The Flotation of Waterlogged Organics: The Atocha Example By Corey Malcom Reprinted from: Astrolabe: Journal of the Mel Fisher Maritime Heritage Society Volume 8, No.1 (Summer, 1993), pp.2-7 This article is an expanded version of a paper presented at the 1988 Society for Historical Archaeology s Conference on Underwater Archaeology Recording the deposits found beneath an area of contiguous ceiling planks on the wreck of the Nuestra Señora de Atocha, 1986. In August of 1986, during the second season of fieldwork at the lower hull structure site of the wreck of the 1622 Spanish galleon Nuestra Señora de Atocha, special excavations began on the deposits located beneath four contiguous ceiling planks. This sub-excavation was to differ from other areas of the wrecksite in that the location of every single cultural item was to be recorded. The belief was that the planks had remained undisturbed since the wreck, and had formed a cap for some 80 cubic feet of wreckage. This degree of recording detail had been impractical on other areas of the site because of the presence of a large number of ceramic sherds and wood fragments, whose provenience could be more practically recorded within the areas defined by a grid system (Malcom, 1987). After the first of the planks had been removed, the work commenced as intended. A thorough mixture of bones, potsherds, and wood chips was being uncovered from the marl sea bottom and mapped, all in the hope that this data would provide some indication of the contents of the Atocha s bilge.
The work seemed to be progressing smoothly for the first few dives, but the slower digging style soon revealed problems. While it appeared the project was going well, many smaller examples in this area dense with cultural materials were being passed over. The fact that an airlift was being used as the tool for excavation was one source of the problem. This underwater vacuum cleaner made it impossible to prevent many of the smaller remains from going up the tube and being dispersed by the currents, before they were discovered. Fortunately, the bulk of the scattered material was mostly small wood fragments, bits of ceramic, and fish scales all of limited archaeological value, and a tolerable loss. As the work continued, however, bean-like seeds were found, and it became apparent that the smaller items could no longer be allowed to go. Somehow these seeds and other small items were going to have to be collected from the sea bottom along with their matrix, and later separated and identified in the laboratory. The first samples of these deposits were collected along the centerline of the ceiling timber excavation. Seventeen one-gallon plastic bags of wreck material and sediment were brought up for later processing. At first it appeared that this centerline collection had completed the sampling, but when the next plank was removed, a large concentration of the type of seed encountered earlier was revealed. Approximately nine gallons were removed from these deposits. At this point the sampling procedure was considered complete. The next step was to find a method of extracting the cultural materials from their matrix. The first technique to be tried was flotation a procedure for extracting small carbonized botanical from the soils of terrestrial features by immersing them in water and then collecting the items that float away from the heavier matrix (Struever, 1968). This Screen-washing the collected material
technique has become a standard practice on terrestrial sites, and has proven invaluable in the reconstruction of prehistoric diets. The attempt to use flotation on these deposits recovered from an underwater context revealed an obvious difficulty: the desired materials were waterlogged, and not readily buoyant. Not surprisingly, little work has been done toward the flotation of waterlogged samples. A survey of the literature revealed only one mention of an attempt at this, performed upon deposits from wetlands in Britain using a Calgon and hydrogen peroxide solution, with the conclusion that more experimentation was needed (Coles, 1984). For the Atocha samples, a solution considerably more dense than seawater was needed to carry the organic items to the surface. Tests were conducted to find this medium by placing samples of waterlogged wood collected into various solutions and simply observing what occurred. First, freshwater saturated with Calgon was tried, to no avail. A solution of sodium bicarbonate in water met with the same negative result. Upon the suggestion of a biochemist, a sugar solution was tested. This proved to be the answer, and worked as desired by keeping the wood afloat for nearly one hour. Aside from being effective as a flotation medium, sugar was safe and inexpensive. It has even been recognized as a preservative for wood recovered from underwater sites (Pearson, 1987). Much of the sea bottom in the samples was fine silt marl, which had to be cleaned away to make the process more efficient. A screen-wash system, using 1/16-inch plastic screen was devised. A plastic tub with a spigot set up off the bottom was used as a washbasin. The screen was fastened around the edges with enough slack to allow a depression in the center. The sample was poured out over the screen and freshwater was run over it, carrying away the sand and silt. The elevated drain allowed the fallout to collect on the bottom, while most of the water escaped. This fallout was collected and stored for possible study later. The larger fraction of the sample remaining on the screen was bagged and the tub prepared for the next sample. The Flotation Process. When the washing was completed, the process of flotation was begun. A oneto-one solution by volume of granular white cane sugar mixed in water was prepared in a clear plastic container. The sample of larger items was slowly poured in and when it was completely immersed, the floating remains were collected with a 1/16-inch mesh aquarium net. After the initial collection, the heavy fraction had to be stirred, allowing any lighter remains that were trapped to rise. After all of the lighter items were collected, the solution was
