J Archaeol Sci 34 (2)

Archaeomagnetic dating of samples from a brick kiln discovered at Dogmersfield Park, Hampshire, has been achieved using both field direction and intensity. Detailed rock-magnetic properties and microwave archaeointensities have already been studied and published elsewhere (Ll Casas, J Shaw, M Gich, J A Share, `High-quality microwave archaeointensity determinations from an early 18th century brick kiln', Geophys J Int 161 (2005) 653--661). The archaeomagnetic measurements were assigned time-probability distributions by comparing with predictions from a global model at the sampling site; this procedure suppresses errors arising from relocation to a common central reference location. All three probability distributions consistently indicate the same age (~AD 1700). Once the probability distributions are combined a narrower probability distribution is obtained, stressing the importance of pursuing combined (directional and intensity) archaeomagnetic analyses. The inferred age is also highly consistent with available historical evidence.

It is suggested that although a wide variety of methods have been developed in both soil science and archaeology to extract and measure soil phosphorus (P), resulting in a tremendous amount of data and a wide array of interpretations, there is also considerable confusion over appropriateness of methods and terminology. The primary purpose of the paper is to address these issues by clarifying soil P analyses. Anthropogenic additions of phosphorus to the soil come from human refuse and waste, burials, the products of animal husbandry in barns, pens, and on livestock paths, or intentional enrichment from soil fertilizer. Once added to the soil, phosphorus in its common form as phosphate is stable and generally immobile in soils. Soil P comes in many forms, organized for the purposes of the paper on the basis of extraction and measurement procedures as (1) extraction for available P (Pav); (2) portable field techniques (the spot test or ring test); (3) chemical digestion of a soil sample for total P (Ptot); (4) extractions of inorganic P (Pin) for fractionation studies and extractions to look at individual compounds of P; (5) measurements of organic P (Porg); and (6) extractions for total elemental analysis. To compare the suitability of various extractants as the `best' indicator of human input and activity the authors subjected samples from three very different North American archaeological sites (Lubbock Lake, TX; Hulburt Creek, IA; British Camp, WA) to four methods of soil P extraction: perchloric acid digestion (Ptot), sulfuric--nitric acid extraction Ptot), hydrochloric acid extraction after ignition (Pin), and citric acid extraction (Pav). Further, methods of measurement were compared via colorimetry vs. Inductively Coupled Plasma (ICP) spectrometry, and the two methods of supposed `total P' were both measured via ICP. In general, the stronger extractants yielded more soil P, but the result are not clear-cut. Likely variables include the intensity of occupation, nature of the parent material, and postdepositional weathering (for example, the addition of dust).