Flotation Method in Archaeology

An Efficient, Low-Cost Method to Recover Artifacts, if Used Cautiously

Flotation Device in Archaeology Laboratory
Soil samples are exposed to gentle streams of water in this water screening device. Kris Hirst (c) 2006

Archaeological flotation is a laboratory technique used to recover tiny artifacts and plant remains from soil samples. Invented in the early 20th century, flotation is today still one of the most common ways to retrieve carbonized plant remains from archaeological contexts.

In flotation, the technician places dried soil on a screen of mesh wire cloth, and water is gently bubbled up through the soil.

Less dense materials such as seeds, charcoal, and other light material (called the light fraction) float up, and tiny pieces of stone called microliths or micro-debitage, bone fragments, and other relatively heavy materials (called the heavy fraction) are left behind on the mesh.

History of the Method

The earliest published use of water separation dates to 1905, when German Egyptologist Ludwig Wittmack used it to recover plant remains from ancient adobe brick. The widespread use of flotation in archaeology was the result of a 1968 publication by archaeologist Stuart Struever who used the technique on the recommendations of botanist Hugh Cutler. The first pump-generated machine was developed in 1969 by David French for use at two Anatolian sites. The method was first applied in southwest Asia at Ali Kosh in 1969 by Hans Helbaek; machine-assisted flotation was first conducted at Franchthi cave in Greece, in the early 1970s.

The Flote-Tech, the first standalone machine to support flotation, was invented by R.J. Dausman in the late 1980s. Microflotation, which uses glass beakers and magnetic stirrers for gentler processing, was developed in the 1960s for use by various chemists but not extensively used by archaeologists until the 21st century.

Benefits and Costs

The reason for the initial development of archaeological flotation was efficiency: the method allows for the rapid processing of many soil samples and the recovery of small objects which otherwise might only be collected by laborious hand-picking. Further, the standard process uses only inexpensive and readily available materials: a container, small-sized meshes (250 microns is typical), and water.

However, plant remains are typically quite fragile, and, beginning as early as the 1990s, archaeologists became increasingly aware that some plant remains split open during water flotation. Some particles can completely disintegrate during water recovery, particularly from soils recovered in arid or semi-arid locations.

Overcoming the Shortcomings

The loss of plant remains during flotation is often linked to extremely dry soil samples, which can result from the region in which they are collected. The effect has also been associated with concentrations of salt, gypsum, or calcium coating of the remains. In addition, the natural oxidation process that occurs within archaeological sites converts charred materials which are originally hydrophobic to hydrophiliac—and thus easier to disintegrate when exposed to water.

Wood charcoal is one of the most common macro-remains found in archaeological sites. The lack of visible wood charcoal in a site is generally considered the result of the lack of preservation of the charcoal rather than the lack of a fire. The fragility of wood remains is associated with the state of the wood on burning: healthy, decayed, and green wood charcoals decay at different rates. Further, they have different social meanings: burned wood might have been building material, fuel for fire, or the result of brush clearing. Wood charcoal is also the main source for radiocarbon dating.

The recovery of burned wood particles is thus an important source of information about the occupants of an archaeological site and the events that happened there.

Studying Wood and Fuel Remains

Decayed wood is particularly underrepresented at archaeological sites, and as today, such wood was often preferred for hearth fires in the past.

In these cases, standard water flotation exacerbates the problem: charcoal from decayed wood is extremely fragile. Archaeologist Amaia Arrang-Oaegui found that certain woods from the site of Tell Qarassa North in southern Syria were more susceptible to being disintegrated during water processing—particularly Salix. Salix (willow or osier) is an important proxy for climate studies—its presence within a soil sample can indicate riverine microenvironments—and its loss from the record is a painful one.

Arrang-Oaegui suggests a method for recovering wood samples that begins with hand-picking a sample before its placement in water to see if wood or other materials disintegrate. She also suggests that using other proxies such as pollen or phytoliths as indicators for the presence of plants, or ubiquity measures rather than raw counts as statistical indicators. Archaeologist Frederik Braadbaart has advocated the avoidance of sieving and flotation where possible when studying ancient fuel remains such as hearths and peat fires. He recommends instead a protocol of geochemistry based on elemental analysis and reflective microscopy.


The microflotation process is more time consuming and costly than traditional flotation, but it does recover more delicate plant remains, and less costly than geochemical methods. Microflotation was used successfully to study soil samples from coal-contaminated deposits at Chaco Canyon.

Archaeologist K.B. Tankersley and colleagues used a small (23.1 millimeters) magnetic stirrer, beakers, tweezers, and a scalpel to examine samples from 3-centimeter soil cores. The stirrer bar was placed at the bottom of a glass beaker and then rotated at 45-60 rpm to break the surface tension. The buoyant carbonized plant parts rise and the coal drops out, leaving wood charcoal suitable for AMS radiocarbon dating.