The Long-Term Efficacy of Conventional Treatments for Bynesian Decay

Bynesian decay—originally known as Byne’s disease (see Shelton 2008)—is the common term for the reaction of volatile carboxylic (acetic and formic) acid vapors from the natural decay of cellulose with the surface of objects that are made up of calcium compounds, such as shells and eggs. Its causes and effects have been well described in literature for some years (e.g., Tennent and Baird 1985; Callomon 2002; Shelton 1996, 2008) and it is normally the result of a suite of material and environmental factors (Sturm 2006). Its chemistry is well understood (e.g., Waller et al. 2000, Callomon and Rosenberg 2012, Cavallari et al. 2014), and treatment based on stopping the reactions involved is normally successful.

This report describes an unusually large-scale treatment program involving an entire collection that was carried out in 2011. In addition, monitoring of a batch of specimens from an earlier donation that had been treated in this fashion in 2001 reveals no recurrence of the problem over 20 years later.

In 1999, the Academy of Natural Sciences (ANSP) completed the transfer of a large private mollusk collection from the basement of the former owner’s house in Philadelphia. The collection comprised dry shells that had been stored for several decades in wooden cabinets with tightly fitting drawers. There had been no humidity control and the basement had flooded at least twice due to the failure of sewage pipes. In November 2001, one of the present authors (Callomon) confirmed the presence of Bynesian decay in shells from a specific area of the former storage. They were treated using the conventional method—first soaked in clean water for long enough to allow the formate/acetate crystals to dissolve, and then lightly scrubbed and thoroughly dried. A set of 27 treated specimens was then placed in one of the Academy’s older storage cabinets, which are made of steel with wood-based interiors and drawers (Callomon and Rosenberg 2012) to determine via long-term observation whether that environment alone would cause the decay to reoccur.

In 2010, ANSP received the donation of another large and scientifically important private collection comprising roughly 9,000 gastropod mollusk shells from a collector in Pennsylvania. These shells had also been housed until then in tightly fitting wooden drawers, in some cases for more than 30 years. Again, there had been little control of temperature and none of humidity. One author (Callomon) had some years previously confirmed that Bynesian decay was already present and had advised the owner that at least some of the shells needed treatment.

In both cases, the presence of Bynesian decay was established in situ using the classical test methods. Use of a magnifier shows that anomalous white nodules or stains on the shell surface are crystals resembling fur or snow, and a sour odor and taste indicate the presence of acetic acid.

The second collection, which is the focus of this paper, was sampled on arrival to assess the extent of the decay further. As noted in previous studies, the hygroscopic nature of calcium formate and acetate salts mean that some shells in a particular box or drawer might be affected, whereas their immediate neighbors are not, the victim having become a “sacrificial anode” that draws moisture away from the others. Nevertheless, the presence of the decay was confirmed in more than a third of the shells randomly sampled across the collection. The shells range in size from 4 mm to over 40 cm, though the largest ones (25 cm and over) showed no decay, probably because they had been stored in larger and less enclosed spaces with proportionately more air circulation.

Given the scale of the task and the constraints on time and finance that are familiar to museum collections staff, we had to design a process that would effectively treat the maximum number of specimens within the project period, which was delimited by our availability and that of the temporary space used as a drying area.

A crucial first question was whether we could reduce the amount of work by identifying and separating only those specimens needing treatment. This would require a simple but accurate test for the presence of the decay that could be applied to the entire surface of a shell of any size and not require equipment, such as a spectrometer, that was not readily available. The traditional taste test is very accurate, but violates laboratory safety and hygiene standards. Experimentation revealed that fluorescence under ultraviolet light was not reliable, as areas of natural shell erosion could give off similar light patterns to decay-affected parts. Swabbing with color-indicating compounds would risk staining the shells, as areas affected by the decay tend to have lost their outermost crystalline layer and thus are highly absorbent. Given that the salts present on the shell surface will readily dissolve in water, the simplest test would be to immerse the specimen and monitor the pH of the water. This is identical to the first stage of treatment, however, and given the known extent of the problem it was deemed most time efficient to instead simply treat the entire collection using a mass handling approach.

The amount of water that would be needed in the washing treatment—more than 340 liters (90 gallons) per day—precluded the use of deionized water, so the Academy’s municipal supply water was tested using a pre-calibrated Oakton pH-100™ series meter. Its pH was found to be 6.5, which was set as the base value for the whole program and regularly monitored. To confirm the efficacy of the traditional “soak-and-wash” method, 10 affected specimens were immersed in volumes of water proportionate to their shell lengths. After 24 hours, pH values over the 10 subjects ranged from 7.72 to 6.79, a shift towards basic indicating normal dissolution of carbonate.

We placed the entire collection in three airtight cabinets while a treatment area was prepared. This was centered around a large (133-liter/35-gallon) washing tub created by cutting in two a 242-liter (55-gallon) plastic drum with a hand saw. We lined the tub with a fine mesh net bag made from a discarded plankton net. Water was fed in at the bottom via a flexible hose weighted with a piece of ventilator brick and expelled via an overflow drain made of standard PVC plumbing pipe and fittings (Fig. 1).

