The Laboratory Excavation of a Soil Block from Sylvester Manor

TL;DR

This paper describes the laboratory micro-excavation of a large intact soil block from the South Lawn midden at Sylvester Manor and tests low-technology methods for mapping micro-artifacts, detecting protein residues, and density-separating heavy minerals. Preliminary results suggest profile mapping can distinguish recent single-component deposits from long-term depositional patterns, and that UV fluorochrome tagging and heavy-liquid density beads are promising laboratory approaches.

Key findings

• A 40 × 50 × 65 cm (D × W × H) soil block weighing approximately 800 lbs was removed intact from the South Lawn midden and excavated intermittently over one year in the laboratory.
• Direct profile mapping of particles in the 1–16 mm range (0 to -3 phi) showed coral, shell, and bone frequency decreasing with depth, matching Michie’s “single component assemblage in motion” pattern; charcoal frequency increased with depth, matching either a “multicomponent assemblage” or an “assemblage settling at the base,” though sample size made exact interpretation uncertain.
• A coumarin fluorochrome, Polyfluor YG®, prepared as a 0.125% w/v solution in ethyl alcohol, fluoresced on gelatin-treated test soils and detected a positive protein response under a butchered sheep’s scapula; the xanthoproteic test was positive.
• Fastfloat® sodium heterotungstate heavy liquid, starting density 2.95 g/ml at 25°C, with glass density beads calibrated from 1.80–2.86 g/ml, was shown as a promising method for isolating heavy minerals and iron corrosion products.

Methods

• The intact soil block was taken from the south central section of the South Lawn midden at Sylvester Manor, Shelter Island, New York, a 17th-century Dutch provisioning plantation; the buried midden covers roughly 40 m².
• In the laboratory, the block was tilted at 45° in a cradle for plan and profile excavation, with LED, incandescent, and longwave UV lighting; two 5 cm vertical slices were removed to create three profiles, the landscape layer was removed in four 5 cm levels, and the midden/A-B layer was micro-excavated in 23 1 cm levels.
• Analyses included low-power (10×) direct particle mapping on glass sheets under visible and UV light, heavy-liquid density flotation with Fastfloat® and density beads, and fluorochrome tagging with Polyfluor YG® followed by a xanthoproteic microchemical test for protein.

Limitations

• The paper states that findings are preliminary, that results are not guaranteed, that further research is needed to understand charcoal distribution and the reliability of profile mapping, and that particles extended below the sampled B horizon; sieve analyses from corresponding micro-excavation levels were still in progress.
• The paper states that sample size limitations of the soil block make the exact interpretation of the charcoal pattern uncertain, and that the protein investigation is ongoing.
• The text makes evident that the study relies on a single soil block from one site, that the micro-artifact size range used (roughly 16 mm down to 1 mm) is coarser than most micro-artifact studies and may affect results, and that the author is a conservator with relatively little field excavation experience, making the approach potentially idiosyncratic.