passed through the net as it was drained. The heavy fraction was washed in fresh water and bagged. Now the remains were ready for examination. The two components that each sample had been reduced to quite obviously contained cultural materials. Bone, stone ballast, and ceramic sherds were visible in the heavy fraction, and the light fraction appeared to consist entirely of wood and seeds. washed, but the fragile organic remains required a more gentle rinsing. This was accomplished by placing the material in a paper coffee filter and allowing clear water to percolate through. Once rinsed, the filter still containing the sample was placed on a pad of damp paper towel, allowing it to remain wet, but absorbing excess water. This arrangement was placed under an illuminated magnifier and sorted by hand. A dental pick was used to lift the more delicate objects. Most of these small items would attach to it through surface tension, without having to be poked or grasped. Detail of a Sample of the Heavy Fraction Detail of a Sample of the Light Fraction Work has been done to separate individual material types through additional chemical flotation (Bodner and Rowlett, 1980), but it was decided to sort the Atocha samples by hand. This required that the remaining sticky sugarwater be removed from the light fraction. The heavy fraction was easily The Sorting Table and Magnifier Since these items needed to remain wet to prevent shrinkage and distortion, small plastic cups of fresh water were labeled by material type for each sample number. These were placed on the sorting table, and the materials went directly into them as they were found.
When the sorting was completed, the results were well beyond what was expected. What was intended to be a way to effectively recover seeds observed on the sea floor yielded a diverse array of cultural materials. The heavy fraction contained stone ballast, bone, metals, and ceramics mixed in a hash of shells and foraminifera. The light fraction was composed of wood chips, seeds, plant remains, rosin, charcoal, hair, and insects. Not only did this procedure reveal interesting objects, it worked as planned, with the bony remains of fish scales as the only common item between the two fractions. Royal Palm (Roystonea sp.) Seeds. These bean-like seeds were readily visible underwater, and alerted excavators to proceed carefully. Chili Pepper (Capsicum sp.) Seeds After these items were recovered, the categories beyond our analytic capabilities were sent to the appropriate specialists for identification. These items were the collections of bone, seeds, insects, and hair. This sampling procedure produced a diversity of shipwreck materials in a state of preservation that surpassed all expectations. The flotation process inspired by the observation of one variety of seed has revealed nine species (Newsom, 1987). Insect Fragments. One species, Bruchus rufimanus, has been identified. Not only was the seed array a surprise, none of the other components in the light fraction, except wood fragments, was expected. When areas of organic preservation are observed on underwater sites, especially those of items too small and numerous, or too integrated within the sea bottom to recover through conventional means of excavation, the sugar flotation process is now an option. It has proved its worth on the Atocha, showing that what was to be high-detail excavation was bringing only a minor fraction of the wreck s secrets to light.
Earthenware Olive Jar Sherds Grape (Vitis vinifera) Seeds. A number of Bronze Straight Pins were recovered from the Heavy Fraction. A small, bronze medallion bearing the likeness of Nuestra Señora de Guadalupe was found in the heavy fraction. Small Ballast Stones and Gravel
Ivory or Bone Rosary Beads Carpenter s Wood Chips
This plant, of the genus Bidens was one of three inadvertently sprouted from seeds recovered through the flotation recovery technique. Unfortunately, none of these plants survived to maturity. Seed species recovered from Nuestra Señora de Atocha through flotation: Cucurbita pepo Capsicum sp. Roystonea sp. Corylus sp. Olea europa Vitis vinifera Medicago sp. Desmodium sp. Bidens sp. Pumpkin/Squash Chili Pepper Royal Palm Hazel Nut Mediterranean Olive Wine Grape Livestock Fodder Livestock Fodder Livestock Fodder
Bibliography: Bodner, Connie C. and Ralph M. Rowlett 1980 Separation of Bone, Charcoal and Seeds by Chemical Flotation. American Antiquity 45:111-116. Coles, John 1984 The Archaeology of Wetlands. Edinburgh University Press, Edinburgh, Scotland. Malcom, Corey 1987 Mapping the Nuestra Señora de Atocha. Paper presented at 1987 SHA/Conference on Underwater Archaeology. On file at Mel Fisher Maritime Heritage Society, Key West. Newsom, Lee 1987 Personal communication. Letter dated December 28. On file at Mel Fisher Maritime Heritage Society, Key West. Pearson, Colin 1987 Conservation of Marine Archaeological Objects. Butterworths: London. Struever, Stuart 1968 Flotation Techniques for the Recovery of Small Scale Archaeological Remains. American Antiquity, 33:353-362.