials recovered is generally substantial, and all too frequently, curation budgets are not sufficient to process the materials fully (Sullivan and Childs Reference Sullivan, Terry Childs, Terry Childs and Warner2019:81–83). One consequence of this is that objects get put on the shelves without curation staff being fully aware of what they have. In most cases, this is not a concern. Indeed, many a thesis has been produced on collections that have been sitting on repository shelves (at the University of Idaho, approximately 50% of our theses are collections based). However, as our work has demonstrated, not fully knowing what is on our shelves can lead to problems. A broken ampule of phosphide or a leaking bottle of 100-year-old urine may only lead to a nasty cleanup job, but it could be much worse. A lack of understanding of the contents of containers that “still have stuff in them” can be an unrecognized health risk for lab staff.
It is not a realistic expectation for repositories to immediately check all of their collections for bottles that may have bits of residue in them (illustrated in Figure 9). Indeed, there are probably hundreds, if not thousands, of containers of this kind on repository shelves, and the cost of testing such materials would be prohibitive (in a commercial lab setting, we estimate that the cost would be about $5,000 per artifact). However, collections managers can take some small steps to manage the possible health implications presented by such containers.
A first step is for repositories to make a point of identifying when they have containers with contents in them—although it is not enough simply to have contents noted in the comments section of a site catalogue. Instead, we suggest that repositories keep separate listings of such items and post those lists in lab workspaces with the appropriate OSHA Safety data sheet. This way, all personnel could at least (1) be aware that there are items that are not the typical empty bottles and (2) know where these objects are located. We also suggest that lab managers include some focused instructions about dealing with these artifacts. This could be as simple as instructing lab technicians not to open any sealed containers they encounter and to contact their supervisor for specific processing instructions when they encounter containers that contain obvious residue. Related to this point, lab managers should have contact information available for their local hazardous-materials disposal agencies. Finally, when practical, it is important to have container contents tested by analytical chemists so that one knows what one has, can make informed decisions about curating these items, and perhaps take appropriate steps to dispose of the contents, if necessary.
What we want to achieve with these recommendations is some reasonable awareness of, and responses to, an issue that many collections managers and lab personnel do not think about (we also note that field techs should be given training on how to properly handle artifacts with contents in them). As we have demonstrated, there are clearly items sitting on repository shelves that can be quite nasty: the arsenic in Kellogg's Ant Paste had been sitting on a shelf for about 50 years. An important step is simply to become aware that such items are in collections. Being mindful of the possible health implications of such containers should be a small but significant reminder for all lab personnel.
Acknowledgments
We would like to thank the following individuals and organizations for providing us with the materials that were included in this work: Sarah Heffner, USACE Sacramento District; Kelly Dixon, University of Montana; Barb Voss and Christina Hodge, Stanford University and the Market Street Chinatown Archaeological Project; Laura Phillips, Burke Museum of Natural History and Culture; Ruth Trocolli and Christine Ames, Washington DC Historic Preservation Office; Marc Münch, Idaho Transportation Department; Andrea White, City of St. Augustine; and Dave Valentine for his consistent contributions of artifacts for us to test. Thanks also to Heather Olson for being a sounding board for portions of this work; and our colleague Ashley Kerr for providing the Spanish translation of our abstract. No permits were required for this project.
Funding Statement
We thank the John Calhoun Smith Fund and Carl and Kirby Dyess for supporting portions of our work over the past 15 years.
Data Availability Statement
All reports on the chemical analysis of materials reported on in this article are available in the Department of Chemistry, University of Idaho.
Competing Interests
The authors declare none.
Open access