The shells were individually numbered, with the number written on them and their label in pencil where possible. For the smallest shells, a paper chip bearing the number in pencil was included with them during treatment. To ensure accurate reunification of the treated shells with their lotmates and the correct label, each specimen lot was photographed with a digital camera before treatment, together with its label. The shells had already been measured by the original owner, as seen on the label (Fig. 2).

Many shells had been acquired with their chitinous operculum mounted with glue in the normal way on a wad of cotton pressed into the aperture. This was removed, the cotton discarded, and the operculae stored with the numbered label in a Ziploc bag.

To ensure penetration of the water, specimens were immersed with their apertures pointed upward (Fig. 3) for 24 hours, and each was manually agitated once during that period to dislodge bubbles. The water was then changed by running the feed hose at the rate of 3.8 liters (3 gallons) per minute at the beginning of each day for a period of 30 min, following which the specimens were soaked for a further 24 hours. Upon final removal, each specimen was lightly scrubbed with a nylon hand brush or toothbrush under running water. The periostracum, where one was present, seemed to have inhibited decay and the immersion treatment had not noticeably affected its adhesion.

A low-temperature drying oven was initially considered as a way to ensure even drying of the specimens, but the sheer volume of material, the labyrinthine internal shape of the gastropod shell, and the presence in some specimens of organic remains meant that confirming complete dryness in individual batches would be too time consuming. A longer period of air drying in an open space would thus be safer and more practicable. Larger specimens (over 50 mm length) were placed in vertical racks (Figs. 4–6) for a minimum of 5 days before being returned to the holding cabinet. Smaller specimens were placed in sealed jars with a substrate of indicating anhydrous calcium sulfate desiccant (DrieRite™ brand) for 5 days (Fig. 7).

The treatment program lasted from 22 May to 28 August 2012. The regular treatment cycle described above was carried out on 54 days during that period and a total of 3,397 specimens were treated.

Random testing of posttreatment specimens using the conventional methods described above confirmed complete mitigation of Bynesian decay. Affected areas were paler in color (Fig. 8) and less glossy than the rest of the shell, and in severe cases the color of the shell surface was obscured because of the erosion of the outermost layer (Fig. 9).

The entire collection was placed in open buffered-card trays in the Academy’s standard all-metal storage cabinets (Callomon and Rosenberg 2012) and was regularly worked on until recently by its original owner and Academy staff. Annual additions were made of newly acquired shells, but the problematic storage at the original site was no longer in use and incoming material showed no signs of decay.

In the summer of 2022, after 10 years, the entire collection was re-examined for Bynesian decay. Though many formerly affected specimens were easily identifiable by the scarring (Fig. 9) they did not show any signs of new decay. However, in a control section of the original collection comprising specimens of little value that had not been treated, decay was visible in numerous shells that had not been seen at the beginning of the study (Fig. 10). This shows that inherently hygroscopic Bynesian decay is not halted only by movement to a stable environment, even one in a lower-humidity setting.

At the same time, the 27 shells from the earlier collection that had been treated in 2001 and returned to a nominally wood-based environment were re-examined (Figs. 11, 12). They are stored in a part of the building that has uncontrolled humidity, though it is somewhat lower than outside. No trace of new deterioration was found, suggesting that it does not easily recur in treated material, even in less-than-ideal storage, as long as humidity is not unusually high for prolonged periods.

This project illustrates the distinction, well known among designers, between craft and production work. It is sometimes safer and more economical to set up a production line for, say, sorting, cataloging, and—in this case—treating material on a large scale, than to tackle lots individually or in batches. A production approach might seem to involve some waste and unnecessary effort at first, but its efficiency can yield a far higher output overall. It usually involves dedicating space, equipment, and people to a single project for a fixed period, and the limitation on time often restricts the rigor that can be applied to individual items. Even working as efficiently as possible, however, we estimate the overall labor cost of treating this collection at roughly $20,000, and that was before any cataloging could begin.

The generation of highly active shell collectors to which these donors belonged is now passing, and institutions such as the Academy of Natural Sciences are regularly being offered large collections. Institutions are committed to expanding their holdings, and particularly those of older material with reliable data, but the costs associated with cataloging, housing, and maintaining a collection continue to grow faster than do endowments and operating funds. In the present case it was known that the collection had Bynesian decay, but its unusually expansive coverage of a single family made it an important acquisition. Nevertheless, its owner did not finance the remedial work, even when the scope of the problem had become clear, and in that regard this paper could be seen as a cautionary tale. Wherever possible, arrangements for the donation of a collection should be made well in advance and should include a discussion of the likely expense.

The authors thank Andrew Bentley (Biodiversity Institute and Natural History Museum, University of Kansas) and Jean-Marc Gagnon (Canadian Museum of Nature) for thoughtful and productive comments.